John M. Kolyer, Donald E. Watson (Auth.) - ESD From A To Z - Electrostatic Discharge Control For Electronics-Springer US (1996)

ESD
FROMATOZ
ELECTROSTATIC DISCHARGE
CONTROL FOR ELECTRONICS
SECOND EDITION
ESD
FROMATOZ
ELECTROSTATIC DISCHARGE
CONTROL FOR ELECTRONICS
SECOND EDmON
John M. Kolyer
and
Donald E. Watson
....
~,
KLUWER ACADEMIC PUBLISHERS
BOSTONIDORDRECHTILONDON
Distributors for North, Central and South America:
Kluwer Academic Publishers
101 Philip Drive
Assinippi Park
Norwell, Massachusetts 02061 USA
Telephone (781) 871-6600
Fax (781) 871-6528
E-Mail <kluwer@wkap.com>
Distributors for all other countries:

Kluwer Academic Publishers Group
Distribution Centre
Post Office Box 322
3300 AH Dordrecht, THE NETHERLANDS
Telephone 31 78 6392 392
Fax 31 786546474
E-Mail <services@wkap.nl>
tt Electronic Services <http://www.wkap.nl>
Library of Congress Cataloging-In-Publication
A C.I.P. Catalogue record for this book is available

from the Library of Congress.
ISBN-13: 978-1-4612-8499-4 e-ISBN-13: 978-1-4613-1177-5

DOl: 10.1007/978-1-4613-1177-5
Copyright © 1996 by Chapman & Hall
This second printing by Kluwer Academic Publishers 1999
Softcover reprint of the hardcover 2nd edition 1999
All rights reserved. No part of this publication may be reproduced, stored in a retrieval
system or transmitted in any form or by any means, mechanical, photo-copying, recording,
or otherwise, without the prior written permission of the publisher, Kluwer Academic
Publishers, 101 Philip Drive, Assinippi Park, Norwell, Massachusetts 02061
Printed on acid-free paper.

Contents
Preface To Revised Edition v

Introduction vi
Chapter 1. How to Use This Book 1
Chapter 2. Basic Physics 3
Chapter 3. Fundamentals of ESD Control 14
Chapter 4. Real and Conceptual Tools from A to Z 22
Chapter 5. The Charged Device Model (CDM) 44
Chapter 6. The Static-Safe Package (SSP) 47
Chapter 7. The Static-Safe Workstation (SSW) 50
Chapter 8. ESD Troubleshooting: Illustrative Examples 56
Chapter 9. Model Specifications 77
Chapter 10. Industry Standards 118
Chapter 11. Living Without MIL-STD-1686 122
Chapter 12. Program Organization and Implementation 130
Chapter 13. Disposition of Mishandled Hardware 145
Chapter 14. Checklists, What to Buy and Do, Conclusion 148
Chapter 15. The Future 159
References 163
Appendix 167
Paper No.1. Selection of Packaging Materials for Electrostatic
Discharge-Sensitive (ESDS) Items 169
Paper No.2. Permanence of the Antistatic Property of Commercial
Antistatic Bags and Tote Boxes 175
v
vi CONTENTS
Paper No.3. Hazards of Static Charges and Fields at the Workstation 181
Paper No.4. Perforated Foil Bags: Partial Transparency and
Excellent ESD Protection 201
Paper No.5. Cost-Effective Methods ofTestingIMonitoring
Wrist Straps 207
Paper No.6. Methodology for Evaluation of Static-Limiting
Floor Finishes 218
Paper No.7. Tote Box Material: How Good Is It? 226
Paper No.8. Electrostatic Discharge (ESD) Control in an
Automated Process 233
Paper No.9. Corrosion and Contamination by Antistatic Additives
in Plastic Films 238
Paper No. 10. Controlling Voltage on Personnel 244
Paper No. 11. ESD-Control Myths, Old and New 251
Paper No. 12. ESD Testing of Silicon Wafers 258
Paper No. 13. CDM and Work Surface Selection 263
Paper No. 14. Is Your Work Surface CDM Safe? 269
Paper No. 15. Realistic Testing of ESD Materials 272
Paper No. 16. Testing Surfaces for ESD Safety 281
Paper No. 17. Humidity and Temperature Effects on Surface Resistivity 286
Paper No. 18. Packaging for High-Voltage Discharge Protection 294
Paper No. 19. Hidden Charges on ESD-Protective Packaging 299
Paper No. 20. Toward an Ideal ESD-Protective Package 305
Index 321
Preface to the Revised Edition
In the past five years, the field of electrostatic discharge (ESD) control has under-
gone some notable changes. Industry standards have multiplied, though not all of
these, in our view, are realistic and meaningful. Increasing importance has been
ascribed to the Charged Device Model (CDM) versus the Human Body Model
(HBM) as a cause of device damage and, presumably, premature (latent) failure.
Packaging materials have significantly evolved. Air ionization techniques have
improved, and usage has grown. Finally, and importantly, the government has ceased
imposing MIL-STD-1686 on all new contracts, leaving companies on their own to
formulate an ESD-control policy and write implementing documents. All these
changes are dealt with in five new chapters and ten new reprinted papers added to
this revised edition of ESD from A to Z.
Also, the original chapters have been augmented with new material such as more
troubleshooting examples in Chapter 8 and a 20-question multiple-choice test for
certifying operators in Chapter 9.
More than ever, the book seeks to provide advice, guidance, and practical ex-
amples, not just a jumble of facts and generalizations. For instance, the added
tailored versions of the model specifications for ESD-safe handling and packaging
are actually in use at medium-sized corporations and could serve as patterns for
many readers.
The new material in this edition is intended to aid your decision-making process
as directly as possible, given that you must make your own judgments under such
constraints as cost, personnel skill level, and available facilities. As in the first
edition, the essence of this book is the integration of conceptual "tools" (Chapter
4) into an overall approach to ESD control as codified in model specifications
(Chapter 9) and, now, in policy and program documents (Chapter 11).
The frontispiece, new for this revised edition, represents the essence of ESD
damage: a generalized electrical field or air discharge strikes an edge contact of
a board, runs along a circuit line (in white), and figuratively explodes the heart
of a device. To prevent this microdisaster is the object of all our procedures and
paperwork.
We hope our book will prove helpful, and we wish you success in the ever-
evolving, and always challenging, field of ESD control.
vII
Introduction
This book does not pretend to be an encyclopedic test, which would have to be a
compilation of contributions from experts in various facets of ESO control. It is
closer to being a handbook, but we feel that the most accurate descriptive term is
an approach: a coherent, logical, and cost-effective system. As the building blocks
of this system, about fifty real and conceptual tools from A to Z appear in bold
type throughout the text of the book in order to emphasize their usefulness and to
provide an organizing principle. Insofar as "A to Z" hints of thoroughness, the
thoroughness is not in the length of the A-to-Z list but in the applicability of our
approach to all ESO-control situations.
The first edition of ESD from A to Z differed from the self-published edition
of 1989 in that various improvements were made in the test; some of the papers in
the Appendix were condensed to remove extraneous data, and an extensive index
was added to make the book more useful. This revised edition adds much new
material.
The book is arranged as follows. First, Chapter 1 tells how to use the book, and
Chapters 2 and 3 give basic principles and fundamentals. Then the real and con-
ceptual tools are listed and explained in Chapter 4, and three of them are of such
importance that they are given their own chapters: Chapter 5 for the charged de-
vice model (CDM), Chapter 6 for the static-safe package (SSP), and Chapter 7 for
the static-safe workstation (SSW). Since the best way to teach is by examples, the
tools are put to use in Chapter 8 on ESO troubleshooting; this chapter is a collec-
tion of actual case histories illustrating the flexibility of our approach. This brings
us to the bottom line: codification of our ESO control methods into specifications
for packaging and for in-plant handling (Chapter 9) and into a program of maxi-
mum simplicity and minimum expense (Chapters 10, 11, and 12). The important
topic of disposition of mishandled hardware is discussed in Chapter 13. Finally,
Chapter 14 wraps up the book with checklists of what to buy and do and a conclu-
sion summarizing the elements of our approach, and Chapter 15 gives specula-
tions of the future of ESO control.
References are listed at the end of the main text, before the Appendix. As an ex-
ample of the numbering system, Ref. 4-3 is the third reference cited in Chapter 4.
The Appendix contains twenty of our published papers or articles which give
experimental justification for our rules and techniques as well as details of test
methods and discussions of special topics such as tote boxes and the permanence
Ix
x INTRODUCTION
of the antistatic property of antistatic materials. These publications are referred

to throughout the text.
ESD control is notoriously controversial, and every subject, e.g., workbench
surfaces, wrist-strap monitors, ionization systems, and protective bags, has its
warring factions of opinion. We've never been shy about taking positions, and this
book expounds our particular ways of doing things. Our approach is a combina-
tion of logic, worst-case but realistic testing, and simple, effective ESD-control
procedures which are supported by extensive data in the Appendix and have proved
themselves over many years of experience. Our object is to err on the "overkill"
side of protection while still being cost-effective. We emphasize (and spend money
on) the essentials and deemphasize (and save money on) the nonessentials. "Buy
the basics, forget the frills." This attitude makes our approach especially suited to
small companies with limited ESD-control budgets.
Throughout the text, we use "ESDS (electrostatic discharge-sensitive) item" as
a generic term including devices, subassemblies, and assemblies. Our program is
designed for devices sensitive to 100 V by the Human Body Model (HBM). Sepa-
rate programs for different sensitivity classes within a single manufacturing area
are impractical, in our opinion, so we recommend that the methods of this book be
applied to the whole operation if any devices which are Class 1 by MIL-STO-1686
are being assembled, stored, or shipped.
In keeping with the practice of publications in which the papers in the Appendix
were printed, trade names are given only for test equipment. ESD-protective mate-
rials and equipment may be found in buyers' guides, such as those published by
Evaluation Engineering Magazine. The test equipment cited by trade name is en-
dorsed for its function and design, but we imply no claims about durability, haz-
ards, etc., nor do we suggest that patents be violated in buying or using the
equipment. Giving brand names for other that test equipment would be futile be-
cause some products would be obsolete by the time the reader sees this book. In
fact, the ever-changing nature of ESD-control products makes it necessary to ap-
pend a "living list" of them, Reference Document 4, to Model Specification 1 in
Chapter 9.
Please read and follow Chapter I, "How to Use This Book," before delving into
details. We wish you good luck with your program, but luck really shouldn't be
necessary in what purports to be a systematic art if not quite a science. Armed with
the arsenal of real and conceptual tools in Chapter 4, you'll solve ESD problems
unique to your processes. Then, with well-trained personnel and a well-organized
and well-managed program (Chapter 12), incidents of mishandled hardware will
be rare. Under such conditions, ESD control can even be fun!
Chapter 1
How to Use This Book
First, take an hour to skim through the Appendix and read the abstracts and con-
clusions carefully. Note the locations of detailed infonnation, e.g., description of
test methods, for future reference.
Next, study Chapters 2, 3, and 4. Chapter 4 gives the Basic Rule and the key
concepts, including damage mechanisms, from which all else follows. When you
come to the Charged Device Model (CDM), static-safe package (SSP), and static-
safe workstation (SSW), stop and read Chapters 5, 6, and 7. Remember that every-
thing we're doing is aimed at detecting, avoiding, or preventing the two hazards of
fields and discharges.
Study Chapter 8 to see how unique problem situations can always be handled by
using the gamut of ideas and methods from Chapter 4. As problems arise, consult
this chapter again to refresh your memory.
Use the model specifications in Chapter 9 as the basis for your own, adding
more detail as necessary.
Use Chapters 10, II, and 12 as a guide when you're setting up a new program or
improving an existing program.
Note the approaches in Chapter 13 for the disposition of mishandled hardware;
you'll face this problem sooner or later.
Use the check lists in Chapter 14 when you're buying equipment, solving prob-
lems, initiating a program, etc.
Read Chapter 15 on the future for general infonnation.
In summary, learn the location of infonnation in the book so that it can be an
ever-useful guide for troubleshooting, material/equipment buying, specification-
writing, and program management.
When the book has been thoroughly understood, the reason for each provision
of the model specifications in Chapter 9 should be evident. The reader should see
how every rule, though based on limited data and more or less empirical, is in-
tended to implement the Basic Rule. Remember that control measures may be
primary (essential) or secondary (redundant or backup); for example, in Model
Specification I, requirement paragraph 3.6.11 for the CD Rule is a primary pre-
caution, but 3.6.12.2 for unnecessary touching ofleads and 3.6.12.3 for unnecessary
1
2 ESD FROM A TO Z: ELECTROSTATIC DISCHARGE CONTROL FOR ELECTRONICS
triboelectric charging by body movement are secondary precautions. However, all

regulations, whether primary or secondary, are meant to reduce the chance of a
damaging discharge and to minimize fields and keep ESDS items out of intense
fields. We emphasize again that as a whole system of ethics flows from the golden
rule, our whole system of ESD control flows from the Basic Rule. All personnel
should be taught this philosophy because people are inclined to disregard rules
they don't understand. A successful program requires an enlightened work force.
We recommend that your personnel either study this book or be trained with course
material based upon it.
We emphasize that this book cannot be a simple "instruction manual" because
ESD-control programs must be tailor-made to fit circumstances; see the Conclu-
sion section at the end of Chapter 14. Therefore, you must absorb all the informa-
tion and advice given herein and recast it into your own personal approach, which
will agree with ours in outline but differ in details. The bottom line will be your
own specifications derived form the models in Chapter 9. See the real-life ex-
amples of tailored specifications in Chapter 11.
Chapter 2
Basic Physics
Purpose of This Chapter

This chapter will briefly highlight the simple electrostatic principles behind ESD
damage and control. The usual diagrams filled with plus and minus signs are omit-
ted because you can see them in treatises on electrostatics such as Ref. 2-1 through
2-4. Incidentally, you'll find the history of static electricity in these books espe-
cially interesting, dating as it does from the discovery of charging of the plasticlike
amber ("elektron" in Greek) by materials such as cat fur at the opposite end of the
triboelectric series (see Chapter 4).
Static Electricity and Electrons

Static electricity is electricity that is "static," meaning motionless instead of flow-
ing. A surface becomes statically charged when a rubbing or separation process
causes the material to gain electrons (a negative charge) or lose electrons and suf-
fer an electron deficit (a positive charge). When two dissimilar surfaces come into
microscopic, molecular-level, intimate contact, the more electron-desiring surface
"steals" electrons from the surface with less electron affinity; rubbing may simply
increase the number of small areas in close contact (Ref. 2-5). Generally, the more
dissimilar in electron affinity the surfaces (the triboelectric series again), the more
the charging.
Note that metals charge by rubbing as do nonconductors, as can be observed by
static meter if metallic disks are held by insulating handles and stroked together.
We tend to think of metals as not triboelectrically charging because they usually
are grounded so that the charge runs off almost instantaneously.
Now, what is an electron, and why can it "rub off' and transfer between sur-
faces? An electron is an elementary particle that is conceived as orbiting around
the far larger and heavier cluster of nuclear particles (protons and neutrons) of an
atom. An analogy is a tiny, man-made satellite racing about the earth. Given enough
velocity, the satellite could fly off into space, and similarly the electron can exist
independently away from its nucleus. This independent nature of the electron is
3
especially prominent in highly conductive materials, e.g., metals, over which and
through which electrons roam freely. In contrast, electrons cannot travel through
or over nonconductors such as common plastics (unless voltages are very high, as
will be discussed later), so "pools" of electrons--or, conversely, areas of deple-
tion-lie stagnant on nonconductive surfaces after triboelectrification (from Greek
tribein, meaning "to rub").
At relatively low voltage, under approximately 2000 V, or 2 kV, these stagnant
electron pools, or "dried-out" areas yearning to be moistened, cannot flow, and
they make their presence known only by an electrical or electrostatic field (E field),
which is either positive or negative. This field is "a voltage gradient between two
surfaces" (Ref. 2-6) and is a force field that pushes or pulls electrical charges. If
one surface is a charged plastic tote box, for example, and the other surface com-
prises the floor, ceiling, and walls of the room, the field will be intense only near
the box. Then, a particle in that field will be polarized, with the electrons gathered
on the side of the particle toward the box if the field is positive or away from the
box if the field is negative. Polarization results from electrons forcefully repelling
one another across space, or else from an electron deficit just as forcefully tugging
at distant electrons.
Indeed, individual electrons are analogous to misanthropes. They can be forced
into a crowd but are anxious to disperse. Thus, if a charge (quantity of electrons) is
deposited on a conductor, the electrons almost instantly scatter over the surface to
give a uniform distribution, with each particle as far as possible from its fellows.
These antisocial particles not only wish to avoid each other but yearn to hide away
in "holes" (mobile electron vacancies) in semiconductors or in the safe harbor of
the outer shell of any electron-hungry atom.
The Water Analogy

We have spoken of electrons resting in pools or flowing as if they were water and,
in fact, the water analogy is quite useful at low voltages (less than 2 kV or so).
"Electricity is really rather simple, because it behaves like a liquid. It flows like
water." (Ref. 2-7). Thus, the voltage or potential gradient between two charges
(quantities of electrons, expressed in coulombs) is paralleled by the difference in
height of the surfaces of water in basins of the same dimensions. 1\vice the quan-
tity of electrons means twice the height of water in the basin analogy.
Inside MOSFETs, a voltage gradient of typically 100 V across the gate will
punch through a very thin layer, typically 1000 angstroms, of silicon oxide. In Fig.
1, this oxide layer is represented by a dam, and cement and silicon oxide are coin-
cidentally both of the nature of brittle ceramics. The water level in the reservoirs
on both sides of the dam is the same (same voltage) in Fig. 2-1, so the fragile dam
is unstressed.
But now the finger of a negatively charged person touches a MOSFET lead to
dump in water (electrons) by the HBM (discussed in Chapter 4) as shown in Fig. 2-2.
BASIC PHYSICS 5
Electrons
Fig. 2-1.
Fig. 2-2.
Fig. 2-3.
Fig. 2-4.
What happens? The water level on one side rises to break the dam (Fig. 2-3); the oxide
is punctured and the MOSFET has failed.
So much for the HHM. Now, let's consider the FIM, also described in Chapter
4. As shown in Fig. 2-4, stagnant pools of electrons on an approaching noncon-
ductor, e.g., a plastic circuit board, are repelling the electron pools in the reservoirs.
BASIC PHYSICS 7
The nearer reservoir pool surges higher, because the E field is more intense nearer
the pools on the charged plastic, and again the dam breaks as in Fig. 2-3-another
MOSFET failure. This creation of a voltage differential by an E field is called
"induction. "
The last but not least damage mechanism to be illustrated by the water analogy
is the CDM, described in Chapters 4 and 5. As seen in Fig. 2-5, stagnant pools of
COM
Field
Fig. 2-5.
COM
Field
,
/
.' /;
/
Fig. 2-6.
electrons lie near the rim of one reservoir, representing charges on a nonconduc-
tive part of a device, e.g., the plastic case of a dual-in-line package (DIP). The E
field, being motionless, causes no surge, and the slight difference in water heights
seen in Fig. 2-5 cannot crack the dam. However, when a grounded finger touches
a DIP lead and draws off electrons, which are always anxious to rush to ground
(Fig. 2-6), a large water height differential results, and again the dam crumbles as in
Fig. 2-3.
Note that the analogy works backward for opposite charges. A positive finger in
Fig. 2-3 sucks water out, the water on the right instead of the left side surges up in
Fig. 2-4, and the grounded finger dumps in electrons in Fig. 2-6.
The morals of this water analogy are: (I) a charged finger (HBM) or tool (MM,
see Chapter 4) should not be allowed to transfer electrons to or from ESD-sensi-
tive devices, (2) an intense E field should not be brought near an ESD-sensitive
device, and (3) leads of charged ESD-sensitive devices should not be grounded by
touching them with a finger or other conductor.
The water analogy is useful for training operators. Just remember to stress that
electricity is a sort of magical blue fluid that, unlike water, forcefully repels itself
while, conversely, desiccated regions attract the fluid. Of course, this analogy, like
any other, cannot be stretched too far and, in this case, "too far" means high voltage.
High Voltage
Above 3 kY, as an arbitrary dividing line, electricity is no longer "really rather simple."
Charges on nonconductors are not merely pools radiating a force field but are able
to discharge through the air by ionizing it. A large-scale example of a high-voltage
discharge is lightning, and a small-scale example is the spark thrown from a finger
to a doorknob. Furthermore, electrons at high voltages can move over nonconductive
surfaces to some extent and even punch through layers of material when the dielec-
tric strength is exceeded. The world of high voltage is a strange one.
To appreciate the nature of high static voltages, e.g., 40 kY, triboelectrically
charge a slab of rigid polystyrene packaging foam by stroking it on fabric or car-
pet. Bring the foam near the back of your hand, and you'll feel the hairs lift be-
cause a charge opposite to the charge on the foam is induced on the hairs, and then
opposite charges attract each other. If the foam is negatively charged, electrons are
driven off the hairs to leave them positively charged, and vice versa. Induction, as de-
monstrated here with hairs on the hand, is the CUlprit in FIM damage as we have seen.
Now we come to a more dramatic effect, the air discharge. When an NE-2 neon
bulb with one lead grounded is brought within 6 inches of the highly charged
foam, you will hear a "snap," and the light will flash; an invisible "lightning bolt"
has traversed half a foot of space. If the bulb, which flashes at 80 Y, had been a
MOSFET, it would have been destroyed.
Note that induction precedes such a discharge. Electrons are driven off into
ground when a negative charge approaches a grounded conductor-or electrons
BASIC PHYSICS 9
are drawn from ground when the charge is positive-and when the gap becomes
small enough, ionized air carries current as a spark.
You'll find that you can bring a grounded metallic plate with surface parallel to
the foam surface to 0.5 inch from the foam with no discharge, whereas a crackling
discharge will occur as far as 6 inches from a corner of the plate. The sharp corner
is acting as a lightning rod by concentrating the lines of force of the E field and
ionizing the air (corona discharge). Saint Elmo's fire is the corona glow on tips of
a sailing ship's rigging.
With a static meter, you can see the charge induced on an electrically isolated
plate, e.g., a 1 ft2 sheet of lin-inch aluminum mounted on a base of polystyrene
foam, when charged foam is brought near the plate. If negatively charged foam is
held in front of the plate, a meter behind the plate will show a negative charge
because electrons were driven to the back of the plate, and this charge will subside
when the foam is withdrawn. However, if the foam discharges to the isolated plate
by being brought near a corner, the plate will hold a negative charge that fades
slowly as positive air ions are attracted to neutralize the charge. This charged plate
is a capacitor, with the dielectric being air and the other plate, or plates, being a
tabletop, the walls of the room, and so forth. Such a charged conductor could
discharge to an ESD-sensitive item to cause damage by the MM mechanism (Chap-
ter 4). If a charged person were the capacitor, the mechanism would be the HBM
which, like the MM, is a special case of DI (Chapter 4).
By the way, all the triboelectricity (transferred electrons) on a person charged by
walking resides on his shoe soles. If the soles "steal electrons" from a carpet, the
negative E field on the surface of the shoe soles drives electrons away from the
skin of the sole of the foot, so that the sole of the foot becomes positive while the
rest of the skin becomes negative (Fig. 2-7). This is, of course, another example of
+ + + + + +
""
. ""
""""
"""""""""""""
""""
"" ' ",
'"
Carpet
Fig. 2-7.
induction. If a spark is thrown to ground by the person's finger, the minus signs
above the foot in Fig. 2-7 vanish (electrons have run off to ground), leaving the
positive charges still trapped on the sole of the foot. If the person stepped out of his
shoes and onto an insulating surface, his whole skin would be positive. This pro-
cess is called compound induction, and isolated conductors charged by compound
conduction are an ESD hazard (see Paper 19 in the Appendix). Meanwhile, the
negative charge on the nonconductive shoe sole persists because it is, you will
remember, static (immobile).
Now, ground your formerly isolated metal plate, and a static meter held behind
it will show no charge at all when a charged piece of foam approaches the front of
the plate. You'll find that a grounded metallic screen will work, too. The grounded
plate or screen provides shielding, which is one of the tools of ESD protection
discussed in Chapter 4.
From the above experiments, you can imagine how much the ESD hazard is
aggravated at low relative humidity, e.g., 10-20% , when voltages run high and
synthetic fabrics and other materials crackle with air discharges. Unprotected ESD-
sensitive items are at great risk, like bystanders caught in the crossfire of a riot.
Low-voltage H8M, FIM, and CDM are enough to worry about without DI from
all directions! However, remember that humidification will only lessen the ESD
problem, not eliminate it, as pointed out in Chapter 4.
Now, let's turn to the sources of static charges, of which common triboelec-
trification, e.g., an operator's clothing accidentally stroking a plastic tote tray is
only one. All possible static generators, some subtle and unexpected, must be iden-
tified in your workplace.
Static Generators
Contact Charging and Triboelectric Charging. As mentioned above, rub-
bing may merely increase contact area, so these two charging phenomena are hard
to disentangle. For a brief but in-depth discussion of contact charging, including
the concepts of Fermi level and valence bands versus conduction bands of elec-
trons, see Ref. 2-8. Among many practical examples of triboelectric charging is
the peeling of insulative tape from circuit boards, which can generate several kV.
Also, a nonconductive liquid can charge a solid, e.g., high-purity water impinging
on a silicon wafer (one remedy in this case is to add carbon dioxide to the water to
make it conductive).
Induction Charging. We have seen this phenomenon in the FIM (water anal-
ogy, above) and shoe-sole charging (Fig. 7).
Freezing. Large voltages have been seen between ice and water phases during
freezing. A theory is that "splinters" ejected from a freezing drop are charged op-
positely to ice particles remaining in the drop. This process electrifies clouds in
thunderstorms (Ref. 2-8).
BASIC PHYSICS 11
Particle Beams. Particles can be charged by bombarding them with other

particles such as electrons, ions, or atomic fragments like alpha particles. When
air is bombarded, nuclear air molecules (positive ions) and free electrons are
created. Then the electrons combine with other air molecules to give negative ions
(Ref. 2-8). The result is a stream of air with ions of both polarities.
Radioactive Decay. This is the source of alpha particles used in nuclear air
ionization, as mentioned above.
X-Rays. Particles also can be charged by bombardment with X-rays, which

are a form of radiation similar to light but of shorter wavelength.
Grit-Blasting. This instance of charging by relatively large particles is only a

special case of triboelectric charging but is worth special mention because it can
wreak ESD havoc (Example 25, Chapter 8).
Spray. By the "balloelectric effect," atomizing a liquid by spraying gives a

cloud or fog of charged drops. One explanation is that ion distribution in the bulk
liquid is irregular, so fine drops may capture excesses of either positive or negative
ions. Another explanation is that a dipole layer on the surface of the liquid binds
negative ions to its underside; then, spraying "peels off' this layer to give negative
droplets and leave positive charges in the spray container (Ref. 2-8). Canned cryo-
genics, used for troubleshooting and environmental testing with microelectronics,
not only may emit charged droplets but may cause condensed water droplets to
freeze on the surface and cause more charging.
Thermionic Emission. Hot filaments can emit positive ions (Ref. 2-8 and
Example 19 in Chapter 8). Beware of this effect with heat guns that blow hot air.
Photoelectric Effect. Photons ("particles" of light) may strike a surface to

eject electrons and leave a positive charge (Ref. 2-8). See Example 27 in Chapter 8.
Field Emission. An E field polarizes particles and, if the field is intense enough,
electrons may be lost from the negative end of the particle while positive ions may
be extracted from the positive end. If more electrons than positive ions are lost, the
net charge on the particle is positive, and it will be drawn in the direction of the
field (Ref. 2-8). The "direction of the field" is by definition the direction in which
a positive charge is forced to move.
Corona. This is the basis of electrical ionizers in which air molecules become
charged (ionized) in the vicinity of sharp, high-voltage emitter points.
Deformation and Cleavage (Piezoelectric Effect). Piezoelectricity was dis-

covered by the Curie brothers in 1880. Deformation of a solid can cause surface
charges because of reorientation of polar molecules (when crystal symmetry is

present) or displacement of ions (Ref. 2-8). In fact, a novel ESD simulator uses
hand pressure, through a lever, to deform piezoelectric crystals and generate 2-5
kV (Ref. 2-9). An example of charge creation by deformation of noncrystalline
material is the radial stretching of vinyl tape in a wafer dicing process; ESD-sensi-
tive dies affixed to the tape are vulnerable to CDM damage.
Hot, Blowing Air. Our tests showed no charging above 100 V in Example 20,
Chapter 8. In careful experiments, a blast of shop air gave very low voltages on
metal surfaces, presumably because of triboelectric charging by submicroscopic
particles, while filtered air caused no detectable charging (Ref. 2-10). Kinetic en-
ergy calculations indicate that pure gases cannot charge surfaces.
Evaporation. "Cleaning or drying fluid by evaporation" is listed as a "typical

prime charge source" in Ref. 2-6. Evaporation of solvent from coatings is an ex-
ample. You should check such processes with your static meter.
ESD Is a Surge
Electrostatic discharge has been called a "spark," sometimes with the admission
that the only "spark" is a high-temperature breakdown of gate oxide, metallization
melting, or other internal device failure. We are concerned, however, with the ex-
ternal ESD events that lead to internal device failure. If "spark" is defined as an air
discharge, the FIM obviously is sparkless, as is the HBM at low voltage, e.g., 1
kV, while the CDM mayor may not involve a spark (see Chapter 5).
To prove that the HBM can be a sparkless surge, don't ground your wrist strap
but attach it to the lid of a Static Event Detector™ (3M Co.), sensitive to 80-100 V,
with the body of the detector grounded. Sit in a chair, shuffle your feet on a carpet,
and quickly lift your feet. You have charged yourself (by induction from the tri-
boelectric charge on your shoe soles) and, as your feet rise, your capacitance
falls, your voltage surges, and the detector is tripped. A MOSFET, represented by
the detector, would have failed. Obviously, there was no gap for a spark to jump
between you and the detector lid. A surge was to blame.
Clearly, ESD, in the sense we use it, is a voltage surge that mayor may not
include a spark (air discharge). Indeed, in the failure of voltage-sensitive devices
like MOSFETs, the minimum required current (electron flow) to cause damage is
extremely feeble and far below sparking level.
Conclusion
To summarize, static electricity is a "self-repelling fluid" that creates a force field
in space. This field induces charges on nearby conductors by polarizing them but
is blocked (shielded) by a grounded conductive sheet or screen. The static electrical
BASIC PHYSICS 13
"fluid" lies in "stagnant pools" (negative charge) or is conspicuously lacking in

"dried-out" areas (positive charge), on nonconductors at low voltage (arbitrarily
less than 3 kV). However, the "fluid" can escape by leakage through the air over
relatively long paths at high voltages, e.g. 40 kV, greatly exacerbating the ESD
problem.
In essence, static electricity is easily conceived as a matter of attraction and
repulsion, though its sources, e.g., the balloelectric effect, may be hard to explain.
From a relatively few basic concepts flow all the damage mechanisms and defense
strategies of ESD control, as you will find as you continue reading this book.
Chapter 3
Fundamentals of ESD Control
This chapter is reprinted, by pennission, from the Technical Record of the Expo
'92 International Conference on Electromagnetic Compatibility, sponsored by EMC
Technology Magazine, Reston, VA, May 18-22, 1992, pages 154-161.
Introduction
First, let's define "ESD." Electrostatic discharge (ESD) means a discharge (flow
of electrons) to or from a charge (deficit or surplus of electrons) that fonnerly had
been static (immobile).
Electrons are immobile under two conditions: (1) They reside on a nonconductive
surface, e.g., common plastic, over which they can't flow, or (2) they are trapped
on a conductive but electrically isolated object, e.g., a screwdriver blade with an
insulating plastic handle.
Now, if that screwdriver blade, bearing a sufficient charge, is brought near the
lead of an integrated circuit (IC), a spark jumps (if the voltage is high enough) and
we have an ESD. Or if an ungrounded electrically isolated person becomes stati-
cally charged by walking on a carpet and then throws a spark to a doorknob, we
have another ESD.
Even if the ESD is imperceptible and sparkless, it may still damage a highly
sensitive Ie. In any case, an ESD is fonnerly static (resting) electricity in motion,
and this motion, like the swing of a wrecking ball, can do damage and must be
controlled.
Threats
How do we control ESD? First, let's review the threatening fonns it may take.
These fonns are called ESD damage models, and Fig. 3-1 is a diagrammatic sum-
mary of the major ones. For a full discussion, see books such as Ref. 3-1.
For simplicity, Fig. 3-1 uses a silicon oxide layer, as on metal oxide semicon-
ductor field-effect transistor (MOSFET) gates, to represent all ESD-sensitive device
14
FUNDAMENTALS OF ESD CONTROL 15
elements, but these also include PN junctions, thin resistive films, and piezoelectric
crystals (Ref. 3-2). The destructive "zap" involves almost no current for voltage-
sensitive devices but more considerable current for current-sensitive devices.
With reference to Fig. 3-1 for definitions of abbreviations, the person-to-door-
knob ESO previously mentioned is an example of the HBM and also, if the knob
is not grounded, the FM. In the FM, the capacitance, or electron-sink ability, of
the knob (or of circuitry on a board) is a substitute for ground and accepts or
provides electron.
DI
PLATE
CONDUCTIVE
- FFB
SURFACE
~
CDM J}- ~
~
OR mIF
SYMBOLS ABBREVIATIONS
m OXIDE LAYER w,.PUNCTURE IN z

HBM a
DELTA VOLTAGE
HUMAN BODY MODEL
ril [?
AND
CONTACTS TO
OXIDE LAYER
~ RAPID
DISCHARGE
MM • MACHINE MODEL
FM s FLOATING MODEL
FIM ~ FIELD INDUCED MODEL
-L PUNCTURED COM • CHARGED DEVICE MODEL
- GROUND ~ OXIDE FFB = FIELD FROM BOARD MODEL
.
~ (OBVIOUS E = ELECTRICAL (E) FIELD
...L. FAILURE) LF • LATENT FAILURE
CAPACITANCE DI DIRECT INJECTION
~
Fig. 3-1. Threats.

The screwdriver-to-IC-Iead ESD mentioned above is an example of the MM

and also, if the device is ungrounded, the FM.
The 8BM, MM, and FM are all variants of 01. In Fig. 3-1, the two minus signs
(which could as well be plus signs) mean that the discharge has charged only one
side of the device so that there is a delta voltage across the silicon oxide layer.
When this delta rises much above 100 V, the dielectric strength of l000-angstrom
oxide is exceeded and a puncture or destructive "zap" occurs.
Like 01, the COM involves a discharge, but this time the device itself is charged
rather than an external conductor. The discharge unbalances the charge on the
device (one minus sign vs. two in Fig. 3-1) to create the fatal delta voltage.
So far, a discharge has caused the trouble, but an electrical (E) field is also a
danger. In the FIM, the E field emanating from a plate or other conductive object
brought near the device polarizes the device so that, especially if one side is
grounded, a delta voltage across the oxide results. Or the ungrounded device may
be immersed in the E field first and then grounded later to cause the delta voltage.
In the FFB (Ref. 3-3), the E field is on the same circuit board as the device; this
happens when the board becomes triboelectrically charged by rubbing against a
surface such as a stainless-steel oven shelf. Then a COM-like discharge does the
damage.
Note that conductive surfaces participate in the discharge in the COM or FFB.
We'll say more about this under "Defenses' below.
Whatever damage model is to blame, devices "wounded" too subtly for detec-
tion by routine tests may "die an early death." Such latent failure (LF) is rare but
real, and it must be considered in rigorous ESD-control programs protecting highly
sensitive, expensive, or mission-critical items. For an excellent review of LF, see
Ref. 3-1.
Behind the variety of damage models in Fig. 3-1 lie just two basic threats: fields
and discharges. If fields are weak enough and if discharges are feeble enough,
there can be no ESD damage. Now the question becomes: How do we tame these
twin threats of fields and discharges?
Defenses
Basic defenses are diagrammed in Fig. 3-2. Design protection comes first, and
grounding the input through a diode shows the principle. In this oversimplified
representation, the diode is gated and opens to drain off excessive inputs. Typi-
cally, the damage threshold is raised from 100 V to 2000 V, so the item is still
ESD-sensitive and, in many cases, e.g., dies being assembled, input protection is
inapplicable. Therefore, design protection is no panacea, and the need for ESD
control by other means remains.
The first of these other means is grounding of conductors. All metallic or car-
bon-loaded conductors, from large structures such as ovens to the smallest tools,
must be grounded in a static-safe workstation (SSW) to prevent the MM (discharges
~~ ±
~DESIGN
~~N~··
PROTECTION
::~lIl -orA~ DISSIPATIVE

...~~--
AV
vs SURFACE
I •
,-,'
LP ~
SLOW DRAIN
;::;;~
/'AI1 SPECIAL
//11\\",,-
-
V-I \ \ ~'lzED
-\ + AIR
HUMID FlOOR +
AIR
+++
U FINISH
~
"/
NEUTRALIZATION
FLOOR OF CHARGES
LlMITINGTC
SYMBOLS ABBREVIATIONS
~ CURRENT-LIMITING RESISTOR CM .. CONTINUOUS MONITOR
(FOR WRIST STRAP)
t ZENER DIODE
VS • VOLTAGE SUPPRESSION
LP • LEAKAGE PATH
TC .. TRIBOELECTRIC CHARGING
SLOW DISCHARGE E • aECTRICAL (E) FIELD
I!N .. DELTA VOLTAGE
OTHER SYMBOLS: SEEAGURE1
Fig. 3-2. Defenses.
from charged conductors) and FIM (fields from charged conductors). To prevent
the H8M, operators' skin also must be grounded. Skin-grounding preferably
includes a continuous monitor because ESD damage happens in a tiny fraction of
a second and no occasional checks, however frequent, can put failed wrist straps
out of service fast enough (Ref. 3-4). The resistors in Fig. 3-2 limit the current to
ground to a safe maximum of 1.0 rnA in case the operator touches a hot lead-and
safety comes ftrst in any operation. Alternate methods of grounding personnel,

e.g., through weakly conductive shoe soles on conductive floors, are sometimes
used. For example. shoe-grounding may be preferred when operators are extremely
mobile.
Though fields from conductors are eliminated by grounding. this method is
useless against fields from nonconductors such as common plastics because non-
conductors. by deftnition, are ungroundable. If we cannot make the nonconduc-
tor surface-conductive by applying an antis tat, as described below, we must accept
the existence of a field and either (1) shield against it or (2) suppress its voltage.
Examples of shielding (Fig. 3-2) are placing grounded screens over cathode ray
tubes (CRTs) or using metallized laminates for ESD-protective packaging.
In voltage suppression (VS), as diagrammed in Fig. 3-2, the E field is "col-
lapsed" by proximity to a ground plane. This proximity raises C, and therefore
lowers V. in the equation Q = CV. where Q = charge. C = capacitance, and V =
voltage (the measure of fteld strength).
A good example of VS is the operator discipline of wearing long hair tied close
to the skin instead of hanging free. The ground plane of the skin suppresses the E
field from nearby hair to a low, acceptable level (Ref. 3-5).
In the special case of the FFB, E fields on boards must be suppressed by metal-
lic ground planes in the board or neutralized by ionized air as discussed below.
The control of E fields presents an opportunity to save money through a rational.
analytical approach. The fact is that fields cannot be entirely eliminated. but a
harmless level offield strength can be determined and achieved cost-effectively. A
simple method for measuring E fields and declaring them safe is available (Ref. 3-6».
When this method was applied to chairs. the field from ordinary chairs was found
to be harmless at the bench surface where the work was being done. so an invest-
ment in expensive conductive. groundable chairs was avoided.
The next defense in our arsenal is the slow-drain principle to prevent the CDM
(or its variant. the FFB). which may be the most troublesome damage model.
Substitution of static-dissipative for conductive surfaces tames discharges from
device-killing surges to harmlessly slow drains. When the resistance of the leak-
age path around an oxide layer (Fig. 3-2) is lower that the resistance through the
lead to ground, as occurs in MOSFETs. the internal leakage current will outrace
the external drain current so that the oxide is never overstressed by a delta voltage.
Still another defense is the limiting of triboelectric charging (TC) as diagrammed
in Fig. 3-2 by a nonconductive shoe sole separating from a tile floor. The sole will
always become charged to some extent. but a special floor ftnish with a micro-
scopically thin layer of antistat on the surface (Ref. 3-7) will limit charging. Car-
pets also can be antistat-treated. Alternatively. static-dissipative soles can be worn
on conductive flooring or carpet to virtually eliminate TC.
Even if wrist straps with continuous monitors are used, a static-limiting floor
finish on conventional tile is desirable as an inexpensive "safety net." For ex-
ample. a visitor who barges into a static-safe workstation (SSW) and touches a
500 V-sensitive device might not "zap" it if static-limiting floor finish has mod-
erated his triboelectric charge from 1000 V to 300 V.
We have seen that a special finish reduces TC on floors, and the general defense
is to treat nonconductive surfaces of all kinds, e.g., upholstery, walls, or plastic
equipment housings, with a topical antistat that attracts atmospheric moisture to
form a "sweat layer" that drains away incipient charges.
An even more general defense that reaches all surfaces and "nips TC in the bud"
is humidification. However, high relative humidity has drawbacks. It reduces TC
but does not eliminate it, so no operator disciplines should be relaxed. However,
these disciplines may be violated, contrary to the rules, if humidification lulls
personnel into a false sense of security. Furthermore, humidification is expensive
and may cause personal discomfort (because of mugginess), corrosion, delamina-
tion of circuit boards, and solderability problems.
Another general defense is air ionization. As shown in Fig. 3-2, the principle is
to bathe charged surfaces in a stream of both positive and negative air ions. In the
diagram, the negative charge (minus symbols) on the device, which could cause a
CDM discharge, will soon attract positive air ions and be neutralized.
A basic limitation of ionization is that it does not prevent TC and can only
control charges that already exist and emanate an E field to attract the neutralizing
air ions. Thus, the method is not a panacea, but it does playa significant role in
ESD control as will be discussed later in this book.
Last but not least among defenses is operator skill, achieved by training and
evidenced by certification. This brings us to the subject of documenting and orga-
nizing your program.
Documentation/Organization
We've touched on the major defenses with simple illustrations but, in fact, ESD
countermeasures can become quite complicated. This is particularly true for spe-
cial processing such as cleanroom operations or automated assembly. Then, how
should all the various ESD-control procedures by documented? And how does a
manufacturer gear up organizationally?
Fig. 3-3 shows some key elements. First come the military documents if you
have a government contract. Or, even if you don't, the combination of MIL-STD-
1686 and MIL-HDBK-263 makes a good "bible." (Note: But see Chapter 10.)
Next is your own ESD-control program plan, which implements the govern-
ment documents. It carries out their intent by translating general "commandments"
to "situation ethics" fitting your circumstances, e.g., budget available and facili-
ties already in place. These circumstances also include your established business
system, whose policies will necessarily impact the plan.
From the plan flow nuts-and-bolts documents such as drawings and specifica-
tions, which are continually revised and improved by inputs from experimenta-
tion, failure analysis and, in accordance with Total Quality Management (TQM)
REQUIREMENTS SUGGESTIONS
(MIL-STD-1686. (MIl-HDBK-263.
•
DIDa. ETC.) MIl-HDBK-n3. ETC.)
ESD-CONTROl
PROGRAM PLAN
• BUSINESS
SYSTEM
APPROVED
!
MATERiAlS AND DRAWINGS EXPERIMENTATION
EQUIPMENT r-- SPECIFICATIONS
(MIL-STD-1000. ETC.) (ENGINEERING)
7
(LIVING LIST)
FAILURE
ANAlYSIS
FAClllTIZATION
(BENCHES.
OPERATOR
WORK
~ OPERATOR
EXPERIENCE
GROUNDING. ETC.) INSTRUCTIONS (TOM)
..
-
L
I
HANDLING. TRAINING
ASSEMBLY. AND
STORAGE CERTIFICATION
~ J
FINAl
TESTING OA
l
\
'"
MARKING.
PACKAGING. ~
SHIPPING
! COMPANY
HIGH YIELD,
HIGH RELIABILITY.
"'---- STATIC-
AWARENESS
PROGRAM
NO LATENT FAILURE
Fig. 3-3. Documentation/Organization.

precepts, operator experience. A "living list" of approved ESO-control materials

and equipment is an important appendage.
A specification gives basic requirements but should not-and, as a practical
matter, cannot-provide the fully detailed "how-to" that workers need. Therefore,
operator work instructions evolve from, and expand on, the specification.
Now training and certification of personnel enter the flow, bringing us to the
actual activities: facilitization, handling, assembly, storing, testing, marking, pack-
aging, and shipping. All these are watched over by Quality Assurance (QA), while
a company program with bulletins and posters keeps everyone static-aware.
Last but not least, top management must be sold on the need for ESO control
and give the program full support.
The goal, of course, is high yield (no overt ESO damage) and high reliability (no
hidden ESO damage causing latent failure in the field). By understanding the
threats, developing defenses, and winning the support of management, you'll have
a viable and effective ESO-control program.
ESO-control program management, including documentation and organization,
is a whole art or science in itself. Reference 3-8 is an authoritative book on the
subject.
Conclusion
This thumbnail sketch has given textbook examples, but actual ESO-control situ-
ations are often unique, with no ready-made answer. You must arm yourself with
"real and conceptual tools" and run your own tests to get the facts for a cost-
effective solution (Ref. 3-9). Without facts, the only alternative is to try to buy
your way out of the problem with ignorant overkill.
At one extreme of ESO expertise is a naive customer sitting at a costly ESO-safe
assembly bench heaped with superfluous equipment touted by glib salesmen, and
at the other extreme is an ESO "guru" working at a wooden table without even a
wrist strap because he knows exactly what he's doing. Your ESO-safe workers
should be somewhere in the middle: knowledgeable but backed up by "safety nets"
to allow for human error. In "Defenses," above, static-limiting floor finish was
mentioned as an example of such a safety net.
To put ESO control succinctly: "Work smart," and "buy the basics, forget the
frills."
Most important of all, make every operator a "little guru" and "captain of his/her
own workstation," because operator skill is the key element in a successful pro-
gram. Remember that an inept sea captain can wreck the most modem ship equipped
with every sophisticated radar and sonar safety device.
Chapter 4
Real and Conceptual Tools from A to Z
First, about fifty principles, rules, definitions, test methods, etc., are surveyed "from
A to Z" under the heading "Real and Conceptual Tools: Definitions." This is not a
glossary of ESD terms but a limited list of those that are especially important to
our approach; definitions of other terms, such as abbreviations for types of de-
vices, can be found in textbooks on electronics.
Second, the terms are elucidated under the heading "Real and Conceptual Tools:
Discussion," and three of the most important ones, the COM, SSP, and SSW, are
given chapters of their own, Chapters 5, 6, and 7. These tools serve to troubleshoot
in novel situations (Chapter 8), to develop specifications (Chapter 9), and to build
a logical and thorough ESD-control program (Chapter 12). As mentioned in the
Introduction, these key terms appear in bold type throughout the text.
ESD control has its complexities but is basically a simple matter of avoiding
discharges and dangerously intense E fields--not all E fields, which would be
impractical. (Note: In our usages, "E field" and "field" are synonymous.) If all
objects in the SSZ are groundable and grounded, and if the CD Rule is followed,
most of the ESD problems will have been solved. But the reader shouldn't be
lulled into false confidence by this statement, because achieving these basic re-
quirements of grounding and following the CD Rule is more easily said than
done, human nature being what it is, and a residue of problems, exemplified by
those in Chapter 8, will remain. These may be solved by diligent use of our con-
ceptual tool kit.
Table 4-1 categorizes the tools as primarily applicable to either fields or dis-
charges except when no distinction can be made. There is often much overlap;
humidification, for example, reduces not only fields but also discharges (which
may be visible sparks when the air is dry enough), but it is most often thought of as
lessening the static charge on nonconductors, especially common plastics, to
reduce the danger of the FIM. The point of the table is that every concept, mate-
rial, or object in this chapter plays a role in preventing OJ or the "internal OJ"
caused by induction ("pushing" or "pulling" of electrons by a field to create
polarization within a device).
22
REAL AND CONCEPTUAL TOOLS FROM A TO Z 23
Table 4-1. Categorization of Tools.

TOOL
REAL (AN OBJECT

APPLICABILITY OR MATERIAL) CONCEPTUAL
Primarily pertinent to fields Antistatic (materials) Apparent charge

Faraday cup CD Rule
Field meter E field
Insulator FFB
Nonconductive (materials) Field
Nonconductor F1M
Static-dissipative (materials) Humidification
Topical antistat Ionization
Shielding
Static charge
Static field
Surface resistivity
Triboelectric charging
Voltage suppression
Primarily pertinent to dis- Antistatic lotion ARTG

charges Conductive (materials) Capacitance
Conductor CDM
Continuous wrist-strap monitor DI
Personnel Voltage Tester Discharge
Static-limiting fioor finish Discharge test
Zapfiash Grounding
HBM
Shunting
Walk test
Equally pertinent to both Coupon Analysis

fields and discharges Faraday cage Basic Rule
MOSFET ESDS
SSP ESDSitem
SSW Latent failure
SSZ Operator disciplines
Special test
Standard test
Real and Conceptual Tools: Definitions

Analysis: A paper study with no testing required.
Antistatic: Having a surface resistivity of at least 109 but less than 1012
ohms/square (our definition; same upper limit as EIA 541 or MIL-HDBK-263
"dissipative"). The old DoD-HDBK-263 definition was 109_10 14 ohms/square.
Antistatic Lotion: Hand lotion applied to the wrist to lower skin resistance
and help meet ARTG for operators wearing wrist straps.
Apparent Charge: The charge in volts of a surface or object as read by a

hand-held field meter. The word "charge" is used for convenience; the correct
units of charge are coulombs (or usually nanocoulombs for ESD work).
ARTG: Allowable resistance to ground for operators' skin. For ESDS items
sensitive to 100 V this is 10 megohms.
Basic Rule: An ESDS item shall never, even for a nanosecond, be exposed to
an E field of more than a prescribed strength or have its sensitive leads or terminals
touched to, or receive a discharge from, any surface at more than a prescribed voltage.
Capacitance: The capacity or ability of an object to hold a charge. In ESD
control, the usual unit of capacitance is the picofarad, pF.
CDM: Charged Device Model, a damage mechanism.
CD Rule: Charge-distance rule. An empirical relation of safe distance from a
surface versus the apparent charge on that surface. This rule protects items sensi-
tive to 100 V from damage by E fields by the FIM.
Conductive: Having a surface resistivity of less than lOS ohms/square.
Conductor: A conductive material. Common examples in ESD control are
metals and plastics loaded with graphitic ("metallic") carbon.
Continuous Wrist-Strap Monitor: Equipment for continuously checking the
resistance of an operator to ground to make certain that the ARTG is never ex-
ceeded even for a nanosecond as required by the Basic Rule. A continuous wrist-
strap monitor system includes the wrist strap with dual-conductor cord.
Coupon: An exaggeratedly ESD-sensitive model for an ESDS item. Cou-
pons are passed through an automated process to locate hazards and, when these
have been corrected, to give assurance that the process remains ESD-safe.
Of: Direct Injection, a damage mechanism. Note: This and the COM, FFB,
and FIM are called "damage mechanisms" as opposed to "failure mechanisms,"
which are the ways an ESD-affected device fails internally, e.g., dielectric break-
down (MIL-HDBK-263).
Discharge: The flow of electrons between surfaces. This flow may take place
between touching surfaces or across a gap via ionized air as a visible spark or an
unseen silent discharge.
Discharge Test: A test in which a finger of a charged person, a tool held by a
charged person, or a probe connected to a charged capacitor is discharged to the
surface of an ESD-protective container, e.g., a bag or tote box, holding a MOSFET
in a special fixture. Damage to the MOSFET constitutes failure of the test. A
detector simulating a MOSFET may be used.
E field: Electrical field, which is the region surrounding an electrically charged
object in which another electrical charge can be induced and will exert a force.
ESDS: ESD (electrostatic discharge)-sensitive.
ESDS Item: ESD-sensitive device, component, subassembly, or assembly.
Used as a general term for convenience.
Faraday Cage: A container whose wall is electrically continuous and has at
least the electrical conductivity of O.OOO25-inch (O.25-mil) aluminum foil. Small
holes, such as made by staples, are allowed. Wire screen, e.g., FED-SPEC-RR-W-365 ,
Type VII, 18 X 16 regular, is an excellent Faraday cage, equivalent to at least
6-mil foil. (Note: This is our own definition.)
Faraday Cup: A shielded cup with attached electrometer for measuring the
charge on any object dropped into the cup.
FFB: Field-From-Board damage mechanism.
Field: An E field in the context of this book.
Field Meter: An instrument for measuring E fields.
FIM: Field-Induced Model, a damage mechanism.
Grounding: Metallic connection with the earth to establish zero potential with
respect to earth. Grounds for ESD purposes, e.g., for an SSW, include water pipes
and building structural steel. Grounding also means the electrical connection of
conductive, static-dissipative, or antistatic materials with earth in any manner.
HBM: Human Body Model, a damage mechanism (special case of DI).
Humidification: Controlling relative humidity, e.g., to 30% or 40% minimum
at 70°F, to reduce the triboelectric charging of materials in the SSW.
Insulator: A nonconductor.
Ionization: The ionization of air for the purpose of neutralizing charged surfaces.
Latent Failure: Failure of an ESDS device because of previous degradation
and weakening by one or more ESD events.
MM: Machine Model, a damage mechanism.
MOSFET: Metal-oxide semiconductor field-effect transistor. Used in the dis-
charge test and in special tests.
Nonconductive: Having a surface resistivity of at least 1012 ohms/square (our
definition, consistent with that of MIL-HDBK-263 and the Electronic Industries
Association).
Nonconductor: A nonconductive material.
Operator Discipline: A rule of conduct for operators that tends to minimize
the chance of ESD damage to ESDS items being handled.
Personnel Voltage Tester: Trade name for a voltmeter with high input im-
pedance and low output capacitance, for measuring the instantaneous voltage on
people.
Shielding: Attenuation of an E field by an object between the charged surface
and a point in space at which the field strength is measured. The more conductive
the object blocking the field, the better the shielding.
Shunting: "Shorting out" or connecting leads together as a means of lowering
the ESD sensitivity of an ESDS item.
Special Test: A test designed to resolve a specific ESD-problem situation.
Typically, the test conditions are worst-case, yet realistic, so that confidence is
high if the test is passed.
SSP: Static-safe package. This is an SSZ within a Faraday cage or within
material providing sufficient shielding to pass a discharge test on the particular
configuration of shipped/stored ESDS item and packaging.
SSW: Static-safe workstation. This is a volume in space which is designated

and equipped to maintain an SSZ.
SSZ: Static-safe zone. A volume in space at every point of which the Basic
Rule is followed.
Standard Test: A test procedure from a specification or other ESD-control
document.
Static Charge: The quantity of electricity, measured in coulombs, nano-
coulombs, etc., accumulated on a nonconductor. This is a classic definition, but
static charges occur, of course, on electrically isolated (ungrounded) conductors
and exist for a brief time, but drain off quickly, on grounded antistatic or static-
dissipative materials.
Static-Dissipative: Having a surface resistivity of at least lOS but less than
109 ohms/square. This is the old DoD-HDBK-263 classification.
Static Field: An E field caused by a static charge.
Static-Limiting Floor Finish: Floor "wax" with a static-dissipative or anti-
static surface as well as a propensity to reduce triboelectric charging of people
walking or shuffling their feet while sitting.
Surface Resistivity: The electrical resistance between two electrodes pressed
against a surface and forming the opposite sides of a square of any size. The units
are ohms/square, and the basic reference is ASTM D257.
Topical Antistat: A moisture-attracting substance applied to the surface of
nonconductors (insulators) to make them antistatic.
Triboelectric Charging: Generation of static charges by friction or separa-
tion of dissimilar materials or even two pieces of the same material.
Voltage Suppression: The "collapse" or reduction in strength of an E field
around a charged surface when that surface is close to another surface, especially
a grounded conductive one.
Walk Test: Realistic test for approximate voltage attained on walking persons
by triboelectric charging between shoe soles and floor surface.
Zapflash: Trade name for a battery-powered continuity checker with a bulb
which glows dimly at about 7 megohms and becomes brighter with decreasing
resistance.
Real and Conceptual Tools: Discussion

Analysis: Conclusions are arrived at logically, using methods such as extrapo-
lation and criteria such as similarity. For example, we know that 0.25 mil of alumi-
num foil is a sufficient Faraday cage for general use, so extrapolation tells us that
a I/s-inch wall of aluminum on an electronic box is even better and need not be
tested for shielding. Or, if one antistatic plastic composition with a surface resis-
tivity of 10 10 ohms/square bleeds off static charges satisfactorily, we can con-
clude that a different composition with similar surface resistivity may be substituted
(assuming no shortcomings in other properties).
Antistatic: DoD-HDBK-263 defined "antistatic" as having a surface resis-

tivity of less than 1014 ohms/square. However, our measurements with blade elec-
trodes and a megohmmeter (Paper No. 2 in the Appendix), or with a Voyager
SRM-110 surface resistivity meter, showed that values of 10 12 ohms/square or
more are too high for static charges to bleed off within a few seconds as desired.
Exactly what bleed-off time is needed is a moot question, but we insist that anti-
static surfaces not retain over 300 V apparent charge when stroked with polyes-
ter fabric or Aclar (polychlorotrifluoroethylene) film and held in one hand while
being measured with a field meter in the other and waiting 5 seconds. Numbers
aside, the antistatic surface must be sufficiently conductive for a feeble current to
flow across it (Ref. 4-1).
(Note: The word "conductive" in the last sentence is not in bold type because it
is not used in our special sense of having a surface resistivity of less than lOS
ohms/square. )
Antistatic Lotion: Some operators have skin contact resistance above 10 meg-
ohms, which exceeds the ARTG (see Paper No.1 0 in the Appendix), so this lotion
is a necessity for them, and it may have to be applied twice a day. Thus the con-
tinuous wrist-strap monitor may become a lotion monitor, signaling when the
lotion must be renewed. Fortunately, most operators enjoy using the lotion, which
has a pleasant feel and can be obtained with either herbal or floral essence. Select
a brand free of lanolin or silicones that might contaminate surfaces and interfere
with adhesion for bonding or printing.
Apparent Charge: This is the uncorrected reading of a field meter, in volts,
taken for a surface following the meter manufacturer's instructions. This reading
reflects the intensity of the E field at the point in space where the sensor plate of
the meter is held. The "voltage" of a nonconductive surface is of doubtful mean-
ing because the static charge may vary greatly form spot to spot: excess electrons,
or regions of electron deficiency, may exist in clumps. Besides, the surface charge
per se is irrelevant. What matters is the strength of the field, but this is hard to
measure even at one point because the meter itself perturbs the field and affects
the reading. Also, the lines of force of the field are warped by nearby surfaces.
However, despite these uncertainties the apparent charge is meaningful for our
CD Rule because this rule is a worst-case approximation with latitude for numeri-
cal error. See "Field Meter."
ARTG: An ARTG of to megohms is recommended in Paper No. to in the
Appendix. Note that, without the worst-case situation of a garment sleeve stroking a
workbench surface, resulting in a very rapid voltage rise, the ARTG might be set
much higher than to megohms. Operators' sleeves are supposed to be of antistatic
smock material or else rolled up (see Example 3, "Operator Clothing," in Chapter 8),
so this worst-case condition should not exist. However, we want backup precau-
tions; if the law on rolling up sleeves is violated, an ARTG of 10 megohms still
provides protection. Our program is designed to provide safety nets in this way when-
ever they are cost-effective. Another good example is static-limiting floor finish.
Basic Rule: For ESDS items sensitive to 100 V by the OBM, this rule may be
rewritten: "An ESDS item shall never, even for a nanosecond, be exposed to an E
field in violation of the CD Rule or have its sensitive leads or terminals touched
to, or receive a discharge from, any surface at more than 50 V."
"Even for a nanosecond" refers to the rapidity at which ESD damage can occur.
Only 25 V applied for 100 nanoseconds, at the chip, can destroy memories or
microprocessors (Ref. 4-2). In the waveform required by MIL-STD-883, Test
Method 3015, referenced in MIL-STD-1686, the rise time for the OBM is a maxi-
mum of 10 nanoseconds (Ref. 4-3). A typical ESD event involves rapid discharge
with rise times of l00{}-3000 volts/nanosecond (Ref. 4-4).
Although a charge of 300 V on nonconductors is harmless for devices sensitive
to 100 V by the OBM (Figure 3 of Paper No.3 in the Appendix), it is usually not
practical to distinguish nonconductors from other materials, and nonconductors
shouldn't be in an SSW. Therefore, 50 V is stated as the limit for any surface.
A basic rule suggested by a supplier of conductive ESD-control materials is:
ESDS items "may be handled or approached only with highly conductive grounded
objects," and "storage or transportation can only be done in Faraday cages (maxi-
mum of 1()4 ohm/square material)." This is the "conductive approach." Note that
our definition of Faraday cage is more rigorous; see the discussion.
In contrast to the conductive approach, a supplier of antistatic materials has
long insisted that all surfaces in the SSW must be "nonsparking." This is the "an-
tistatic approach," and its merit versus the conductive approach is a notable con-
troversy in the very controversial field of ESD control. Which of these two
approaches is right? The conductive approach is correct in theory but lacks the
"forgivingness" or latitude for error needed in imperfect, real-life situations. Spe-
cifically, the CDM and DJ are not taken into account. ESDS items, materials, or
even careless operators might be charged, and conductive surfaces can increase
ESD damage by rapid discharges as discussed under "CDM" below. Also, con-
ductive items, e.g., bags, can carry lethal currents if a 1lO-volt hot lead falls onto
them (Paper No.7 in the Appendix); this danger is improbable. but safety is so
important that even unlikely possibilities must be considered.
There are also other arguments against conductive materials, such as sloughing
of metal flakes from metallized bags or particles from carbon-loaded plastic bags
or tote boxes; these conductive flakes or particles could cause shorts by falling
into open devices or onto printed circuit boards to bridge circuit lines. On the other
hand, antistatic materials of the conventional type (depending on a fugitive anti-
stat) have a permanence problem (Paper No.2 in the Appendix), and their rubbed-
off or volatilized antistats can cause contamination (Paper No.9 in the Appendix).
Nothing is perfect.
On balance, we use and recommend a mix of antistatic, static-dissipative, and
conductive materials, as stated just before the "Conclusions" section of Paper No.
3 in the Appendix. We prefer to avoid conductive materials (note that "conduc-
tive" is defined by the conductivity of the surface, so an externally antistatic foil-
laminate bag, for example, is not conductive), but we use them when no alternatives
are conveniently available. After all, conductive items such a screwdrivers and
other hand tools are ubiquitous in SSWs. We merely try to minimize the addition
of more conductors.
In conclusion, the Basic Rule guards against fields, by means of the CO Rule
and control of triboelectric charging, and also against discharges, by means of
grounding of materials and operators, by operator disciplines, and by using cou-
pons to make automated processes ESD-safe.
Capacitance: Capacitance C is related to the charge Q and the voltage V of an
object by the equation: Q = CV. This means that for a given Q, or quantity of
electrons present or missing, a decrease in C, which is the capacity to hold the
charge, results in a rise in V, which is the "looseness" of electrons (or "anxious-
ness" of their lack) that manifests itself as an E field. Thus, when the capacity to
hold the charge decreases, the charge is less tightly gripped or restrained and radi-
ates as a field; when the capacity increases, this field collapses.
In practice, a charged object is one plate of a condenser (capacitor) separated by
air as the dielectric layer from a ground plane such as a workbench surface or a
floor, which is the other plate of the condenser. When the distance between the
plates doubles, Q remains the same, C is halved, and V doubles. This relationship
is discussed in connection with the ARTG in Paper No. 10 in the Appendix. The
most familiar, and important, example of V rising as C falls is a charged person
lifting his foot.
COM: For a discussion of CDM, see Chapter 5. The nature of the COM has
been questioned, with the suggestion that damage occurs in the charging process
rather than during a discharge, so that the COM resembles the FIM. However,
our observations on the FFB, which is a kind of COM, indicate that a discharge
isn't necessary for ESD damage but increases its likelihood (Paper No.8 in the
Appendix).
Insofar as 01 is a part of the COM, a conductive surface tends to be more
damaging than an antistatic surface (Chapter 5). This was demonstrated in tests
which support the intuition that an antistatic surface slows the rate of a discharge
and is safer than a conductive surface. Hence we discourage the use of conduc-
tive surfaces but don't forbid it; see "Basic Rule" above.
CO Rule: This empirical rule is based on MOSFET tests as described in Pa-
per No.3 in the Appendix. Since the MOSFET board used in the tests is an exag-
geratedly sensitive design (large "antennas"), the CO Rule is worst-case, and this
worst-case quality relieves uncertainty in field-meter readings as explained under
"Apparent Charge" above. The equation d = V°.5/1.8, where d is the minimum
safe distance in inches and V is the apparent charge in volts, can be expressed as
a table as included in Model Specification 1 in Chapter 9.
The CO Rule is incorporated in a company in-plant-handling specification, as it
is in Model Specification 1, and several years of experience have proved this rule
to be extremely useful. It has helped solve many problems, a few of which are
described in Chapter 8. The rule is especially valuable in preventing needless ex-
pense; fears of low-intensity fields are discredited so that purchase of equipment
to control these unreal hazards is avoided. For example, conductive chairs were
shown to be unnecessary in Example 5 in Chapter 8. In other words, the rule fore-
stalls "overkill" and maximizes cost-effectiveness.
Based on the monetary savings provided by the CD Rule, field meters should
be generously distributed to operators. One field meter for each SSW is ideal. Not
only do the meters allow cost-effective solutions to ESD problems via the CD
Rule, but they are a prime tool of eternal vigilance to exclude common plastics
and other sources of fields from the SSW.
In conclusion, the CD Rule will be one of your best friends in ESD control. Use
it often.
Conductive, Conductor: Conductivity is defined by surface resistivity be-
cause surfaces tend to be the locus of currents in ESD-control work; examples are
vapor-deposited metal layers or antistat sweat layers. In the test (ASTM D257 or
equivalent), current flows through the bulk of a conductive material such as car-
bon-loaded polyolefin, not just on the surface, but what matters from a practical
standpoint is not where the current flows but how large it is, because the property
of real interest is the rate of bleed-off of static charges.
Incidentally, this facility to carry current may be too large for safety of person-
nel; see "Basic Rule" above.
For this reason of safety to people and because of slightly more danger to ESDS
devices (see "CDM" above), the use of conductive surfaces is discouraged but
not prohibited in our approach.
Do not use a field meter to measure the true voltage on a conductor. The read-
ing probably will be deceptively low, so that DI may be unrecognized as a threat.
See the precautions under "Field Meter."
Continuous Wrist-Strap Monitor: This subject is discussed in detail in Pa-
pers No.5 and lOin the Appendix. The Basic Rule requires continuous monitor-
ing "even for a nanosecond," and if a wrist-strap system is found to have failed in
a daily check, for example, the operator's skin voltage may have been sufficient to
cause damage by DI for not just a nanosecond but for a whole shift!
More and more continuous wrist-strap monitors, of the resistive type as we
require, are being used in assembly areas. There is no logical alternative, espe-
cially for high-reliability products in which latent failures, however few, could be
disastrous. However, try monitors before buying to be sure the audible alarm is not
annoying, etc.
Coupon: Paper No.8 in the Appendix describes the coupon approach and its
successful application to an automated assembly process. The complete version of
this paper also mentions damage mechanisms and ESD hazards.
01: Electrons flow directly in or out of a device. This is the most straightfor-
ward and common cause of ESD damage, with people being the major threat.
Therefore, controlling voltage on personnel (Paper No. lOin the Appendix) is one
of the most important parts of an ESD-control program.
The HBM is a model for 01 caused by people, but the conditions are not realis-
tically worst-case; see "HBM" below.
Discharge: Discharges and fields are the two ESO threats controlled by the
Basic Rule. A discharge may be very rapid, e.g., with a rise time of 1000-3000
volts/nanosecond, or slow, e.g., 1000 volts/minute, in which case it is called "bleed-
off' or "drain." Conductive surfaces favor rapid discharge, and there is evidence
that these surfaces are therefore more hazardous to ESDS items than are anti-
static surfaces (see "COM" above).
A discharge is a flow of electrons, with the quantity of electrons being mea-
sured in coulombs and the rate of flow (current) in amperes = coulombs/second. A
discharge creates a field which may itself cause ESO damage.
Discharge Test: In variations of this test, we use a resistance of 1500 ohms as
in the HBM (Chapter 9, Model Specification 2, para. 4.3.3, option I), 150 ohms
(same reference, option 2), or 0 ohms (Test No.1 in Paper No.4 in the Appendix).
For tote boxes, electrodes are used instead of a capacitance probe (see under "Test
Methods" in Paper No. 7 in the Appendix). The test becomes much more rigorous
as the resistance drops.
The shielding/discharge test with a capacitive sensor and oscilloscope (see Fig.
2 of Paper No.3 in the Appendix) has become a standard for bag materials (in EIA
541, "EIA" being the Electronic Industries Association), but this test is misleading
as explained in Paper No.3 in the Appendix. The flaws in this test are (1) it gives
falsely high (good) readings when metallization is exposed rather than buried, and
(2) the flat electrodes don't provide a spark discharge. A "real-life" spark from a
person's finger evaporates the metal from some see-through metallized bags so
that they afford poor protection; see Table 3 of Paper No.4 in the Appendix. An
analogy of an unrealistic test that is still on the books but has been largely sup-
planted by more realistic tests is ASTM 0635 for flame resistance of plastics.
Nylon passes this test, but glass-reinforced nylon, which if anything should be less
flammable, fails. Why? Because the low melt viscosity of nylon allows the molten
region to drip off, carrying the flame with it, unless glass fibers support the melt so
that the flame can propagate! The shielding/discharge or pulse test of EIA 541 is
just as unrealistic. Another test corrected ASTM 0635 by adding cotton below the
burning specimen with a provision that burning, falling drops must not ignite the
cotton for the test to be passed. Similarly, a correction to the pulse test would be to
add a spark discharge as we have done in our modified version. See Chapter 7 for
further discussion of unrealistic versus realistic tests.
It is interesting that researchers at a major corporation producing see-through
metallized bag material noted evaporation of the metallization by a spark and even
studied the area evaporated versus voltage of the discharge, but this information
was suppressed because it might have harmed sales. Our position is that spark
discharges are a significant threat and bags should resist them. See Paper No. 20
in the Appendix.
E Field: Fields are, of course, the threat, along with discharges, that the Ba-
sic Rule controls. For practical purposes (the CD Rule), field strength is mea-
sured in terms of apparent charge and the distance from the charged surface. See
the discussion under "Apparent Charge" above.
ESDS, ESDS Item: ESDS devices fail in two common ways. Voltage-induced
failure dominates for discrete MOS (metal-oxide semiconductor) devices. A volt-
age of 80-100 V exceeds the dielectric-breakdown strength of the common thick-
ness of 1000 angstroms of oxide. After breakdown, the part usually shows
permanently degraded leakage characteristics and is more susceptible to future
damage; it has a reduced operating margin and is likely to fail.
The other common failure mode is current-induced failure. This appears in
Schottky or PN (bipolar) junction ICs, which are subject to destructive reverse
breakdown in which the instantaneous power of the discharge (more than 5 kW
for less than I J.lsec) generates enough heat to melt silicon and metal. Thin-film
resistors are damaged by surface breakdown that bypasses part of the resistor and
leaves a shunt path, or else metal is vaporized or melted from the thin film. In
either case, the resistance value usually shifts, which is serious in a precision (less
than 0.1 % variation) part (Ref. 4-5).
Typical failure distributions for bipolar ICs are 90% junction burnout and 10%
metallization burnout. For MOS integrated circuits, this changes to 63% metalli-
zation burnout and 27% oxide punchthrough (Ref. 4-2). For a full discussion of
failure mechanisms, see MIL-HDBK-263.
MIL-STD-1686, which superseded DoD-STD-1686, defines three classes of
ESD-sensitivity: Class 1 (0-1,999 V), Class 2 (2,000-3,999 V), and Class 3
(4,000-15,999 V). Class 1, which is the concern of this book, includes microwave
devices (such as Schottky barrier diodes), M OSFETs, surface acoustic wave (SAW)
devices,junction field-effect transistors (JFETs), charged coupled devices (CCDs),
precision voltage regulator diodes, operational amplifiers (OPAMPs), thin-film
resistors, integrated circuits (ICs), hybrids using Class 1 parts, very high speed ICs
(VHSIC), and some silicon-controlled rectifiers (SCRs).
Note that components mounted on an assembly are relatively ESD-safe only if
the assembly has protection circuits at all sensitive nodes. The risk of ESD damage
can actually increase for mounted devices because each printed conductor con-
nects to several devices and a discharge to that conductor stresses several devices,
not just one. (Note: The word "conductor" in the last sentence is not in bold type
because it is used in the special sense of "circuit line.") Another factor is that
circuit lines can act as antennas to intensify the effect of fields and promote dam-
age by the FIM. Furthermore, CMOS (complementary metal-oxide semiconduc-
tor) circuits subjected to ESD while powered are subject to an additional risk called
"latchup," which is an avalanche effect causing overheating and catastrophic fail-
ure (Ref. 4-5). We repeat: ESDS devices mounted on boards can be at even more
risk than when unmounted (Ref. 4-6).
Faraday Cage: As defined in technical dictionaries, a "Faraday cage" is a
room enclosed with an earth-grounded metal screen, and the only commercial ESD-
protective bag in accordance with this definition (except for grounding) is con-
structed of aluminum wire screen (as used for windows) sandwiched between layers
of antistatic polyethylene (see Paper No.4 in the Appendix). However, for most
purposes, 0.25 mil of aluminum foil, as in commercial foil laminate bags (see
Table 10 of Paper No.3 and Table 3 of Paper No.4 in the Appendix) is sufficient.
We recommend these foil laminate bags for all in-plant handling as well as ship-
ping (but nonfoil package constructions may meet the discharge test of Model
Specification 6). However, consider the screen bag when visibility of the contents
is needed and extremely high ESD sensitivity, high cost, or need for high reliabil-
ity of the enclosed item justifies a relatively expensive package.
Note that being a Faraday cage or equivalent is not enough for an SSP, because
the inside of an SSP must be an SSZ in which triboelectric charging cannot violate
the Basic Rule by creating hazardous fields. The old MIL-B-81705, Revision B,
Type I, was a perfectly good Faraday cage, but its nonconductive liner was unsuit-
able for an SSZ; in a test, miniature circuit boards with MOSFETs were shaken in a
bag and found to be damaged due to triboelectric charging (Ref. 4-7). Our own
tests verified that this liner was nonconductive. The new MIL-B-81705, Revision
C, Type I is satisfactory. It contains a layer of opaque metallization, not foil.
In conclusion, use foil laminate bags or opaque, metallized bags (MIL-B-81705C,
Type I) with antistatic surfaces both inside and outside, or, in special cases, use
screen bags with antistatic plastic sandwiching the screen (see Paper No.4 in the
Appendix).
Faraday Cup: For examples of the use of this test equipment in studying the
propensity for triboelectric charging, see Papers No.6 and 7 in the Appendix. A
field meter measures charges indirectly and imperfectly, whereas the Faraday
cup gives accurate reading on any charged object dropped into it. However, when
the intensity of the E field is the property of real interest, as it is using the CD
Rule, the relevant parameter is the apparent charge read by a field meter and
expressed in volts.
FFB: This damage mechanism is a special case of the FIM. Triboelectric
charging of circuit boards, e.g., by sliding on stainless-steel conveyor belts or
oven shelves, creates a field which damages ESDS devices by induction as ex-
plained under "FIM" below.
Field: See "E Field."
Field Meter: This has been and still is the single most useful tool in ESD
control.
Field meters commonly are calibrated at the factory with a charged I-ft2 plate,
but in actual measurements the target areas vary, with a large effect on the reading
(Ref. 4-8). Thus, the actual voltage, or charge, on the surface can't be accurately
determined, but what matters for our purposes is the apparent charge as dis-
cussed under "Apparent Charge" above.
Though the real charge is irrelevant, we do want a properly determined appar-
ent charge as a measure of field strength, and there are a few requirements for
using the meter correctly.
First, the meter must be calibrated. This is suitably done annually, with a cali-
bration sticker being placed on the meter.
Second, if the meter has a grounding outlet, it must be grounded with a lead or,
if the meter has no such outlet, the operator holding the meter must be grounded.
Third, the meter must be zeroed on a grounded conductive object, suitably the
grounded operator's hand. Note that meter readings are relative. If neither op-
erator nor meter is grounded and the operator is charged to -1000 V and has
zeroed the meter on his hand, he will read a grounded I-ft2 surface as +1000 V,
or he will read a surface at + 1000 V as +2000 V. Thus, to purposely read the
approximate charge on himself, the ungrounded operator can take a reading on a
grounded I-ftl surface and reverse the polarity. Or a larger surface can be used
for a rough result.
Fourth, the apparent charge must be read according to the meter manufacturer's
instructions; for example, the meter might read true at 6 inches from a surface, or
the distance might be increased to 12 inches, to keep the needle on scale, and the
reading doubled.
We recommend that every SSW have a field meter to be used in complying
with the invaluable CO Rule. However, the meter must be used only for apparent
charges. It is not to be used for reading voltages on conductors because the true
voltage is not shown by the meter unless the conductor happens to be a I-ft2 plate
(assuming the meter was calibrated with this). For example, because of its small
area, a wire at 500 V might read only 50 V on the meter. According to the CO
Rule, this reading of an apparent charge of 50 V prescribes a safe distance of 4
inches, which is fine, but a requirement of the Basic Rule for ESOS items sensi-
tive to 100 V is that a sensitive lead may not touch a surface at more than 50 V in
real voltage. In this example, if the bare wire were touched, the item would be
damaged by OJ. In other words, field meters, as their name indicates, are for
judging fields, via apparent charge and distance, not for measuring true volt-
ages! To measure voltages on conductors, use a contact or noncontact voltmeter.
A noncontact voltmeter is a type of field meter, but it is specially calibrated and
applicable only to certain sample geometries; it is not an ordinary, general-purpose
field meter.
Operators should be given detailed instructions on using the field meter
because it is such an important tool. For familiarization, they should charge a
sheet of non conductive plastic, e.g., polyethylene, by rubbing with polyester
or woolen fabric, and note the unevenness of the voltage readings over the
surface because electrons or electron-deficient areas are clustered irregularly;
in such cases, the highest reading, which is the most conservative for protec-
tion purposes, shall be used as the apparent charge in implementing the CO
Rule. Also, operators should rub the plastic against grounded stainless steel
and note that grounded conductors, though they never show a charge them-
selves because it immediately drains off, are very effective in charging non-
conductors including circuit-board laminates. Intuition may suggest that
grounded metals are harmless in regard to ESD, but this idea is dangerously
wrong! Not only can they cause triboelectric charging, but they can be tar-
gets for rapid discharges in the COM: see "COM" above. The skin is another
conductor that charges nonconductors by rubbing, as the operator can
demonstrate by wiping a hand over the surface of an uncharged plastic sheet.

(Use an ionizing blower to neutralize the plastic between triboelectric charg-
ing tests, or else the charge may be "blotted" off by pressing the palm of the
hand, without rubbing, against the plastic surface.)
If the sensor of the meter is accidentally touched to a highly charged conductive
surface, the meter may lock up and refuse to be zeroed; if this happens, give the
meter time to recuperate. Sometimes zeroing is difficult because fingers touching
the sensor have left body oils on it. When this happens, clean the sensor with a
swab using a solvent such as isopropyl alcohol and be careful not to touch the
sensor again.
FIM: Even relatively insensitive devices are at risk if the "antennas" are large
enough and the field is strong enough. The possibility of FIM damage is why
fields are a hazard to be avoided in accordance with the Basic Rule.
Grounding: In theory, if all materials, including cases on parts, in the SSW
were conductive and grounded, as in the idealistic "conductive approach" (see
"Basic Rule" above), all surfaces would be at the same (zero) potential, and
fields and discharges would be nonexistent so no ESD damage could occur. In
practice, however, antistatic and static-dissipative materials are much used
because these materials are safer for people and ESDS items than are conduc-
tors, and these materials are grounded in various ways, e.g., by being held by a
grounded operator or by lying on a grounded antistatic or static-dissipative
workbench surface.
The grounding lug of the workbench itself must be connected through a resistor
to an earth ground, e.g., a water pipe, as required in Model Specification 1 in
Chapter 9. As stated in the same specification, some equipment, e.g., ovens, may
be connected directly (hard-grounded) to commercial power ground.
HBM: In the HBM (MIL-STD-883, Method 3015), the touch of a charged
person's finger is simulated by discharging a lOO-pF capacitor through a 1500-
ohm resistor. But human capacitance reaches 250 pF, and body resistance can be
as low as 100 ohms (Ref. 4-9). Or Ref. 4-2 gives these extremes as up to 500 pF
and down to 50 ohms. Furthermore, a charged metal tool might touch an ESDS
lead or terminal (the MM, Machine Model) so that the resistance would be less
than 1 ohm. When a person discharges through a tool, the rise time may be much
faster than a finger because of the sharpness of the discharge tip (Ref. 4-2). An-
other problem with the HBM is that data vary from one simulator model to an-
other, so that the need for a calibration procedure in MIL-STD-883C has been
suggested (Ref. 4-10).
Ratings of sensitivity by this test are commonly called "V-zap levels," and the
simulators may be termed "zappers."
Humidification: This ESD-control technique has been much maligned, for
example in comparison to ionization. One reason for criticism is that triboelec-
tric charging persists even at high relative humidity as shown in the following
table from Ref. 4-5:
STATIC VOLTAGES
ACTIVITY 20%RH 80%RH
Walking across vinyl floor 12kV 250 V

Walking across synthetic carpet 35 kV 1.5kV
Arising from foam cushion 18kV 1.5 kV
Picking u polyethylene bag 20kV 600 V
Sliding styrene box on carpet 18kV 1.5 kV
Removing Mylar tape from PC board 12kV 1.5kV
Shrinkable film on PC board 16kV 3 kV
Triggering vacuum solder remover 8kV I kV
Aerosol circuit freeze spray 15kV 5 kV
However, the fact remains that triboelectric charging becomes especially

troublesome below about 20 or 30% relative humidity, as shown by the high volt-
ages attained at 20% RH in the above table. Incidentally, precise relative humidity
values should include an associated temperature. A temperature factor in involved
in surface resistivity as shown in Paper No. 17 in the Appendix. In the above
table, a temperature of about 70°F is assumed. Another problem at very low rela-
tive humidity, e.g., 10% at 70°F, is that the sweat layers on antistatic materials
may work less effectively.
Several publications, e.g., Ref. 4-11, have mentioned the seasonal correlation of
naturally occurring high humidity (usually in summer) and increased yields of
ESDS products because of less ESD damage. Thus humidification, e.g., to 30 or
40% relative humidity, minimum, is surely desirable, but drawbacks include (1)
expense of facilities for adding water to the air, (2) possible adverse effects such as
delamination of polyimide circuit-board laminates or corrosion of metals if the
humidity becomes too high, and (3) the psychological factor of false confidence
inspired in operators and even engineers. If humidification can be done cost-
effectively, unobtrusively, and without ill effects such as corrosion, do it. But keep
in mind that ESD problems will not be eliminated but only lessened. No disci-
plines may be relaxed. As proof that ESD damage can occur at 40-50% relative
humidity, note the considerable charges, voltages, and MOSFET damage recorded
in Papers No.6 and lOin the Appendix.
Insulator: See "Nonconductor."
Ionization: As we use this term, it refers not to the phenomenon in general
but its application, if the form of air ionization, to ESD control.
Ionization is not a cure-all, as stressed in the following opening remarks we
made in the Ionization Workshop at the 1988 EOS/ESD Symposium in Los Angeles:
Ionization for no valid, well-defined reason is useless and may actually be part of the
ESD problem by creating charges on ungrounded conductors; that such charging can
occur is indicated by GIDEP Alert H7-A-85-02, issued in 1985, on an electrically unbal-
anced AC ionizer. Hence, ionizer manufacturers have shown much concern about bal-
ancing, which in some models is now automatic.
In contrast to humidification, which reduces triboelectric charges by 'nipping them in

the bud,' i.e., by inhibiting them in the fIrst place, ionization can neutralize a charge only
after it has formed. Thus, ionization is a process of correction rather that prevention.
However, ionization does have its place in ESD control when used judiciously. Local
ionizers are helpful in tribocharging processes such as grit-blasting or peeling tape from
a roll, and room ionization does control standing charges on common plastics, garments,
etc. Room ionization is claimed to have an incidental 'air-cleaning' effect (repulsion of
particles from surfaces or precipitation or particles away from work zones) in cleanrooms.
However, particles adhering to surfaces, such as walls, by 'static cling' can be undesir-
ably released by area ionization so that contamination is increased at least temporarily.
In conclusion, ionization is not a harmless vitamin pill but is strong medicine which as
possibly dangerous side effects and should be used only when risks are carefully con-
trolled and are outweighed by benefIts.
Electrical ionizers have the possible problem of charging isolated conductors

as mentioned above, whereas nuclear ionizers have a possible problem of polo-
nium-210 contamination; one brand of nuclear ionizers was recalled by the Nuclear
Regulatory Commission in February 1988 (Ref. 4-12). Polonium-21O is harmless
unless inhaled or ingested, but the oral radiotoxicity is similar to that of pluto-
nium-ifthe polonium-21O escapes from insoluble microcapsules that would be
expected to pass harmlessly through the gastrointestinal tract (Ref. 4-13). The ra-
diation hazard seems low, but it must be considered.
Slow-pulsed DC ionizers present a special problem, as explained in Example 21
of Chapter 8. We have reversed the recommendation of these ionizers which we
made in Paper No.3 in the Appendix, because when writing that paper we wrongly
assumed that electrically isolated conductors would be grounded in all SSW s. In
practice, conductors such as microscopes with rubber feet were not grounded and
were observed to be throwing arcs to adjacent conductors when near a pulsed DC
ionizer operating at 22 cycles/minute. Since AC ionizers do a similar job less dan-
gerously, we advise using them instead of the slow-pulsed DC type. However,
faster-pulsed DC ionizers may be acceptable; see Example 21 in Chapter 8. Note
that even AC ionizers can charge electrically isolated conductors if the ionizers
go out of balance. Hence, we require autobalancing ionizers in paragraph 3.6.8 of
Model Specification 1 in Chapter 9.
Beware of ozone generation as mentioned in paragraph 5.2.1 of Model Specifi-
cation 1 in Chapter 9. The rule of thumb is that if ozone can be smelled its concen-
tration is too high.
Ionizing blow-off nozzles are a special case. Gases when free of solids and liquids
do not cause tribocharging (Ref. 4-14), and "many shop and dry nitrogen lines have
been tested and no charging has been observed on aluminum and Teflon surfaces"
(Ref. 4-15) though a later study did show low voltages with shop air (Ref. 2-10).
Therefore, "installing ionizers on air lines because of charging fears is unwarranted"
when dust is not of concern (Ref. 4-15). However, flying dust or other particles can
cause significant triboelectric charging, so ionizing blow-off nozzles are a desir-
able precaution, and their use is suggested in MIL-HDBK-263. Ordinary air ionizes
better than dry nitrogen, so a nitrogen nozzle can be configured to allow air to be
aspirated into the gas stream ahead of the ionizing element; this feature appears in
electrical designs but not in nuclear ones, so the electrical type should be used on
nitrogen lines for best results. See Example 38 in Chapter 8 on air ionization.
Note that there are several possible causes of static charge formation in a blow-
off operation. These include (1) triboelectric charging by flying dust or other
particles or by sliding of displaced objects, (2) evaporation of nonaqueous sol-
vents, (3) deformation of solids (piezoelectric effect), and (4) cryogenic charging,
e.g., by freezing of water. In the last case, small splinters supposedly are ejected
from the freezing drop, and they either carry off or leave an excess of electrons so
that the remaining drop freezes to a charged ice particle. (See Chapter 2.)
In conclusion, ionization has its pros and cons. Pros are: (1) background static
charge levels are reduced, (2) there can be an air-cleaning effect, and (3) ionization
is the most practical method of neutralizing charged nonconductors; indeed, ioniz-
ers are sometimes essential, as in solving the problems in Examples 25 and 34 of
Chapter 8. The cons are: (1) dangerous space charges which charge ungrounded
(electrically isolated) conductors are possible, (2) ozone may be generated above
the OSHA limit, (3) the neutralization process (rate of charge decay by air ions
impinging on a surface) is relatively slow so that ESD damage can occur in the
course of triboelectric charging (the field created by pulling tape from a roll will
light a neon bulb in the midst of a stream of ionized air), (4) work areas are incom-
pletely covered by blowing ionized air because of shadowing by obstructions, and
(5) ionization is ineffective when a static charge is hidden by voltage suppression.
As an example of the latter, a charged sheet of plastic lying on a workbench surface
is immune to neutralization by ionized air because the field is collapsed and cannot
attract air ions of the opposite polarity. However, the field reappears and may cause
damage by the FIM when the plastic is lifted from the surface. To put it another way,
the plastic sheet lying on the surface makes an effective capacitor so that C is rela-
tively high and V is very low in the equation Q = CV (see "Capacitance" above).
In view of these advantages and disadvantages, we recommend using ioniza-
tion only "when risks are carefully controlled and are outweighed by benefits," as
we said above. Select ionizers from a major manufacturer with an established repu-
tation, and insist on monitoring and autobalancing features to minimize the space
charging hazard (see Example 42, Chapter 8).
In the years from 1989 to 1995, room ionization as a general precaution has
become increasingly popular, especially in electronics factories in the Far East
with large installations, e.g., 50,000 fe. Also, ionization to reduce airborne par-
ticle counts in c1eanrooms has gained more acceptance. Improved autobalancing,
erosion-resistant emitter points, and other evolutionary developments have en-
couraged the acceptance of ionization techniques.
Latent Failure: This is one of many controversial subjects in the field of ESD
control. Latent failure is accepted as being a real phenomenon (for example, Refs.
4-2,4-16, and 4-17). But the question remains: How frequent are latent failures?
The consensus of opinion by leading ESD experts is that these failures are indeed
real but are very rare. Nevertheless, one latent failure in a high-reliability aerospace
system, for example, might be catastrophic, so full ESD precautions must be taken
even if the chance of latent failure is remote. The above comment on rarity per-
tains to total failures, but some device experts claim that malfunctions such as
intermittency are not rarely but usually caused by ESD events in the manufactur-
ing process. In critical applications such as life-support systems or computerized
weaponry, such malfunctions, though short of complete failure, might themselves
be disastrous. Also, replacing ESD-damaged chips in a defective end product gets
"very expensive terribly fast" (Ref. 18). In conclusion, the possibility of latent
failure, or at least latent malfunction, is a good reason for rigorous ESD control.
MM: In the Machine Model, popular in Japanese ESD literature, a metallic
tool participates in DI. The resistance is nearly 0 ohm.
MOSFET: This has been the "white rat" in much of our research. It is appro-
priate for worst-case testing when the ESDS devices being protected are less sen-
sitive than the MOSFET. Other methods of recording voltage transients may be
used, but MOSFET damage is graphic and convincing because an actual compo-
nent is degraded. Endless special tests with MOSFETs can be devised; see Ex-
amples 10, 12, 18, 19,21,26,27,30,31,33, and 35 in Chapter 8.
Nonconductive, Nonconductor: MIL-HDBK-263 describes nonconductors
as having a surface resistivity of 10 12 ohms/square or more, because a surface
resistivity below this level is needed for rapid bleed-off of static charges as dis-
cussed under "Antistatic" above. On a nonconductor, a static charge, true to its
name, is static or motionless more than long enough (for only nanoseconds are
needed) for its associated field to threaten ESDS items by the FIM.
Operator Disciplines: This subject is discussed in Chapter 7 on the SSW, and
disciplines are incorporated in Model Specification 1 in Chapter 9. No matter how
excellent the ESD-control methods and equipment, the operator is all-important,
as the pilot of a plane is all-important, because the best equipment in the world can
be misused; ESD damage can be done, or the plane can crash. See the discussion
of the importance of operator skiIIs in Reference Document 6 in Chapter 9.
Personnel Voltage Tester: In using this instrument, a charged person merely
pushes a button to record his instantaneous voltage. For examples of test data, see
Paper No. lOin the Appendix. This is voltage on the skin, of course, not on garments.
Shielding: Even materials which are less than conductive provide partial
shielding from fields; for example, an antistatic polyethylene bag gave 75% shield-
ing in a shielding/discharge test (Table 9 of Paper No.3 in the Appendix). How-
ever, conductors are required for a high level of shielding. We prefer a "buried"
layer of at least 0.25 mil of aluminum foil (or the electrical equivalent) for bags or
other containers for in-plant handling; see "Faraday Cage" above. For trouble-
shooting or solving unique ESD problem situations, shielding can be provided by
various conductive barriers, e.g., a chain link fence protecting an SSW in Example
29 of Chapter 8 or a connector shell protecting ESDS pins in Example 30.
Like grounding, ionization (used only when necessary), or the use of topical
antistat, shielding is a basic ESD-control technique.
Shunting: Shunting of ESDS leads or terminals does not assure complete
ESD safety. For the best measure of safety, the consensus of opinion at present is
that antistatic or static-dissipative shunt materials, e.g., foams, are insufficient,

and conductive materials are required. However, the conductive shunt should be
covered with antistatic or static-dissipative material to "soften" external dIs-
charges if these may occur. Our test on black conductive (graphitic carbon-loaded)
"noncorrosive" foam with low sulfur and chloride content showed it to be ad-
equate for at least short-term high-humidity service, as was found for a solid shunt
bar of conductive plastic in Example 14 of Chapter 8. However, long-term gal-
vanic corrosion is possible as explained in that example, and the issue of galvanic
corrosion must be considered in selecting and using any shunt material. See MIL-
STD-889 and regard carbon-loaded plastic as "graphite."
Special Test: Numerous special tests appear in Chapter 8. The strategy is to
start with conditions that are extremely worst-case, giving the highest confidence if
the test is passed, and then, if the test is failed, back off to less exaggerated condi-
tions that still give high confidence. An example is the replacement of a highly ESDS
coupon design with a less-sensitive one, which still gave high confidence when is
passed the test; see Summary point No.5 of Paper No.8 in the Appendix.
A special test can solve any ESD-control problem if the investigator is suffi-
ciently ingenious. Remember that the object of testing is to reveal ESD hazards and
then impart confidence in the means taken to solve them. The only danger of worst-
case testing, besides revelation of unpleasant truths, is that an overkill solution might
be impractical or expensive but, if necessary, further tests at progressively milder
conditions will give enough data for a cost-effective compromise between confi-
dence and expense to be chosen. Overkill is desirable when cost-effective, as in the
case of foil laminate bags that give far superior protection without costing more than
see-through metallized bags (see Chapter 6) or a stone house built near a quarry.
SSP: We prefer a Faraday-cage package but allow a loophole, involving a
discharge test (paragraph 4.3.3 in Model Specification 2 in Chapter 9), for suppli-
ers to use nonfoil construction. See Chapter 6 for a full discussion.
SSW: See Chapter 7 for a full discussion.
SSZ: The SSZ is the interior of the SSP, or a zone within the SSW, wherein
the Basic Rule is never violated. See Chapters 6 and 7.
Standard Test: Examples of government or industry standard tests are: the
"BM test in MIL-STD-1686, surface resistivity by ASTM D257 or DIN 53842,
decay time by Method 4046 of FED-STD-l 0 1, resistance tests according to NFPA
99 (formerly NFPA 56A), and a test for triboelectric charging inside bags in
Appendix G of EIA 541. (EIA 541 is an Electronic Industries Association standard.)
Our own standard tests include the discharge test in Model Specification 2 in
Chapter 9, the surface resistivity test with a Voyager SRM-ll 0 meter or equivalent
(using our own criteria for classifying surfaces as antistatic or nonconductive),
the wet-bulb/dry-bulb method (psychrometry) for relative humidity, and the roller
test, drag test, scuff test, and walk test for static-limiting floor finishes in Paper
No.6 in the Appendix.
Static Charge: For practical purposes, electrons can be imagined as "rubbing
off' one surface to "contaminate" another, thus leaving the first surface positive
and making the second one negative. Only if the surface resistivity falls below
about 10 12 ohms/square will the electrons cease to be static (stationary) and slide
from the surface to give the rapid bleed-off that an antistatic surface needs to be
useful in ESD control.
A static charge commonly results from triboelectric charging, but there are
other sources, including freezing of water, evaporation of solvents, ion and elec-
tron beams, atomizing of liquids (spray charging), photoelectric charging, corona
charging, and deformation of solids (piezoelectric effect of crystals). See Chapter 2.
Static-Dissipative: We use the old definition from DoD-HDBK-263.
Static Field: See "E Field" above.
Static-Limiting Floor Finish: This subject is thoroughly discussed in Paper
No.6 in the Appendix.
A static-limiting floor finish is a backup or safety-net precaution like
humidification. Another is room ionization, which we don't recommend except
for cleanrooms when advantages, e.g., cleaning of the air, are demonstrable and
risks minimal. Unlike these other methods, the special floor finish is inexpensive
and quite unobtrusive. Tile floors must be "waxed" somehow, so why not use a
static-limiting floor finish? Choose one that is durable and easily maintained by
standard, simple procedures. Triboelectric charging of shoe soles is the property
of interest, but surface resistivity is related to this for a given finish and is conve-
niently monitored; 10" ohms/square by a Voyager SRM-110 meter is a typical
allowable maximum before the floor coating must be restored. (Note: Among dif-
ferent finishes, the correlation between triboelectric charging and surface resis-
tivity is imperfect; see Paper No.6 in the Appendix.)
Remember that basic floor-finish properties, as well as ESD properties, are vi-
tal. For example, slip resistance must be retained throughout the wear cycle for
employee safety.
Be careful to follow the floor-finish supplier's instructions for maintenance of
the finish. For example, don't use diluted finish as a cleaner, because the resulting
sticky deposit will collect dirt; use a special floor-finish cleaner, containing an
antistat, which will restore the antistatic property, without stickiness, to the surface.
Surface Resistivity: Values obtained for this property depend on electrode
configuration, pressure on the electrodes, roughness of the surface being mea-
sured, applied voltage, and other factors, so that methods of measurement are a
deep subject (Ref. 4-19). However, for our purposes, an approximate, easily made
determination by a Voyager SRM-II 0 meter or equivalent is sufficient. The criti-
cal property is bleed-off time for static charges and, since we don't have an exact
criterion for bleed-off time, which is arbitrary, we don't need precise values for the
property, surface resistivity, that controls bleed-off time. As we said in a discus-
sion of test methods (Ref. 4-1), "We need only to know if a feeble current will run
across [a plastic surface). So trying to split hairs on the exact resistance is not very
productive thinking." In other words, an acceptable surface resistivity value, e.g.,
below 10 12 ohms/square, is just an indication that enough current will flow across
a surface so static charges can't accumulate.
One surface resistivity meter, such as the Voyager SRM-ll 0 or SRMJRTG Meter,
which measures resistance to ground also, is desirable for every assembly area
where ESDS items are being handled. The meter will see much use in checking
floor finish, questionable plastics, performance of antistatic materials when the
relative humidity drops, etc.
Topical Antistat: Topical antistat treatment is almost a panacea for ridding
the SSW of nonconductors. We say "almost" because topical antistats have their
deficiencies: invisibility, fugitivity, and possible contamination.
Because of the invisibility of the antistat, treated surfaces must be "so labeled," as
required in paragraph 3.6 10 of Model Specification 1 in Chapter 9. Fugitivity means
that the antistat will disappear by vaporizing or wearing off at some unpredictable
rate, so periodic checks must be made with a field meter; fortunately, this is easily
done. For possible contamination effects, see Paper No.9 in the Appendix. If corro-
sion by chloride ion is feared, e.g., when silicon wafers are being handled, use one of
the "chloride-free" (actually, very low in chloride) topical antistats on the market.
Despite the above limitations, topical antistats are a good friend in ESD con-
trol, but don't overdo your reliance on them. Eliminate as many nonconductors as
possible before antistat-treating the necessary ones remaining in the SSW.
Triboelectric Charging: The triboelectric series is only approximately valid;
for example, an anomaly is noted in the complete version of Paper No. 6 con-
densed in the Appendix. However, for reference, here's a brief version including
modem materials and deleting the irrelevant fur and sealing wax:
Positive End
Air
Hands
Glass
Hair
Quartz
Nylon
Wool
Silk
Aluminum
Paper
Cotton
Steel
Wood
Rubber
Gold
Polyester
Polyurethane
Polyethylene
Polypropylene
Vinyl
Aclar (polychlorotrifluoroethylene)
Silicon
Teflon
Negative End
In theory, when two substances from the list are rubbed together, the one higher
on the list becomes positively charged. Thus Aclar, which is a useful material
when triboelectric charging (or "triboelectrification") is wanted, almost always
takes a negative charge. Note that the same material, e.g., polyethylene, can charge
itself so that one area of a sheet may be positive and another negative. Also, "circles"
occur in which A is charged positively by B, B is charged positively by C, but C is
charged positively by A instead of being charged negatively as it "should" be.
Experimentally, triboelectric charging is notoriously erratic.
Triboelectric charging tests include the bag shaker method using quartz and Teflon
disks to bracket most of the triboelectric series (Appendix G of EIA 541) and the
roller test or drag test illustrated in Figs. 5 and 6 of Paper No.6 in the Appendix.
Voltage Suppression: This is both a concept and a technique for ESD control.
A good example of voltage suppression is the minimal apparent charge that can
be observed on painted, grounded metals (allowed in the SSW as necessary non-
conductors in paragraph 3.6.10 of Model Specification 1 in Chapter 9) or plastic-
coated metals (also allowed in the same paragraph of Model Specification 1; note
the vinyl-coated steel tote box in Paper No.7 in the Appendix). Another example
is the relatively low apparent charge on hair hanging close to the skin (Example
2 in Chapter 8).
Use voltage suppression whenever possible to "kill" fields and achieve com-
pliance with the CD Rule.
Walk Test: This test is illustrated in Fig. 7 of Paper No.6 in the Appendix. It
may be augmented or replaced by measurements with a PVT-300 Personnel Volt-
age Tester as utilized in Paper No. lOin the Appendix. However, when using the
PVT-300, be sure to take at least 10 readings in order to catch, or partially catch,
the voltage spikes occurring when the shoe leaves the floor and capacitance drops.
The walk test is a good example of a bottom-line practical and realistic test that
directly measures the property of concern-voltage on people in this case. Labo-
ratory tests tend to be more reproducible but less relevant to real-life use of materials.
Zapflash: This inexpensive and convenient continuity-checker can also be
used to demonstrate various ESD phenomena such as voltage surges on a person
caused by triboelectric charging when he shuffles his feet on a carpet. We recom-
mend that every SSW have a Zapflash for verifying grounding as well as a field
meter for use in complying with the CD Rule.
"Zapflash" is a trade name of Anderson Effects, Redlands, CA. Besides show-
ing continuity up to about 7 megohms, the Zapflash can be used for checking
batteries, lamp filaments, and fuses, for detennining diode or DC voltage polarity,
for revealing AC power leakage to grounded equipment cases or machinery (which
could be a DI hazard), and for identifying the hot lead of an AC outlet.
Chapter 5
The Charged Device Model (CDM)
This model was born in 1974 with the proposal by Speakman that a part such as an
integrated circuit might be damaged or destroyed by rapid discharge of static elec-
tricity accumulated on the part's own body (Ref. 5-1). In 1980, the CDM was
judged the predominant failure mode at AT&T, as typified by dual-in-line pack-
ages (DIPs) sliding within their plastic packaging tubes to become triboelectrically
charged with most of the charge residing on the lead frame (Ref. 5-2). In 1992, it
was said, "It has become clear in the past five years that the CDM, not the 8BM,
is responsible for the vast majority of ESD damage, certainly to discrete devices
and probably to entire circuit cards and assemblies" (Ref. 5-3). Reflecting industry
concern about the CDM, certain packaging materials are now being advertised as
"CDM-safe."
Because of concern about the CDM, we attempted to define how conductive a
"conductive" surface must be to cause the "rapid discharge" required by this
damage mechanism. Our results appear in detail in the appended Papers No. 12
and 13 and in summary form in Papers No. 14, 15, and 16. The importance of
CDM safety in packaging is mentioned in Paper No. 20.
In essence, we found that thin-film resistors (current-sensitive) were relatively
CDM-resistant, but MOSFETs, represented by a Static Event Detector™ (3M
Co.), were very vulnerable. In fact, even antistatic polyethylene could be a threat
to MOSFETs at 50% relative humidity though not at 35% (Paper No. 16). Resis-
tance to ground measured through a device lead was 10 10 ohms for no CDM fail-
ures, 109 ohms for some failures, and 105-108 ohms for consistent failures (Paper
No. 16). Skin resistance through the device lead was }Q6-107 ohms (Papers No. 13
and 16), identifying fingers as a prime CDM danger. Hence we say, "A grounded
operator is the lesser of two evils versus an ungrounded operator" among the mot-
toes in Chapter 14.
We've mentioned the CDM danger to DIPs, and a remedy is to let the parts fall
onto an antistatic surface, preferably 1010-1011 ohms/square, when they are emp-
tied from a tube or rail. But the Field-From-Board (FFB) mechanism (Chapter 4)
is another problem. In this damage mechanism, ESDS parts such as integrated
circuits are mounted on a circuit board which becomes triboelectrically charged;
44
THE CHARGED DEVICE MODEL (CDM) 45
Finger
of Grounded
Operator
Negatively
Charged
Board with
Devices
Fig. 5-1. Field-from-board (FFB) mechanism.
then, touching a lead or terminal with a tool or a finger could cause a damaging
discharge.
Figure 5-1 diagrams the FFB. The negative charge on the board has polarized an
approaching fingertip by driving electrons off onto the operator, and a discharge,
which may involve a spark depending on voltage and capacitance, is imminent.
What must be done to assure that handling of ESDS items is CDM-safe and
FFB-safe? As mentioned in Chapter 4 in the discussion of the Basic Rule, we
favor a practical, cost-effective mix of antistatic, static-dissipative, and conduc-
tive materials with minimization of the conductive ones. Our suggestion is that
ESDS items with a sensitivity of over 100 V by the HBM can be handled safely
with operators' bare fingers if circuit boards are frequently checked for E fields
with a static meter and "washed" with ionized air from a blower as required. The
CDM and FFB require two elements-a static charge and a conductive surface-
and in this case we are removing the charge.
However, ifthe ESDS items are supersensitive, e.g., 12 V, extreme precautions
will be needed (Chapter 15), and it will be prudent to attempt to eliminate both
static charges and conductive surfaces. This means that operators must wear fin-
ger cots, which should be antistatic (at least 109 but not more than 10 12 ohms/
square), not static-dissipative, as demonstrated in Paper No. 16. Now the operator
is no longer the lesser of two evils but has been eliminated as a CDM/FFB hazard.
We've confined the wearing of cots to handling supersensitive items because
the cots are inconvenient to use and interfere with the tactile sensation, or "feel,"
needed for delicate assembly work. Perhaps antistatic cots of thin, tight-fitting
latex (or elastomer resembling latex) like the fingertips of surgical gloves would
solve the "feel" problem.
In conclusion, we suggest that you read the detailed information in the Appen-
dix and be on guard for the CDM danger in your operations. Strictly enforce the
CD Rule, with special attention to circuit boards with mounted devices. "No charge,
no discharge." Also, insist on ESD-protective packaging with a CDM-safe (anti-
static) surface.
Chapter 6
The Static-Safe Package (SSP)
By the Basic Rule, an ESDS item must always be in an SSZ. This SSZ will be in
(1) an SSP, (2) an SSW, or (3) elsewhere under continuous operator supervision
(paragraph 3.5.2 of Model Specification 1 in Chapter 9). An example of "else-
where" is an open tote box being carried between SSWs by an operator whose
wrist-strap cord is connected to an overhead grounded trolley.
In this chapter we're concerned with the SSP, which is defined in Chapter 4. The
SSP must have an outer Faraday cage or at least pass a discharge test in accor-
dance with paragraph 4.3.3 of Model Specification 2 in Chapter 9.
But how is the SSZ maintained inside the SSP? Remember that we're guarding
from fields and discharges in accordance with the Basic Rule. Since the Fara-
day-cage wall of the SSP excludes external hazards, and there's no source of in-
ternal discharges except internal fields, the only internal hazard is the creation of
fields by triboelectric charging. But this is a big "only," because triboelectric
charging is a mysterious or at least erratic process affected by the least amount of
surface contamination. For example, shoe soles can become charged the opposite
of their original polarity when they become invisibly soiled, and sometimes a sheet
of plastic, oddly enough, can be folded over and rubbed to charge itself, becoming
positive in one spot and negative in another, perhaps because no two surfaces are
quite identical (Ref. 4-14).
Thus, the problem of designing an SSP is attaining high confidence that tri-
boelectric charging inside the package will be minimal. Approaches are: (1) run-
ning laboratory triboelectric charging tests, e.g., the test in Appendix G of EIA
541 or the roller test illustrated in Fig. 5 of Paper No.6 in the Appendix, to screen
candidate materials for an intimate wrap, (2) running a more realistic special test
in which coupons with MOSFETs are shaken inside the package as in Ref. 4-7,
(3) mounting the packaged ESDS item so that it cannot slide within the package
during shipment, and (4) protecting the item (but not completely, as mentioned
under "Shunting" in Chapter 4) by shunting the leads or terminals. In approaches
I and 2, the surface(s) of the item to be packaged should be tested in the triboelec-
tric charging test; the EIA 541 bag-shaking test uses quartz and Teflon to bracket
the triboelectric series (see "Triboelectric Charging" in Chapter 4) because they're
47
48 ESD FROM A TO Z: ELECTROSTAnC DISCHARGE CONTROL FOR ELECTRONICS
at opposite ends of it, but, since the series is fallible, this method of scoping all
materials may fail in practice.
At the present state of ESD-control knowledge and material development, we
recommend using antistatic materials as the intimate wrap, keeping in mind the
question of their permanence (Paper No.2 in the Appendix) and taking no chances
of contamination (Ref. 6-1) or corrosion (Paper No.9 in the Appendix). If at all
possible, secure the item so that it can't slide when the package is vibrated in
shipment, and use the technique of shunting. Probably the worst hazard is static-
generating foam such as polystyrene foam "peanuts." If such a material must be
used, make very sure it's antistatic by a practical shaking test followed by a check
with a field meter. Remember that the package may find itself in relatively dirty
air. Will the cushioning material remain antistatic at, say, 10% relative humidity
at 70°F? Testing is necessary. In our experience, a good brand of antistatic bubble-
wrap is a safer cushioning material than "peanuts." The bubble-wrap can't break
open to expose fresh, nonconductive surfaces, and its resilience tends to hold
packaged items in place. Use it whenever possible. Cardboard boxes with a buried
layer of aluminum foil and with antistatic bubble-wrap inside are excellent pack-
ages.
Though we allow a variety of shielding materials to accommodate suppliers, as
long as the discharge test in Model Specification 2 can be passed, we much prefer
a Faraday cage (by our definition). In a hierarchy of tests (Paper No. 4 in the
Appendix), a foil laminate bag was arbitrarily ranked as "good," while a bag con-
struction with heavy foil (opaque, of course) or metal screen (partially transpar-
ent) was "excellent." Bags with see-through metallization were at best "fair." (Note:
In later tests, a "metal-in" see-through bag was "good"; see Paper No. 15.) Signifi-
cantly, two groups of investigators at major corporations arrived independently at
the same conclusion that foil provides very effective shielding whereas see-through
metallization, which is on the order of only 100 angstroms thick, is dubious. At the
1984 EOS/ESD Symposium, the Rockwell group said, "Only foil-containing lami-
nate bags are recommended for Faraday-cage protection against worst-case fields
and discharges" (conclusion 12 of Paper No.3 in the Appendix), and the British
Telecom group said, "Highly static-sensitive components should always be pro-
tected by metal-foil bags" (Ref. 6-2).
See-through metallized bags have become popular because of their partial trans-
parency, and a capacitance-probe (shielding/discharge) test (in EIA 541) has
been contrived by the bag manufacturers and accepted by naive users. The first
problem with this test is that the voltage is too low (1000 V); the British Telecom
tests (Ref. 6-2) used higher voltages, e.g., 1600 V, which can easily be encountered
in practice. Furthermore, Ref. 6-2 shows that as the area of the electrodes falls, the
shielding effectiveness of see-through metallized bags decreases. This is in the
direction of a spark discharge, and the second problem with the test is that spark
discharges, which are not accounted for by the large flat electrodes used, burn off
metallization and can damage ESDS items in the bag; a person can walk on a
carpet at 2~30% relative humidity at 70°F and touch his finger to a see-through
THE STATIC-8AFE PACKAGE (SSP) 49
metallized bag with the metal protected only by a thin polymeric coating to "blow"
a MOSFET inside; see Fig. 2 and Table 3 of Paper No.4 in the Appendix. Use
only "metal-in" see-through bags, with the metallization protected by a O.S-mil
Mylar (polyester) film.
Admittedly, the MOSFET was configured in the worst possible way (leads con-
nected to upper and lower "antennas"), so that the chance of ESD damage was
maximized, but why take chances? We agree with Ref. 6-3 that "packaging cost
differences of as much as one or two dollars for product protection in these areas
(precision medical, communications, personal hazard, or defense application) are
a minor consideration ... EMIIRFI testing may also be appropriate in such cases."
Yes! And foil or screen gives EMI/RFI protection along with protection from spark
discharges. Furthermore, foil laminate bags cost little or no more than see-through
metallized bags! Screen bags (see Paper No.4 in the Appendix) do cost consider-
ably more, but an extra "one or two dollars" is trivial when the product in the bag
is worth $100,000 or its latent failure might cause a $10,000,000 catastrophe.
Being penny-wise and pound-foolish in this case is not only risky but may be
unethical where health or defense is involved. Note the suppression of data men-
tioned under "Discharge Test" in Chapter 4. Also, MIL-HDBK-773 (Ref. 6-4)
insists on foil bags. However, the opaque metallization in the new MIL-B-8170SC,
Type I, is not foil but is very effective (Paper No. 15).
But we're tired of being a voice crying in the wilderness. Let the world go its
way and make believe that a wispy 100 angstroms of metal, which is not an EMI/RFI
shield and, if unprotected, is easily evaporated by a spark from a fmger, is a "Fara-
day cage"!
For recommended Faraday-cage tote box constructions, see Paper No. 7 in the
Appendix; the shielding layer in these is foil as a minimum.
Here's a final word on corrosion and contamination (the subject of Paper No. 10
in the Appendix). We advise running an accelerated special test with your ESDS
item inside of, or against the surface of, the proposed intimate wrap, as a test was
done with a connector and the shunt bar proposed for it in Example 14 of Chapter
8. Generally suitable high-humidity conditions are MIL-STD-202, Method 106,
with or without the vibration or freeze cycles as appropriate, for one month. This
rigorous accelerated test probably will disclose long-term problems, but we sug-
gest an accompanying real-time test at slightly elevated temperature, e.g., lOO°F,
with samples being checked every few months. The results of the real-time test
will give you a good idea of what is happening to your product in storage. Where
corrosion is concerned, "never assume anything."
For suppliers of ESD-control packaging materials, see buyers' guides such as
the one from Evaluation Engineering magazine, Nokomis, FL.
Chapter 7
The Static-Safe Workstation (SSW)
How the SSW Holds the SSZ

As mentioned at the start of Chapter 6, an ESDS item must always be in an SSZ.
This chapter is concerned with the means of maintaining an SSZ inside an SSW.
First, note that the SSZ may be irregularly shaped and may shift position within
the SSW. For example, if an ionized air blower is being used, the SSZ will end 12
inches from the corona-discharge points (paragraph 3.6.8 of Model Specification
I in Chapter 9), and when the ionizer is shifted on the workbench, the SSZ will
also shift. The operator is responsible for maintaining the SSZ and knowing its
boundaries at all times. We stress again that his or her skill, not the selection of
materials and equipment (though these must be adequate), is the most important
element of ESD control.
Since the SSW is a volume in space, it is not just a working surface. Floors,
ceilings, and walls mayor may not be included, and the volume in space may
contain only part of a workbench or no workbench at all; for example, an oven
might be an SSW. Whatever the SSW is, it must be "designated" (see defmition in
Chapter 4), which means clearly marked, e.g., with strips of yellow tape.
Can an SSW have any fields at all? Yes, but they are known and controlled so
that they're always harmlessly weak within the SSZ as required by the CD Rule.
If no fields whatsoever were permitted within the SSW, the operator himself or
herself would be excluded because hair and clothing "radiate" fields. These fields
must, of course, be sufficiently attenuated by distance at the point where the op-
erator is handling the ESDS item; if the attenuation is insufficient, so that the CD
Rule is violated by too short a distance from apparent charges on hair or cloth-
ing, corrective actions are required, e.g., tying the hair closer to the skin or chang-
ing the clothing, for example from wool to cotton.
Thus the operator's head and body are outside the SSZ while his or her hands
are inside it. Why not try to force the whole operator, chair and all, to meet SSZ
requirements? The reason is cost-effectiveness. We're using operator skills to mini-
mize purchases of antistatic smocks and conductive chairs as well as to avoid
troublesome restrictions such as the wearing of hair nets, and experience has shown
50
THE STATIC-SAFE WORKSTATION (SSW) 51
that this cost-effective approach is feasible. The operator need only remember that
ESDS items must be "kept at arm's length" and not be unnecessarily brought near
hair, clothing, or chair upholstery.
Similarly, walls and floors with fields harmless by the CD Rule may be inside
the SSW. In fact, the SSW should be considerably larger that the SSZ so that
sources of fields near the SSZ are controlled. Consider Example 29 in Chapter 9.
In this case, the boundary of the SSZ was 18 inches from a chain link fence which
provided shielding from fields on passers-by. Should the boundary of the SSW be
the same as that of the SSZ, or should the SSW extend all the way to the fence? It
should extend to the fence so that the 18-inch "buffer zone" along the SSZ is
subject to SSW requirements and new field sources won't be added. If the SSW
ended 18 inches from the fence, common polyethylene packaging film carrying a
high apparent charge, e.g., -10,000 V, might be "legally" heaped between the
fence and the SSW. In theory, the operator would note any violation of the CD
Rule by using his or her field meter and would remove the polyethylene, but if the
SSW had extended all the way to the fence, the plastic wouldn't have been lying
there (assuming compliance with SSW rules) in the first place.
Though the language was unclear, the old "one-meter rule" of DoD-STD-1686,
now superseded by MIL-STD-1686, seemed intended to provide such a buffer
zone. The trouble with this rule was that it was arbitrary, and one meter might not
be a safe distance, by the CD Rule, from a high apparent charge. Since we have
the quantitative CD Rule, the extent of the buffer zone between the working sur-
face and the boundary of the SSW is optional; it may be nil where a workbench
touches a wall but may be a few feet from the other end of the bench as marked by
a yellow line on the floor. In this case, common nonconductive plastic couldn't
"legally" be left lying on the floor near the bench. Without wasting expensive
space, of course, such buffer zones should be maximized.
Certification of SSWs
See Reference Document 2 to model Specification 1 in Chapter 9.
Workbenches: Design and Grounding

A static-safe workbench should (1) hold no apparent charge over 50 V (when
ESDS items sensitive to no less than 100 V are being handled, as assumed in this
book), (2) allow charges to drain from charged objects placed on the work surface
(unless the charges are on nonconductors and are thus truly static charges that
can't flow), and (3) have appropriate electrical supply and ESD grounding facili-
ties including a grounding lug.
Requirement 1 is satisfied by constructing the bench of painted steel which is
ESD-grounded. A painted surface is nonconductive but shows a negligible ap-
parent charge because of voltage suppression by the steel; hence, "painted,
grounded metal" is one of the "necessary nonconductors" allowed in paragraph

3.6.10 of Model Specification 1 in Chapter 9.
Requirement 2 is met by selecting a work surface that is antistatic or static-
dissipative by a surface resistivity measurement (Ref. 6-1) and has a surface-to-
ground resistance (Ref. 6-1) of 1 x 1011 ohms, maximum. A conductive surface is
undesirable because it can be hazardous to workers, by carrying a lethal current,
and also to ESOS items, by allowing rapid discharges (see "COM" in Chapter 4).
The effects of wear, chemicals, staining, and heat on the work surface (Ref. 6-1)
must also be considered, as well as possible corrosion of ESOS items by impuri-
ties from the work surface. Corrosion is a treacherous problem because it takes
time to manifest itself. If silicon wafers will rest on the surface, consider a test for
extractable chloride ions in laminates as described in Example 17 of Chapter 8.
The importance of drain time is another controversial subject. See Ref. 6-2 for a
discussion of the danger of voltage suppression hiding a field when a charged
object such as a tote box rests on a surface so that the field returns to possibly
cause ESD damage when the object is lifted. Paper No. 7 in the Appendix makes
the point that contact area between object and surface complicates the issue; a
warped rigid conductive tote box with three small contact areas will drain far
more slowly that a similar box with a flat bottom giving a relatively large contact
area. In practice, a grounded operator usually touches a box or package before
reaching into it, so a conductive work surface, with its hazard to people and ESOS
items, shouldn't be used just to reduce drain time. We would rather err in the
direction of high drain time than of too-rapid discharge to the surface that might
cause COM damage.
Requirement 3 involves grounding the bench to a water pipe, building struc-
tural steel, or wet ground well as required in paragraph 3.6.5.2.1 of Model Speci-
fication 1 in Chapter 9. A resistor is needed between the bench-top laminate and
ground because direct grounding of the conductive interlayer of laminates is un-
safe (Ref. 6-3). If high voltages are being used and the utmost personnel protection
is needed, use ground fault circuit interruptors. These guard against the remote
possibility of an electrically overstressed resistor being carbonized so that it falls
below a safe resistance (Ref. 6-4).
Incidentally, a Zapflash is useful for locating AC power leakage onto cabinets,
etc.; touch the probe to the surface and the bulb will light if there is leakage. Sub-
sequent testing with a voltmeter will show the voltage.
For details on grounding of workbenches, see Model Specification 1 in
Chapter 9.
Humidification, Ionization,
Static-Limiting Floor Finish
These subjects are discussed in Chapter 4. In brief, use humidification if your
ambient air is often dry, if the expense of humidification equipment is bearable,
THE STATIC-SAFE WORKSTATION (SSW) 53
and if there are no hannful side effects such as corrosion; also, beware of false
confidence inspired in personnel. Use ionization only when there is no alterna-
tive, as sometimes happens; Examples 24 and 25 in Chapter 8 are cases where
triboelectric charging was unavoidable and only ionization could mitigate it.
Make the most of static-limiting floor finish, which is a very economical "safety
net" as seen in the cost calculation in Table 1 of Paper No.6 in the Appendix.
Materials
Allow "necessary nonconductors" as listed in paragraph 3.6.10 of Model Speci-
fication 1 in Chapter 9. Nonconductors are either ignored because their apparent
charge is low and they are safely far from the ESDS item, in accordance with the
CD Rule, or they are controlled by treating with topical antistat or by having
their fields attenuated by shielding, as by the chain link fence in Example 29 of
Chapter 8, or by voltage suppression, as by keeping hair close to the skin in Ex-
ample 2 of Chapter 8.
Allow conductive materials only when no antistatic or static-dissipative alter-
native is convenient. (Note: As of this writing, the more convenient EIA and MIL-
STD-1686 designation of "dissipative" is coming into use to replace these two
classifications with only one. This is a sensible change in nomenclature, because
the difference between antistatic and static-dissipative is arbitrary and is seldom
a useful distinction.) For categories of materials and criteria for their selection, see
Reference Documents 4 and 5 appended to Model Specification 1 in Chapter 9.
When selecting materials, remember that standard tests may not be sufficient!
Laboratory tests often give misleading results because real-life conditions are not
simulated. Hence, special tests with realistic configurations are often needed. An
example of an unrealistic test is the shielding/discharge or pulse test in EIA 541,
which ignores the realistic possibility of a spark discharge as discussed in
Chapter 6.
Another, far more serious example of unrealistic testing involves a type of aero-
space wire. The insulation on this wire has three different properties from conven-
tional wire insulations: (1) possible susceptibility to hydrolysis because it's a
condensation instead of an addition polymer, (2) unusual stiffness because the
insulation is high-modulus and must be wrapped on the conductor as tape, and (3)
high aromaticity (high carbon content in the molecule).
If realistic testing had been done before the wire was installed in airplanes and
space vehicles, the following would have happened: (1) exposure to wann water
and high-pH cleaning solutions, simulating wheel-well conditions in Navy planes,
would have caused hydrolytic breakdown, so the wire would have been restricted
from wheel-well applications, (2) vibration tests with simulated airframe assem-
blies would have shown chafing and shorting when the stiff wires rubbed together
(as opposed to no chafing with conventional wire whose soft insulation makes it
limp), so installation would have been done with this problem in mind, and (3)
flashover and arc propagation tests would have shown that an arc propagated even
in vacuum because of carbonization of the highly aromatic polymer, and care would
have been taken to prevent shorting that might start an arc, or the wire would have
been excluded from use in applications where the current was sufficient to sustain
an arc.
What really happened was that applications went forward based on standard,
unrealistic lab tests. For example, laboratory abrasion tests gave excellent re-
sults because they did not include the chafing situation where stiff wires rubbed
together, and standard tests for electrical properties failed to predict the arcing
problem. Thus untold millions of feet of the wire have been installed, and all
three problems described above have reared their ugly heads in the course of a
few decades (disasters can be slow in coming). The wire is now "in big trouble"
because realistic tests weren't run long ago. But enough. A word to the wise is
sufficient.
Equipment
Model Specification 1 in Chapter 9 requires grounding of all conductors, which
is not always easily done. For example, a stainless-steel oven shelf is grounded
directly, but the blade of a screwdriver is grounded to the operator's hand only
through the "sweat layer" of topical antistat applied to the nonconductive handle.
Microscopes present a special problem requiring judgment in bending the ground-
ing rules; metal eyepiece housings, for example, have small capacitance and are
not expected to touch an ESDS item, so they can be ungrounded. However, we
recommend grounding the major sections of the microscope even if wires must be
added to link them. Use the Zapflash to check continuity.
Soldering iron tips must be grounded so that there is a potential of no greater
than 2 mV between the hot tip and the grounded ESDS item being soldered, if
MIL-STD-2000 is being followed. Also, that standard requires less than 5.0 ohms
resistance between tip and workstation ground.
Electrical equipment must be checked for E fields, of course. Remember that
space charging can be a problem with ionizers, as discussed under "Ionization" in
Chapter 4, and don't let ionizers become part of the ESD problem they're sup-
posed to solve.
Treat plastic parts of equipment, e.g., housings, with topical antistat unless the
plastic is a coating that is relatively thin (less than 20 mils) and the metal is thick
enough (over 60 mils) to provide effective voltage suppression (paragraph 3.6.10(2)
in Model Specification 1 in Chapter 9). Beware of stresscracking polycarbonate
(Paper No.9 in the Appendix).
In special cases, beware of H fields (magnetic fields). MIL-STD-2000 requires
that H fields of soldering irons, soldering machines, and associated processing
equipment be less than 2 gauss measured at any surface of an item being
processed.
THE STA1lC-8AFE WORKSTA1l0N (SSW) 55
Continuous wrist-strap monitors (see Papers No.5 and 10 in the Appendix)

are a must; an example of a procurement document for these is Reference Docu-
ment 3 appended to Model Specification 1 in Chapter 9. Other essential pieces of
equipment are a field meter and a Zapflash.
For categories of equipment and criteria for their selection, see Reference Docu-
ments 4 and 5 appended to Model Specification 1 in Chapter 9.
General Design, Construction,

and Maintenance of ESD-Protected Areas
See Reference Document 1 appended to Model Specification 1 in Chapter 9.
Operator Disciplines,
Training, and Certification
In general, operators must maintain the SSZ and always know its boundaries within
the SSW. ESDS items should never be brought unnecessarily near or touched to
any surface, nor should an operator fidget or shuffle his or her feet unnecessarily.
In other words, though the best materials and equipment have been selected and
the ARTG is being met by using a continuous wrist-strap monitor, don't look
for trouble. See the "Conclusions" section of Paper No.3 in the Appendix. Opera-
tor disciplines are included as requirements in Model Specification 1, paragraph
3.6.12, in Chapter 9, and training and certification of operators are covered in
Reference Document 6 appended to that specification.
Chapter 8
ESD Troubleshooting:
Illustrative Examples
Troubleshooting, by which we mean the solving of unique problems as they arise,

is done using one or more of three approaches (analysis, standard test, or special
test) with the aid of other real or conceptual tools from Chapter 4. In the following
actual case histories, we describe each problem and state which of the above three
approaches was used. Then we present the solution in enough detail to illustrate
the reasoning process.
In principle, any novel ESD-control situation can be handled, as were these
examples, with the A-to-Z arsenal of ideas and techniques of Chapter 4. Solutions
to problems then can be incorporated into the handling specification as illustrated
in paragraphs 3.6.12.6 and 3.6.12.11 of Model Specification 1 in Chapter 9.
1. Ann Hair
Problem: Reference 8-1 suggested that arm hair might be an ESD hazard, with
charges up to 900 V having been measured. We suspected that such voltages are
atypical. What new operator disciplines, if any, should be introduced to control
fields from arm hair?
Solution: The approach was a special test. Polyester garment sleeves or sheets
of common nonconductive polyethylene were stroked across the upper surface of
operators' forearms, and the apparent charge was read. Because of voltage sup-
pression by the skin, in no case did the reading exceed 300 V, which represents a
harmless field for ESDS items sensitive to 100 V (Fig. 3 of Paper No.3 in the
Appendix). Therefore, no new operator disciplines were introduced. However,
operators were warned to keep the potential hazard in mind and run checks with
their field meters on especially dry days. Frequent field-meter checks of the whole
SSW, to enforce the CD Rule, are of course a vital part of our program.
56
ESD TROUBLESHOOTING: ILLUSTRATIVE EXAMPLES 57
2. Head Hair
Problem: Apparent charges of up to + 10,000 V were observed on the hair of
female operators; by the CD Rule, the safe distance from 10,000 V is 56 inches
(paragraph 3.6.11 of Model Specification 1 in Chapter 9). What new operator
disciplines should be introduced?
Solution: The approach was a special test. Apparent charges of up to + 10,000
V were found with hair hanging 6 to 8 inches from the skin, but when the hair was
2 inches from the skin, the maximum apparent charge was only +300 V, which is
a harmless level as noted in Example 1 above, because of voltage suppression.
Therefore, a new operator discipline was introduced: hair must be kept within 2
inches of the skin by tying back as required; this requirement appears in paragraph
3.6.12.6 of Model Specification 1 in Chapter 9.
3. Operator Clothing
Problem: What precautions should be taken regarding the static charges on cloth-
ing? Should special antistatic or other type of ESD-control smocks be required?
Solution: The approach was a special test. Apparent charges were measured
on shirts, blouses, and sleeves being worn by operators, and a few months' worth
of data showed that fields were sufficiently weak in the vicinity of the operator's
hands, where the ESDS item was, to meet the CD Rule as long as sleeves were
short or rolled to above the elbow. Therefore, street clothes are allowed if the
sleeves are short or rolled; with long, unrolled sleeves, ESD-protective smocks are
required (paragraph 3.6.12.10 of Model Specification 1 in Chapter 9). Smocks,
though optional, may raise ESD awareness by being "uniforms" and give factories
a more businesslike appearance. However, these marginal advantages must be
weighed against the initial cost of the smocks as well as the cleaning costs, which
may be considerable. A compromise is to use rolled-up sleeves in engineering and
test labs and smocks in assembly areas.
4. Identification Badges
Problem: Apparent charges of up to -2000 V were observed on plastic identi-
fication badges (Table 4 of Paper No.3 in the Appendix). What precautions should
betaken?
Solution: The apparent charges of up to -2000 V were observed when the
badge was held away from a surface; in fact, by triboelectric charging of badges
with polycholorotrifluoroethylene (Aclar, by Allied-Signal), the apparent charge
could be pushed as high as + 10,000 V. However, in special tests, when the charged
badge was held close to the skin, as a badge is worn in practice, the apparent
charge dropped to 500 V or less because of voltage suppression. The CD Rule
allows a distance of 12 inches from 500 V (paragraph 3.6.11 of Model Specifica-
tion 1 in Chapter 9), and ESDS items are seldom brought much closer than this to
a badge. Furthermore, badges are not highly charged in practice because they're
not rubbed as in our worst-case tests. The conclusion is that badges are not a prac-
tical hazard, but if there is any doubt, they can be easily treated with a topical
antistat.
5. Vinyl Chair
Problem: An apparent charge of +500 V was noted on a vinyl-upholstered chair
(Table of Paper No.3 in the Appendix), and higher voltages could be produced in
contrived triboelectric charging tests. Therefore, should ordinary chairs be al-
lowed in the SSW, or are conductive chairs necessary?
Solution: The approach was analysis. First, the operator's sitting in the chair
attenuates the chair's field by voltage suppression. Second, the bench top pro-
vides partial shielding. Third, field-meter checks in assembly areas have always
confirmed compliance with the CD Rule in regard to chairs. The conclusion is that
ordinary chairs are acceptable (paragraph 3.6.10 of Model Specification 1 in Chapter
9), but an operator discipline requires that ESDS items shall never be held be-
yond the front edge of the work surface (paragraph 3.6.12.2 of the same specifica-
tion).
6. Paper Forms and Labels

Problem: Should paper forms and labels, used to document manufacturing op-
erations, be allowed in the SSW?
Solution: The approaches were analysis and standard test. MIL-HDBK-263
classifies paper products as antistatic materials. Also, we checked some of the
subject forms and found a surface resistivity of 109 ohms/square at approx. 50%
relative humidity and nOE This value is at the low end of the antistatic range, and
experience has shown that unglazed paper products are not an ESD hazard except
in high-speed processing such as printing. The conclusion is that paper forms and
labels, if unglazed, are allowable in the SSW.
7. Charged Walls
Problem: Do charged walls present a field hazard when they bound an SSW,
e.g., when a workbench is against a wall?
Solution: The approach was a special test. Triboelectric charging of walls
with garment fabric or Aclar film gave less that 1000 V apparent charge because
fields tend to collapse on large surfaces (voltage suppression). At 1000 V, the safe
distance in inches by the CD Rule =(1000)0.5/1.8 = 18 (or use the convenient table
in paragraph 3.6.11 of Model Specification 1 in Chapter 9), and in practice ESDS
items are never nearer than this to walls in our operations. Therefore, walls are
judged not a practical hazard. If they were, they would be treated with topical
antistat in accordance with paragraph 3.6.10 of the same specification; note that
the topical antistat must be renewed at least every three months.
8. Charged Windows
Problem: Under dry conditions (relative humidity approximately 20% at 70°F),
an apparent charge of 5000 V was noted on the glass of a window. Triboelectric
charging is not caused by pure air or any other gas (Ref. 4-14), so the charging
was attributed either to electrically unbalanced air or to dust carried by the warm
desert wind which also caused the low humidity. How should the hazard of charged
windows be controlled?
Solution: The approach was analysis. By the CD Rule, the safe distance from
the window in inches =(5000)°.5/1.8 =39, so SSWs were arranged so that their
boundaries were at least 39 inches from windows. Otherwise, a nonsmearing topi.
cal antistat could have been applied to the glass, with field·meter checks to make
sure the charge was sufficiently lowered and remained in control.
9. Cathode Ray Tubes (CRTs)

Problem: CRTs (computer or oscilloscope screens) have a considerable positive
charge as needed for the electron beam to "paint pictures" with the phosphors on
the inside of the glass. Since CRTs are often found inside SSWs today, how can
this field hazard be controlled?
Solution: The approach was a special test. The apparent charge was found
to be about +2000 V with an unshielded tube but less than +300 V with a grounded
conductive mesh screen (Paper No. 8 in the Appendix). The screen was a com-
mercial CRT accessory which reduced the glare as well as providing shielding.
10. Connector Dust Cover

Problem: Was a nonconductive polyethylene dust cover suitable for a connec-
tor on a cable assembly leading to ESDS items inside an electronic box?
Solution: The approach was a special test as well as a standard test (the walk
test depicted in Fig. 7 of Paper No.6 in the Appendix). A person charged himself
to -10,000 V (measured by the walk test method) by shuffling on a carpet. Then
his finger touched the dust cover over the connector face just above a pin which
was connected to the substrate-case lead of a MOSFET at the other end of the 3-
foot cable, the gate lead of the MOSFET being grounded. When measured in the
usual way, the MOSFET was found to be damaged (either shorted or VGS(TH) shifted
0.04 V or more). Another MOSFET was installed and was damaged when the
charged person's finger was brought to within 1 inch of the dust cover. However,
when the cable, which was an unshielded wire bundle, was covered with a grounded
metal shield, five MOSFETs in a row survived the test in which the charged finger
touched the dust cover; then another five survived with the cover removed. When
the dust cover was replaced by aluminum foil wrapping the connector face but the
cable was unshielded, a MOSFET was damaged by the charged finger being
touched to the foil.
These tests show that ESD protection of the connector face was irrelevant. The
problem was that the field from the charged person's hand was being "picked up"
by the unshielded wire to the MOSFET so that this device was damaged by the
FIM; the dust cover could be charged to 10,000 V by triboelectric charging at
low relative humidity (10-15% at 70°F), so it itself might have caused damage by
the FIM with the cable unshielded.
The conclusion was that, when the cable was shielded, the dust cover could be
any material, even ordinary nonconductive plastic. However, good ESD-control
practice demands that an antistatic or conductive plastic cover, either of which is
commercially available, be used. (Note: A static-dissipative cover also would be
fine; this is an example of the uselessness of distinguishing between antistatic and
static-dissipative, which are suitably lumped together as "dissipative" in the cur-
rent revision of MIL-HDBK-263.) The most important result of the test was the
demonstration of the need for shielding the wire bundle of the cable.
Note that these conclusions pertain only to the particular connector-cable con-
figuration tested. For example, if a sensitive pin was less deeply recessed and was
closer to the dust cover, a charged finger might "zap" the pin through antistatic or
even conductive plastic; then a true Faraday-cage cover (buried foil or better)
would be needed.
11. Brushes
Problem: The suitability of four types of brushes for use in scrubbing ESDS
modules in SSWs was to be determined.
Solution: The approach was analysis, special test, and standard test (for sur-
face resistivity). The first brush had a varnished wooden handle, a phenolic band
(seemingly wood-flour filled) holding the bristles, and natural (horsehair) bristles.
At 65% relative humidity and 71 OF, the surface resistivity of the handle was 2 x
1012 ohms/square at 500 V and of the phenolic band was 4 x 10 10 ohms/square. The
bristles, being natural rather than synthetic, were approved by analysis in accor-
dance with MIL-HDBK-263. The conclusion was that the only part of the brush
that was a problem was the handle, which had to be treated with topical antistat
and so labeled (as an insulated tool handle, paragraph 3.6.10(9) of Model Specifi-
cation 1 in Chapter 9).
The second brush had an unfinished wooden handle and a metal band holding
the horsehair bristles, while the third brush was similar but with hog bristles. The
wood was found to be antistatic, and the bristles were approved by analysis as for
the first brush.
The fourth brush was a novel design created to solve a problem. The steel handle
of an acid brush was scraping module surfaces, so a one-inch section of the
nonconductive polyolefin heat-shrink tubing was shrunk over the handle to cover
the first eighth-inch of bristles (natural) and prevent the steel from touching the
surface being scrubbed. But would the polyolefin hold an unacceptable field? Tri-
boelectric charging of the detached polyolefin gave -390 V apparent charge,
which would require a safe distance in inches, by the CD Rule, of (390)0.5/1.8 =11
between the polyolefin and an ESDS item. However, when the polyolefin was
shrunk in place onto the steel handle, which was held by a grounded person, the
apparent charge was reduced by voltage suppression to -110 V, which is accept-
able on nonconductors (Fig. 3 of Paper No. 3 in the Appendix). The conclusion
was that the brush was acceptable for use without antistat-treating the polyolefin,
which in essence falls under the plastic-coated metals provision of paragraph
3.6.10(2) of Model Specification 1 in Chapter 9.
12. Screwdriver Handle

Problem: Could the plastic handle of a screwdriver be a field hazard in the
SSW?
Solution: The approach was to use special tests. The screwdriver in question
was 9 inches long overall with a cellulose acetate butyrate plastic handle 3 1/4
inches long and 1 inch in diameter. At approximately 50% relative humidity and
70°F, the handle was stroked with polyester fabric to give an apparent charge of
+1700 V. MOSFETs were damaged by the field (see Table 4 of Paper No.3 in the
Appendix), and the CD Rule calculation gives the safe distance in inches as
(1700)°·5/1.8 =23. The conclusion is that the handle indeed could be a hazard, so it
was treated with topical antistat and so labeled. This and similar tests led to the
provisions of paragraph 3.6.10 of Model Specification 1 in the Appendix.
13. Smocks
Problem: Were three different designs of antistatic or static-dissipative smocks
suitable for use inside SSWs?
Solution: The approach was a special test. The first smock, containing 1%
stainless-steel fibers, was tested by clamping the cloth between two sets of vise-
like electrodes; at 1~1000 V, the surface resistivity gauged by a megohmmeter
was 2-4 X 1010 ohms/square. Alternatively, the resistance was measured between
dammed-in pools of mercury to give 2 X 1010 ohms/square at 10 V.
The second smock had black conductive threads in a crisscross grid and mea-
sured 6 X lOS ohms/square (static-dissipative) by the clamp method. A third smock
had black conductive threads 4 mm apart and running all the same way so that the
result of the clamp test was 8 X 107 ohms/square with the black threads perpen-
dicular to the electrodes but greater than 1013 ohms/square with the black threads
parallel to the electrodes.
The conclusion was that the first two smocks were suitable in terms of surface
resistivity, but the third smock was doubtful. Note that the use of smocks in general
was judged to be optional in Example 3 above. Before the first design could be
used, the possibility of its shedding stainless-steel fibers would have to be checked,
because these might cause shorts in ESDS items, e.g., between adjacent circuit
lines on printed wiring boards. GIDEP Alert D5-A-84-0l (April 25, 1984) de-
scribes just such an occurrence.
14. Shunt Bar

Problem: Would a black conductive plastic shunt bar cause corrosion of gold-
plated Kovar connector pins?
Solution: The approach was a special test. An edge connector was cut in half,
and one half was fitted with the shunt bar while the other half served as a control.
The pins were cleaned by vapor-degreasing with Freon TES and washing with
deionized water before starting the test. After 31 days under high-humidity, el-
evated-temperature conditions (MIL-STD-202, Method 106, without freeze or vi-
bration cycles), there was no sign of corrosion and no difference in appearance of
the shunted versus control pins. The conclusion was that the shunt bar was suitable
for general use at moderate relative humidity. Gold and graphitic carbon are both
"noble" (MIL-STD-889), and coupling them doesn't promote galvanic corrosion,
but any flaws or pinholes in the gold plating could allow eventual attack on the
underlying basis metal during prolonged high-humidity storage.
15. Cleanroom Gloves

Problem: Was a particular brand of antistatic disposable vinyl gloves suitable
for use in SSWs in cleanrooms? Could nonconductive latex gloves be used?
Solution: The approach was analysis and a special test. The supplier's litera-
ture claimed that a charge of 5000 V substantially drained from the antistatic
gloves in less than 2 seconds when tested by FED-STD-101, Method 4046, and
this rapid bleed-off of charge, along with a reduced propensity for triboelectric
charging, is of course the basis for using antistatic materials in ESD control. A
test with a surface resistivity meter verified the claimed antistatic rating. There-
fore, these gloves were approved.
In a special test on the latex gloves, it was found that voltage suppression limits
their apparent charge to under 300 V, which is a safe level on nonconductors
(Fig. 3 of Paper No.3 in the Appendix) when they are being worn. Therefore, these
gloves are allowed as necessary nonconductors in paragraph 3.6.10(3) of Model
Specification 1 in Chapter 9. They are used when antis tats might rub off and cause
contamination, as of silicon wafers. Precautions in using these gloves are: (1) they
must be put on and taken off 5 feet from the boundary of the SSW, (2) they must fit
tightly for voltage suppression to work, (3) they should not touch ESDS leads
because some of the charge on the latex, though truly a static charge, might
possibly be transferred, and (4) they should not be needlessly rubbed on any surface
because the less they are triboelectrically charged, the better, even though the voltage
is suppressed. Furthermore, it must be realized that charges on items being handled
cannot drain off onto the fingers when these nonconductive gloves are being worn.
16. Cloth Wrist-Strap Bands

with Metal Fibers
Problem: Could these bands shed metal fibers which might cause shorts in ESDS
items as reported for cloth with stainless-steel fibers in GIDEP Alert D5-A-84-0l
(see Example 13, above)?
Solution: The approach was a special test. Three brands of stretchable cloth
bands with stainless-steel fibers were tested by expanding the band and twisting it
to its complete capacity 100 times and then rinsing it with clean Freon and filtering
to catch any shed particles or fibers for microscopic examination. No metallic
fibers or particles above 25 microns in length were found for any of the bands. The
conclusion was that the bands were acceptable.
17. Workbench Laminates

Problem: Were certain benchtop laminates corrosive toward ESDS items, e.g.,
silicon wafers, lying on the surface?
Solution: The approach was to use special tests. First, a clean copper panel
was pressed against the surface of a sample of laminate and exposed to the el-
evated humidity and temperature conditions ofMIL-STD-202, Method 106, with-
out freeze or vibration cycles, for 1 week. The copper surface was iridescent purple
because of oxide, and small amounts of green corrosion products, possibly copper
chloride, were detected by observation at 8 x. Next, using the same brand of lami-
nate, a ring of room-temperature-vulcanizing silicone rubber served to dam a pool
of deionized water, covered to prevent evaporation, for 6 days. Titration (Ref. 8-2)
showed 0.26 mg chloride/in.2 of surface versus negligible chloride for a thermo-
plastic surface used as a control. Five other brands of laminate showed 0.00 mg
chloride/in.2, one brand showed 0.56, and another showed 1.1.
The conclusion was that five of the eight brands tested were substantially chlo-
ride-free and suitable for use in SSWs in cleanrooms where corrosion of silicon
wafers could not be tolerated.
18. AirGuns
Problem: Were two brands of nuclear ionized air guns equivalent in performance?
What was a safe distance, in terms of the field, from these guns and from two
brands of guns of the electrical type?
Solution: The approach was to use special tests. A 6 x 8 inch piece of FR-4
epoxy-glass circuit-board laminate was given an apparent charge of -5000 V by
triboelectric charging and blown with the competitive nuclear air guns at 3 feet.
After 1 second, the residual apparent charge was -4000 V in both cases. After 2
seconds, one was -1500 V, and the other was -1000 V. After 3 seconds, both were
-500 V. The conclusion was that the brands were equivalent. Note, however, that
certain polonium-210 ionizers were recalled in February 1988 because of an in-
stance of escape of microspheres in which the radioactive isotope was encapsu-
lated (Ref. 4-12). A brand of nuclear air gun using a different method of
encapsulating the polonium-21O was still allowed at that time.
The field from nuclear ionizers is very low (Ref. 4-15 indicates charging of
isolated conductors to less than 50 V), and a safe distance of 1 inch from the
nozzle was established in special tests using MOSFETs. However, electrical ion-
ized air guns caused MOSFET damage at 1 inch from the nozzle; the safe distance
was established as 2 inches for one model and 3 inches for the other (Table 5 of
Paper No.3 in the Appendix).
19. Heat Guns

Problem: A heat gun was suspected of charging electrically isolated conductors
so that ESDS items might be damaged by the DI mechanism.
Solution: The approach was to use special tests. The tip of the heat gun was
held 4.7 inches from a vertical 12 X 12 X 1/32 inch aluminum sheet on a stand-off
(plastic beaker). A vertical 6 x 8 inch section of aluminum window screen, also on
a stand-off, was placed midway between the gun and plate. The gun was set on
minimum air intake with the heater either on or off. The following data are for
duplicate tests:
VOLTAGE BY FIELD METER
HEATER SCREEN AFTER 30 SECONDS AFTER 60 SECONDS
Off Ungrounded 0,0 0,0

On Ungrounded +550, +5\0 +650,+650
On Grounded +2\0, +230 +300,+300
Similar positive charging by a heat gun is reported in Ref. 4-15.

In another test, the gate lead of a Motorola 2N4351 MOSFET was connected to
(1) a 2.5 x 4 inch copper "antenna" on FR-4 epoxy-glass circuit-board laminate, or
(2) the same with an additional "antenna" (1 ft2 aluminum sheet) attached, or (3) a
9 ft. length of 12-AWG Teflon-insulated wire with a section of the insulation re-
moved where heat-shrinkable tubing was to be installed. The substrate-case lead
of the MOSFET was grounded. Results follow:
GATE LEAD CONNECIlON OPERATION MOSFET DAMAGE
Hot air blown onto "antenna" for 1 minute 0/5

with tip of gun 5 inches from surface
2 Hot air blown onto aluminum plate in the 5/5 (OS)
same manner
3 Section of heat-shrinkable tubing installed 2/5 (OS)
on bare conductor with tip of gun 3
inches from tubing, hot air blown for
1 minute
Damage to the MOSFETs was arbitrarily defined as a shift in VOS(TH) of more

than 0.03 V. The designation "2fs (OS)," for example, means that 2 of 5 MOSFETs
tested were damaged (voltage shifts of more than 0.03 V), but none was shorted (S
=shorted). Note that the MOSFET we used is highly susceptible to ESD damage
not only because of its low 8BM threshold, about 100 V, but because it is voltage-
sensitive rather than current-sensitive. Thus, a low-capacitance charged wire or
circuit element might damage MOSFETs but be harmless to current-sensitive
devices.
The data showed that significant voltages were produced on an electrically
isolated aluminum plate when the gun blew hot air but not when it blew room-
temperature air. Positive ions in the hot air were involved in the charging pro-
cess as evidenced by about 50% less charging when a screen in the airstream
was grounded than when it was ungrounded; presumably half the ions were caught
(neutralized) by the screen and half passed through the openings. Apparently the
heating element (high-resistance wire at red heat) was creating ions. The same
positive charging by a heat gun was noted in Ref. 4-15, which states that "the
charge generation occurs with the temperature rise of the heating elements."
Aclar plastic was also charged by hot air from the gun in our tests, as was Teflon
in Ref. 4-15. The damaging effect of the charges was proved in the MOSFET
tests, the most realistic of which involved actual shrinkage of sleeving onto wire,
which is the purpose of the heat gun. Conclusions were:
1. The subject heat gun was and ESD hazard for MOSFETs sensitive to
100V.
2. Positive ions in the airstream were involved in charging the surface being
heated.
3. Worst-case laboratory tests should be run with the specific ESDS devices
installed in hardware on which the heat gun might be used. For example, the gun
might prove safe for 1000-V-sensitive devices. However, a surer solution to the
problem would be to replace the gun by a different heating tool, such as a Glo-
Ring, which depends on radiant heating rather than blowing hot air and is said to
be "ideal for many applications including heat-shrinking."
20. Triboelectric Charging by Air

Problem: Pure gases such as air cannot cause triboelectric charging (Ref. 4-14),
but might real air, with its dust content, cause charging as has been claimed (Refs.
4-4 and 8-3)?
Solution: The approach was a special test. A cellulose acetate butyrate tote
box lid, a piece of nonconductive polyethylene bubble-wrap, and a piece of FR-
4 circuit-board laminate were left in a circulating-air oven with an airflow of 26
feet/second at 200°F. The ambient relative humidity was 48% at n°F. After 2.5
days in the oven, the three plastic pieces were removed and found to have less
than 100 V apparent charge, which is harmless on nonconductors (Fig. 3 of
Paper No.3 in the Appendix). The conclusion was that blowing air at 200°F did
not cause triboelectric charging in ovens of the kinds tested. However, note
that a heat gun did produce ions, possibly emitted from the hot filament (Ex-
ample 19, above). Also circuit-board laminates were charged to +1500 V appar-
ent charge by triboelectric charging when they slid on stainless-steel oven
shelves (see the complete version of Paper No.8 condensed in the Appendix),
and some observers may have falsely attributed such charging to hot, blowing
air. Electrically unbalanced air (with a preponderance of either positive or nega-
tive ions) is another possibility.
21. Charging of Electrically Isolated

Conductors by Ionizers
Problem: What was the safe working distance for ESDS items from an AC ion-
izer or a pulsed DC ionizer? The later had been observed to cause sparks to be
thrown between electrically isolated conductors, e.g., metallic bases of micro-
scopes on rubber feet, and grounded conductors.
Solution: The approach was to use special tests. When a MOSFET with a
relatively large circuit line as an antenna was brought near either type of ionizer,
damage occurred at 2 or 4 inches but not at 10 inches, so 10 inches was judged to
be a safe working distance (Table 5 of Paper no. 3 in the Appendix). In another
test, an electrically isolated aluminum plate (12 X 12 X 0.025 inch) was allowed to
sit in the airstream for 1 minute before being touched by the gate lead of a MOSFET,
with the substrate-case lead being held by a grounded operator. This time the AC
ionizer caused damage at 8 inches e/2 MOSFETs damaged) but not at 12 inches
(Ofs damaged), so the safe working distance was revised slightly upward from 10 to
12 inches. However, the pulsed DC ionizer at 22 cycles/minute caused damage at
e/
12 inches 3 , 2 shorted), 18 inches ('/5' 1 shorted). and 24 inches ('/5' none shorted).
but not at 36 inches (Ofs). At 36 cycles/minute. the damage was 3/3 (2 shorted) at 12
inches. 2/3 (l shorted) at 18 inches, and '/2 (none shorted) at 24 inches, so increas-
ing the pulse rate from 22 to 36 cycles/minute made no noticeable improvement.
In conclusion. 1 foot was established as the safe working distance for AC ioniz-
ers (when balanced as was the one tested) and 3 feet for the pulsed DC ionizer. But
3 feet was too great a distance to be practical, so the pulsed DC ionizer was put out
of service; all units were collected and discarded.
GIDEPAlert E9-A-86-011 stated that pulsed DC ionizers did not protect ESDS
devices, but later testing is said to have shown no ion imbalance at 6 inches with
the ionizer operating at 5 Hz (cycles/second) and an imbalance of only 60 V at 2
Hz; no MOSFETs (Texas Instruments 3N163) were damaged (Ref. 8-4). How-
ever, MOSFETs were damaged at 0.4 Hz (22 cycles/minute) as stated in the GIDEP
Alert and found in our tests, above. It would seem that the frequency must be
about 5 Hz, minimum. We advise the potential buyer of pulsed DC equipment to
satisfy himself, preferably with his own tests, that isolated conductors won't be
charged to cause damage by DI.
22. Old AC Ionizers

Problem: Were old bench-model AC ionized air blowers still well enough bal-
anced electrically to effectively neutralize charges and to be safe at 12 inches in
terms of charging isolated conductors?
Solution: The approach was to use special tests with a Model 200 Charged
Plate Monitor from Ion Systems, Inc. For one ionizer, the time to discharge the
plate from -1000 V to -100 V at 2 feet was found to be 4 seconds, and from + 1000
V to + 100 V the time was 5 seconds. These data indicate that positive ions were
more numerous than negative ions in the airstream, and in fact the isolated plate
was charged to +80 V at 1 foot and +59 V at 2 feet. The conclusion was that this
ionizer, whose points were rounded and dirty, was still effective but was on the
borderline of damaging toO- V-sensitive ESDS items at 1 foot via isolated
conductors. A new ionizer of the same model charged the plate only a few
volts at 1 foot.
A second model of ionizer, which was several years old but had been given light
usage, charged the plate only about +20 V at 1 foot and was judged safe at that
distance for ESDS items sensitive to 100 V.
23. Fanless Ionizer

Problem: Was a novel, fanless, nonpulsed, two-point DC ionizer sufficiently effec-
tive to be substituted for conventional AC ionizers with fans?
Solution: The approach was a special test. Triboelectric charging of non-
conductive polyethylene bubble-wrap with polyester fabric gave -3300 V appar-
ent charge on the bubble-wrap, which was placed 2 feet in front of either the DC
ionizer or a conventional AC ionizer for either 30 or 60 seconds. At 30 seconds,
the residual charge was -3200 V with no ionizer, -2600 V with the DC ionizer, and
-1200 V with the AC ionizer. At 60 seconds, the residual charge was -3000 V with
no ionizer, -1700 V with the DC ionizer, and -700 V with the AC ionizer. The
conclusion was that the novel DC ionizer was less than half as effective as the
conventional AC ionizer and was not an acceptable substitute.
24. Grit-Blasting of Circuit-Board Laminate

Problem: A Micro Blaster was found to create +700 V apparent charge on FR-
4 epoxy-glass circuit-board laminate being sprayed with sodium bicarbonate pow-
der to remove conformal coating. How could this unacceptably high charge be
reduced?
Solution: The approach was a special test. A "piggyback" corona-discharge
ionizer reduced the apparent charge to +60 V (see the "Assembly Operations"
section of Paper No. 3 in the Appendix), which is harmless on nonconductors
because the field is far too weak to damage ESDS items by the FIM (see under
"Field Strength versus Distance" in the same paper).
25. Grit-Blasting of Heat Sink on Module

Problem: The nickel plating on the aluminum heat sink on a module was peel-
ing, so the plating was removed by grit-blasting with alumina powder. The heat
sink was grounded, and the rest of the module was protected from the grit by metal
shields taped in place. The result was ESD damage to all 83 CMOS devices on the
board, so that the $150,000 module had to be scrapped. What ESD precautions
should have been taken in the grit-blasting process?
Solution: The approach was analysis; recommendations were made on the
basis of general knowledge without running tests. The recommendations were: (1)
ground the shields and the ground planes of the board as well as the heat sinks, (2)
use an air ionizer in conjunction with the Micro Blaster (as in Example 24, above),
(3) use conductive tape, e.g., aluminum foil, to hold the shields, and (4) monitor
the apparent charge with a field meter during the grit-blasting process.
With the recommended precautions, no device was damaged. This is a dramatic
case of the need for ESD control. The loss of $150,000 (1986 money) was cer-
tainly an attention-getter for management. And note that such overt failure sug-
gests the possibility of undetected damage leading to latent failure when ESD
precautions are lax.
26. Grounding versus DI Damage

Problem: Was grounding necessary for a MOSFET on a model circuit board to
be damaged by the DI mechanism? The answer was wanted for general informa-
tion.
Solution: The approach was a special test using the MOSFET board described
under "MOSFET Damage Tests" in Paper No.3 in the Appendix. A piece of alu-
minum foil (9 x 11 inches) was charged to +4000 V from a power supply and
touched to the lead of the MOSFET board which was supported by an insulator.
The MOSFET on the board was damaged. Similar tests gave the same result. The
conclusion was that voltage-sensitive devices, which need little current flow to be
damaged, are susceptible to DI with none of the pins being grounded. Note that a
similar result was found for FFB damage (see the complete version of Paper No.8
condensed in the Appendix); here, again, no grounding was needed for MOSFET
damage, but the damaging charge came from a field rather than from DI.
27. Ultraviolet Light

Problem: Ultraviolet (UV) light is known to cause the development of a positive
charge on surfaces by the photoelectric effect. Is inspection of ESDS items under
UV light a safe process?
Solution: The approach was a special test. A MOSFET board (Paper No.3 in
the Appendix), coated with Humiseal 1B-31 acrylic conformal coating, was ex-
posed to UV light from a Blak-Ray B-100A lamp (Ultraviolet Products, Inc.) at 12
inches for 5 minutes. The MOSFET was undamaged. A total of three tests gave
the same result (% damage). The conclusion was that inspection of the coating,
which took less than 5 minutes, was not an ESD hazard (assuming that standard
precautions, e.g., grounding of the operator, were taken during this inspection).
28. Desiccator Cabinet

Problem: Could static cbarges develop on nonconductors lying inside a desic-
cator cabinet under dry, slowly flowing nitrogen? The answer was desired for gen-
eral information.
Solution: The approach was a special test. Three 6 x 10 inch specimens were
placed in the cabinet: FR-4 circuit-board laminate, polyethylene film, and polysty-
rene foam sheet. A nitrogen flow of 10 standard cubic feetlhour was maintained
for 24 hours. At the end of this time, the specimens were removed, with care taken
not to slide them, and measured with a field meter. The apparent cbarge was less
than 100 V in all cases.
The conclusion was that static cbarges did not develop on nonconductors sit-
ting in the cabinet for a day. On the other hand, nonconductors could be charged
by sliding on the grounded steel shelves of the cabinet as observed for oven shelves
and conveyor belts (see Example 20 above).
29. Chain Link Fence as Shield

Problem: A chain link fence separated SSWs from an aisle where people in or-
dinary clothing were walking. How far from the fence should the SSZs begin?
Solution: The approach was a special test. A piece of Aclar film was stroked
with polyester fabric to give an apparent cbarge of -4000 V. When the charged
Aclarwas held against the fence (wire diameter 0.114 inch, 12 openings/3 feet mea-
sured horizontally), the field meter on the other side of the fence read-170 V. Thus
the attenuation was 4000/170 =24 times. In another test, the apparent cbarge was
reduced from -10,000 V to -500 V by the fence, so the attenuation was
10,000/500 = 20 times. If the voltage was 10,oooVon passers-by, the safe distance
from the fence by the CD Rule in inches would be (500)0.5/1.8 = 12. If the voltage
was 20,000 V on passers-by, which is considered a worst-case maximum, the safe
distance in inches would be (1000)°·5/1.8 = 18. In conclusion, SSZs were main-
tained at 18 inches, minimum, from the fence.
30. Packaging of Electronic Box

with Connector
Problem: A metal-cased electronic box was a complete Faraday cage except
for a connector with deeply set pins. Could the connector be left uncapped and the
box shipped in a see-through metallized plastic bag so that the serial numbers on
the box could be read without unpacking?
Solution: The approach was a special test. In a variation of the standard dis-
charge test in paragraph 4.3.3 of Model Specification 2 in Chapter 9, a connector
pin was connected to the gate lead of a MOSFET with the substrate-case lead
being connected to the rim (backsheU) of the connector. The MOSFET and the
two wires on it were bagged in a foil laminate bag to protect them from a field
during the test, and the metallized plastic was pulled tight across the connector
face and touched with the probe at 10,000 V under the conditions of the low-
resistance discharge test (paragraph 4.3.3(2) of the specification). This was a worst-
case condition representing the finger of a charged person touching the wrapping.
Five MOSFETs in a row were undamaged (% damage), so the conclusion was
that the metallized plastic bag was satisfactory for shipping (but it might not have
been if the pins had been less deeply set in the connector housing). The particular
combination of shipped item (electronic box) and packaging was an SSP as de-
fined in Chapter 4; though a special test was done, it was construed as equivalent
to the standard discharge test (in Model Specification 2) on which it was based.
31. Shipping Test

Problem: Would MOSFETs sensitive to 100 V be damaged by in-plant shipping
in a 3-mil plain nonconductive polyethylene bag? The answer was desired for
general information.
Solution: The approach was a special test using a variation of the MOSFET
board (Paper No. 3 in the Appendix), with five MOSFETs, each with its own
separate antennas (about 1.7 in. 2 each). When a person held the bag with his fin-
gers over the antennas, walked across a carpet, and handed the package to another
person who seized it above the opposite antennas, a MOSFET was damaged by the
charge on the fingers acting through the polyethylene. In a control test, MOSFETs
were undamaged, as expected, using a foil laminate bag (a Faraday cage).
ESD TROUBLESHOOTING: ILLUSTRAnVE EXAMPLES 71
Next, the MOSFET board in the plain polyethylene bag was placed in a paper
envelope and mailed 5 times across a corporate facility, involving sorting in a
mailroom. None of the MOSFETs was damaged (°/5 damage).
The conclusion is that ESD damage is not inevitable even with very sensitive
devices in a bag with minimal shielding capability, but even the remote possibility
of damage of high-reliability parts is reason enough to use an SSP.
32. Plated Wire Memory

Shipping Container
Problem: The subject container had a heavy aluminum wall which was anodized
so that its surface was nonconductive. Thus the container was a Faraday cage,
assuming proof of electrical continuity, but was not an SSP by company defini-
tions requiring a conductive or antistatic inner surface. The container was used
for shipment. Could the plated wire memory, which was ESDS, have been dam-
aged?
Solution: The approach was analysis. Electrical continuity of the container, as
required for a Faraday cage, had been verified by probing. Since the memory was
rigidly mounted inside the container, no triboelectric charging could have oc-
curred during shipping, and the conclusion was that the package violated the re-
quirements of a company specification for an SSP but was satisfactory in practice.
In fact, the heavy aluminum wall made the container virtually impregnable to fields
and discharges. The plated wire memory was judged to have been ESD-safe.
By paragraph 3.4.2 of Model Specification 2 in Chapter 9, this container would
have been allowed as an SSP because the packaged item was securely mounted to
prevent triboelectric charging by sliding.
33. Various Packaging Configurations

Problem: Various packages proposed by suppliers were not Faraday cages. Might
some of these give sufficient protection against fields and discharges by reason of
configuration such as wide spacing between the ESDS item and the outside of the
package?
Solution: The approach was to develop a standard test, namely the discharge
test of paragraph 4.3.3 of Model Specification 2 in Chapter 9. Thus the definition
of an SSP was broadened beyond the Faraday cage defined in paragraph 3.2.2 of
Model Specification 2 in Chapter 9. Cardboard boxes lined with antistatic poly-
ethylene foam were able to pass the discharge test without a conductive layer for
shielding. The purpose of broadening the definition of an SSP was to allow sup-
pliers latitude in their packaging methods, but a true Faraday-cage package, i.e.,
one with buried foil or screen, is preferred as explained in Chapter 6.
34. Silicone-Rubber Thermal Barrier

on Probe
Problem: In the evaluation of complex hybrid circuits for dynamic parameters
under temperature variations, a probe was encased in silicone rubber as a thermal
barrier. Unexplainable malfunctions and intermittent failures were ascribed to the
field from apparent charges of commonly -5000 V and sometimes as high
as -15,000 V. How could triboelectric charging of the silicone rubber (or other
probe-encasing rubber or plastic material) be controlled?
Solution: The approach was a special test. Since no high-temperature anti·
static material with all the required properties for encasing the probe was known,
the silicone rubber remained in use, but ionization was applied; the airflow from
an AC ionizing blower 3 feet away was directed on the probe. Now the apparent
charge was limited to at most -300 V, which is harmless on nonconductors (see
Fig. 3 of Paper No.3 in the Appendix).
35. Automated Process
Problem: How could the ESD safety of a complex automated process, with vari-
ous hazards involving several ESD damage mechanisms (DI, FIM, CDM, FFB),
be assured with high confidence?
Solution: The approach was a combination of analysis and special tests. Pa-
per No. 8 in the Appendix outlines how the problem was solved by the coupon
approach, which is uniquely applicable to automation and gives very high confi-
dence in ESD safety because of the absence of unpredictable operator error. Mis-
handling cannot occur unless the process goes awry, as would be detected in periodic
coupon checks.
Note that the coupon procedure has been made part of Model Specification 1
(paragraph 3.9) in Chapter 9. This is another example of the way in which specifi-
cation grow by absorbing the solutions found by troubleshooting. In terms of this
book, an expanding Chapter 8 continues to feed into Chapter 9.
36. Cordless Wrist Strap

Problem: Could this novel strap be approved for use?
Solution: In two special tests (paper No. 15), the strap was quite ineffective. How-
ever, the same conductive fibers incorporated into the strap, apparently to provide
corona discharge, were effective in the form of a carpet as seen in the next example.
37. Conductive Carpets
Problem: Could conductive carpets reduce charging of people wearing ordinary

shoes?
Solution: Two conductive-fiber carpets were evaluated by the walk test (Paper
No. 15). Carpet 2 in Table 1 of this paper utilized the same fibers as the cordless
wrist strap in the above example. Conductive fibers may function by drawing the
charge from each tiny nonconductive shoe sole area contacted by a fiber at the
moment of separation of sole from carpet. Though not as effective as static-limit-
ing floor finish, conductive carpets may make the difference between, say, 8 kV
on a walking person who touches a conductor and throws a spark and 2 k V so that
he throws no spark, thus preventing the upset (Ref. 8-5) of equipment such as
computers in carpeted rooms.
38. Ionization of Nitrogen

Problem: Can nitrogen be ionized for blowing over charged nonconductors to
remove static charges?
Solution: In a special test simulating neutralization of charges on circuit boards,
a slab of polystyrene foam was charged to approximately 20 kV by rubbing with
fabric, and either air or nitrogen was passed through a polonium-21O cartridge and
then through rubber tubing of various lengths before striking the foam surface 0.3
inch from the end of the tubing. After the tubing tip was played back and forth for
30 seconds across the foam surface (3 x 6 inches), the surface was checked with a
static meter for essentially complete neutralization (less than 1 kV residual appar-
ent charge). Nitrogen was ionized less effectively than air, as seen in the follow-
ing results.
LENGTH OF ruBING, INCHES
Gas o 2 6 12
Air Yes Yes Yes No
Nitrogen Yes Yes (barely) No No
(Yes =charge neutralized. No =charge not neutralized.)
In further testing with nitrogen, a safety nozzle with a hand lever gave too long
a path from the cartridge for effective neutralization. The conclusion was that a
half-inch length of rubber tubing, with tip serrated to brush gently against the
work, should be attached directly to the cartridge for nitrogen use.
39. Antistatic and Conductive Tweezers

Problem: Were these tweezers suitable for handling ESDS items?
Solution: The surface resistivity of antistatic (vulcanized fiber) tweezers
or conductive (carbon-filled plastic) tweezers was measured by the method of
Paper No. 17 with the tips of the tweezers held in the clamps of the apparatus. The
antistatic tweezers gave the curve of surface resistivity versus relative humidity
labeled "F" in Fig. 5 of Paper No. 15, indicating CDM-safety (see Paper No. 16) at
74 ESD FROM A TO Z: ELECTROSTAnc DtSCHARGE CONTROl FOR ELECTRONtcS
normal relative humidities. In contrast. the conductive tweezers had very low sur-
face resistivUy (5-6 ohms/square) at 0-93% relative humidity and were definitely
a CDM hazard. 1be antistatic tweezers were chosen for handling ESDS dice.
40. Antistatic Tapes

Problem : Were antistatic tapes effective at low relative humidity?
Solution: Three brands of cellophane antistatic tape gave curves of surface
resistivity versus relative humidity at room temperature very much like the curve
for plasticized and coated cellophane in Fig. 7of Paper No. 17. All the tapes were
antistatic down to very low relative humidity (3-5%) and were judged to be ef-
fective under nonnal working conditions.
41. Conductive Footwear

Problem: How effective was novel conductive footwear versus convenrional
heel-grounders?
Solution: Experimental conductive footwear resembled conductive-soled slip-
pers that fitted over ordinary shoes. A lead with a I-megohm resistor connected
the carbon-loaded rubber sole to a conductive pad bonded to the skin of the wearer's
ankle. In the walk test, the new footwear gave no significant voltage, nor did heel-
grounders (conductive straps under meheel) when subjects walked nonnally. How-
ever, when subjects walked on tiptoe, the heel-grounders allowed voltages of up to
200 V on conductive vinyl tile.
A more significant special test was a resistance-ta-ground measurement in which
a subject held a wand connected to the low (negative) lead of a Beckman L-IO
Fig. 8-1. TIle layout.

ESD TROUBLESHOOTING: ILLUSTRAnVE EXAMPLES 75
megohmmeter while the high (positive) lead was connected to an NFPA 99 (or
ASTM F 150) electrode on the tile layout (l ft 2 tiles) shown in Fig. 8-1. The tiles
were electrically connected underneath. Results, averaged for 10 subjects, follow.
NEW FOOTWEAR RESISTANCE AS PERCENT OF HEEL-GROUNDER RESISTANCE
Tile Surface Clean Surface Sprinkled with Cement Dust

Conductively painted 40 25
Conductive vinyl 60 41
The new footwear was more effective than heel-grounders on a soiled surface,
presumably because of the greater contact area (the entire sole) of the new foot-
wear. In conclusion, the new footwear was judged to provide superior grounding.
It also was easy to put on and take off and reasonably comfortable to wear.
42. Space-Charging by Air Ionizer

Problem: Would a ceiling-mounted air ionizer produce an imbalance of positive
and negative air ions so that an electrically isolated conductor at bench level might
become charged (space-charged)?
Solution: Details of a special test are given in Paper No. 15. The conclusion
was that the ionizer, when properly operated, would not space-charge an electri-
cally isolated 20-pF conductor to make it a threat to ESDS items by the DI dam-
age mechanism.
43. Antistatic Nylon

Problem: Is a truly effective antistatic nylon available?
Solution: Antistatic nylon film of the past was ineffective when the relative
humidity fell below 50% (Paper No. 17). By the method of Paper No. 17, a nylon
film with an improved but still fugitive additive was antistatic down to 0% rela-
tive humidity. A permanently antistatic nylon was also effective down to 0% rela-
tive humidity but was suitable only for molded parts, e.g., probe tools or tweezers
for handling die, and could not be made into film. Curve "N" in Fig. 5 of Paper No.
15 is for this permanently antistatic nylon.
The conclusion was that truly antistatic nylon products were commercially avail-
able, but only the molded parts, not the film, were permanently antistatic with no
fugitive antistat.
44. Pennanently Antistatic Vinyl

Problem: Was a new vinyl plastic truly permanently antistatic?
Solution: This material (curve "V" in Fig. 5 of Paper No. 15) was found to be
volume-conductive and truly permanent because it contains no fugitive (volatile
or extractable) antistat. A second polymeric phase dispersed through the vinyl ma-
trix provides weak conductivity through the bulk of the material as well as on the
surface and cannot be extracted by common solvents. Also, in a special test, this
material did not make a nonconductive surface undesirably conductive by con-
tact (Paper No. 20) so that electrical leakage could occur; this is important for
packaged items such as resistors.
The same technology with polymeric antistatic additives has been applied to
other plastics, e.g., nylon (Example 43, above), ABS, and polypropylene, and is a
breakthrough especially applicable to plastic packaging (Paper No. 20).
Chapter 9
Model Specifications
Introduction
These model specifications are not claimed to be "perfect" or all-inclusive. Their
purpose is to illustrate and codify the elements of our approach, e.g., the SSZ
concept and the CD Rule, in specification fonnat. Here the threads of our preced-
ing A-to-Z discussions are pulled together in a system of required procedures.
Note how the results of special studies (troubleshooting) have been incorporated
as pointed out in the last chapter.
Specifications may be written as a list of general requirements without "how-to"
instructions or detailed examples, but the problem with this method is that the
reader is puzzled by sterile, though correct, generalizations. Alternatively, specific
materials, procedures, etc., can be given, but the problem here is lack of flexibility
for meeting general requirements in new, creative ways. Our specifications are
compromises; we give a general requirement and then illustrate it with specifics.
Thus the intent is stated along with methods-but not the only methods-to achieve
it.
Model Specification 1 concerns the SSW and gives rules for maintaining an
SSZ within it using the various "tools" of Chapter 4. Similarly, Model Specifica-
tion 2 utilizes these tools for the SSP. You should be able to adapt these specifica-
tions to your own use by appropriately modifying and expanding them. Use them
freely; see the copyright page.
Model Specification 1. Handling

and Assembly of Unpackaged
ESDS Items
1. Scope
This specification established requirements for in-plant handling aspects of an ESD-
control program as required by MIL-STD-1686 and contains detailed requirements
n
which are called for but not given in that document. The purpose of the ESD-control
program is to protect electrical and electronic items (including components, as-
semblies, and units as defined in 5.1) that are susceptible to damage or degradation
by ESD. The ESD-control program does not apply to explosives or other hazard-
ous materials.
2. Applicable Documents
The following documents, of the latest issue in effect except as otherwise indi-
cated, form a part of this specification to the extent specified herein. In the event of
conflict between documents referenced herein and the contents of this specifica-
tion, the contents of this specification shall govern.
2.1. Government documents.
SPECIFICATIONS
Military
MIL-M-38510 Microcircuits, General Specification
for
STANDARDS
Military
MIL-STD-120 Marking for Shipment and Storage
MIL-STD-1686 Electrostatic Discharge Control
Program for Protection of
Electrical and Electronic Parts,
Assemblies, and Equipment
(Excluding Electrically Initiated
Explosive Devices)
Department of Defense Soldering Technology, High Quality/
DoD-STD-2000-lB* High Reliability
OTHER PUBLICATIONS
MIL-HDBK-263 Electrostatic Discharge Control
Handbook for Protection of
Explosive Devices)
NAVSEA SE 003-AA-TRN-OlO Electrostatic Discharge
Training Manual
*Note: Superseded by MIL-STD-2000 on January 16. 1989. We reference DOD-STD-2000 because

many older Government programs are on contract for it.
MODEL SPECIFICAnoNS 79
2.2. Nongovernment documents.
SPECIFICATIONS
Model Specification 2 Packaging of ESDS Items for
Shjpment, Storage, and
In-Plant Transfer
OTHER PUBLICATIONS·
Reference Document 1 General Design, Construction,
and Maintenance Guidelines
for ESO-Protected Areas
Reference Document 2 Certification of Static-Safe
Workstations (SSWs)
Reference Oocument 3 Procurement Requirements for
Continuous Wrist-Strap
Monitor Unit
Reference Document 4 Approved ESO-Control Materials
and Equipment
Reference Document 5 Criteria for Selection of Approved
ESO-Control Materials and
Equipment
Reference Document 6 Certification of Personnel for
ESOControl
3. Requirements
3.1. Order of precedence. In the event of conflict between the requirements

of this specification, a program contract, and drawings calling out this specifica-
tion, the requirements of the contract and the drawings shall take precedence in
that order.
3.2. Classification of ESDS items. ESDS items shall be classified in accor-

dance with MIL-STO-1686. Items in Classes 1-3 shall be considered ESDS for
the purposes of this specification, and all shall be protected in full accordance with
this specification.
3.3. Design protection. The need for design protection shall be determined in
accordance with MIL-HOBK-263 and NAVSEA SE 003-AA-TRN-OIO.
3.3.1. Protection of components and assemblies. Class 1 components shall

not be used where Class 2 or less sensitive components are available that will meet
"'Reference Documents 1-6 are made independent of the specification so that they can be revised
separately. This is especially important for Reference Document 4, which is a "living document" that is
always growing and changing.
perfonnance requirements. Where Class 1 components must be used, protective

circuitry shall be incorporated at the lowest practical level of assembly to limit the
sensitivity of that assembly to Class 2 or greater. Where protective circuitry cannot
be used, reasons shall be presented to the acquiring activity.
3.3.2. Protection of external unit enclosure terminals. Unit tenninals, includ-
ing test points, located on the surface of the unit enclosure shall contain protective
circuitry making the unit sensitive to no less than 4000 V (Class 3). When protec-
tive circuitry cannot be used to provide this protection, reasons shall be presented
to the acquiring activity.
3.4. Duration of protection. ESDS items shall be protected in accordance

with 3.5 from the time of receipt through final unit assembly, or until processing
eliminates the possibility of ESD damage as defined by the drawing. Items whose
electrical parameters are irrelevant so that ESD damage is of no concern shall be
exempt from this requirement and from all requirements of this specification; ex-
amples are display models and scrap items used for physical as opposed to electri-
cal testing.
3.5. Static-safe zone (SSZ). While being protected (3.4), ESDS items shall
never leave an SSZ, which is defined as a volume in space at every point of which
the Basic Rule (3.5.1) is followed.
3.5.1. Basic Rule. An ESDS item shall never, even for a nanosecond, be ex-
posed to an E field in violation of the CD Rule (3.6.11) or have its sensitive
leads or terminals touched to, or receive a discharge from, any surface at
more than 50 V.
3.5.2. Location ofSSZ. The SSZ may be (1) inside a static-safe package (SSP)
as defined by Model Specification 2, (2) within a static-safe workstation (SSW) as
defined in 3.6.1, or (3) around the ESDS item while this item is being relocated
from one SSW to another under the continuous supervision of ESD-certified per-
sonnel (3.6.13). An ESDS item shall be removed from an SSP only within the
SSZ of an SSW.
3.6. Static-safe workstation (SSW).

3.6.1. Definition of ssw. An SSW is a volume in space which is designated
and equipped to maintain an SSZ (3.5).
3.6.2. General design. construction. and maintenance. General design, con-
struction, and maintenance of SSWs and ESD-protected areas in general shall be
in accordance with Reference Document 1.
3.6.3. Identification and numbering of ssw s. All SSWs shall be identified by
signs and markings, suitably strips of yellow tape at the boundaries, and shall be
numbered for the purpose of keeping records, e.g., of certification (3.6.4).
3.6.4. Certification. SSWs shall be certified annually in accordance with Ref-
erence Document 2.
3.6.5. Grounding
MODEL SPECIFICATIONS 81
3.6.5.1 Grounding of conductors. All conductors shall be grounded so that

their potential versus earth ground never exceeds 50 V as required by the Basic
Rule (3.5.1).
3.6.5.2. Grounding of work surfaces and personnel using static ground. A
static ground separate from commercial power ground shall be required for rea-
sons of safety. For example, the commercial power ground might be floating (dis-
connected) and powered by a hot lead so that operators are endangered.
3.6.5.2.1. Static-ground terminations. Static-ground terminations shall be one
of the following:
a. Building structural steel.

b. Clamp to a metallic cold water pipe.
c. Wet ground-well with copper-clad steel stake.
(Test and process equipment chassis shall be connected to commercial power

ground.)
3.6.5.2.2. Voltage and resistance differential from commercial power ground.
Static ground, measured at the distribution wire (3.6.5.2.3 and Fig. I), shall not
differ from commercial power ground by more than 5 V (dc or rms) or by more
than 25 ohms.
3.6.5.2.3. Static-ground wire system. The grounding lug of the work surface
shall be connected by a lead wire with a resistor (3.6.2.4) to a distribution wire
terminating in ground (3.6.5.2.1). The lead wire shall be a minimum of 18 AWG,
and the distribution wire shall be a minimum of 12 AWG. (AWG = American Wire
Gauge.) Figure 1 shows a typical grounding system.
3.6.5.2.4. Resistor. There shall be no resistor between the termination and the
distribution wire, but there shall be a I-megohm resistor between the distribution
wire and the grounding lug of the work surface.
3.6.5.3. Grounding of personnel.
3.6.5.3.1. Personnel protection. Personnel shall be grounded by static ground
(3.6.5.2) and shall always be protected by sufficient resistance to limit current to a
maximum of 1.0 mA when they are electrically connected to ground, e.g., by a
wrist strap. (Current in mA = 1000 x voltage/resistance in ohms.)
3.6.5.3.2 Allowable resistance to ground (ARTG). The ARTG, as measured
at 0.1-40 V DC between static ground and a metallic probe 1-3 inches in diameter
held firmly in the operator's hand, shall be a maximum of 10 megohms and a
minimum of 1 megohm.
3.6.5.3.3. Wrist-strap monitor. Every wrist strap shall be monitored by a con-
tinuous wrist-strap monitor of the resistive type procured in accordance with
Reference Document 3. One ground connection of the monitor shall be connected
to static ground, suitably by means of the grounding lug of the work surface, and
the other ground connection of the monitor shall be connected to commercial power
ground. See Fig. 1. (Note: Some users have found continuous-monitor features
such as the audible alarm to be annoying. As we say in Paper No.1 0 in the Appendix,
82 ESD FROM A 10 Z: ELECTROSTATIC DISCHARGE CONTROL FOR ELECTRONICS
Distribution wire
Operator Static ground
Other electrical
equipment
(,.,.",., • t-megohm r"stor)
(Floor treated with ltadc-Umltllla floor fiDbb.)
Fig. 9-1. Diagram of grounding of a typical SSW with workbench.
don't buy these monitors, or any other major ESD-control equipment, without a
trial in your own assembly area.)
3.6.5.3.4. Ground/ault circuit interruptors (GFCls). GFCIs shall be installed
when voltages above 225 V are being used, so that personnel will be protected
even if electrically overstressed resistors should be carbonized so that they allow a
current of more than I mA.
3.6.5.3.5. Conductive floor mats. Conductive floor mats or other methods of
grounding personnel other than monitored wrist straps (3.6.5.3.3) shall be used
only when necessary. An example is the grounding of mobile personnel when
wrist straps even with long cords or connected to a sliding trolley overhead are
impractical. Floor mats shall be connected to the static-ground distribution wire
(3.6.5.2.3) through a I-megohm resistor, or they may be connected directly to the
grounding lug of the work surface. Conductive shoes or shoe straps with a mini-
mum resistance of I megohm shall be used in conjunction with conductive floor
mats.
3.6.6. Materials and equipment.
3.6.6.1. Materials. ESD-protective materials used in accordance with this speci-
fication shall have a surface resistivity of less than 1012 ohms/square when mea-
sured at 70°F and a maximum of 50% relative humidity with a Voyager SRM-II 0
Surface Resistivity Meter or equivalent. Materials shall be selected from Reference

Document 4, which is a listing prepared using the criteria of Reference Document
5. The materials are classified as follows:
Conductive (surface resistivity less than lOS ohms/square).
Static-dissipative (surface resistivity at least lOS but less than 109 ohms/square).
Antistatic (surface resistivity at least 109 but less than 1012 ohms/square). Un-
glazed paper is antistatic at normal relative humidity (above about 25% at 70°F)
and is acceptable for work instructions and other documents in the SSW. Unglazed
cardboard is also antistatic and may be used for tote boxes, etc. (Note: EIA 541
and MIL-HDBK-263 combine antistatic and static-dissipative as "dissipative,"
which is sensible as we mention elsewhere, but we follow DoD-HDBK-263 prac-
tice here.)
3.6.6.2. Equipment. Equipment used in accordance with this specification shall
be selected from Reference Document 4, which is a listing prepared using the
criteria of Reference Document 5.
3.6.7. Relative humidity. If the relative humidity is controlled in order to re-
duce triboelectric charging, the absolute humidity shall be equivalent to a rela-
tive humidity of 30%, minimum, at 70°F. (Note: 40% would be even better. The
maximum must be set with consideration of possible adverse effects such as corro-
sion, delamination of circuit-board material, or personnel discomfort.)
3.6.8. Ionization. Air ionizers shall be used only when absolutely necessary,
e.g., in grit-blasting or spray-coating operations. Equipment shall be selected from
Reference Document 4. Ionized air blowers shall be electrical, but nozzles (guns)
may be other electrical or nuclear. (Note: This is our preference based on cost of
use, because the nuclear-type blowers are leased and have an annual cost that ac-
cumulates excessively, in our opinion.) The blowers shall have autobalancing, point-
discharge indication, and need-for-cleaning indication features. All ionization
equipment shall be tested and certified annually. ESDS items shall never be nearer
than 12 inches to the points of a bench-model ionized air blower or 3 inches from
a nozzle; this prohibition protects the items from slight electrical imbalance of the
ionizers. (Note: Even nuclear (polonium 21O) blowers have been accused of creat-
ing space charges and charging isolated conductors to a little under 50 V; see Ref.
4-15.}
3.6.9. Static-limiting floor finish. An approved static-limiting floor finish
(Reference Document 4) shall be applied to all tile and concrete floors (over a
sealer for the concrete) and maintained so that appearance is attractive (gloss = 25,
minimum; see Paper No.6 in the Appendix), slip resistance is adequate (0.20 Topaka,
minimum; see Paper No.6 in the Appendix), and the surface resistivity never
exceeds 1011 ohms/square as measured by a Voyager SRM-IlO Surface Resistiv-
ity Meter or equivalent. Carpeted floors shall be used only when absolutely neces-
sary, and the carpet shall be treated with an approved antistat; the limitation of
charges on people shall be measured by the walk test of Paper No.6 in the Appen-
dix or by measurements with a Voyager Personnel Voltage Tester PVT-300 or
equivalent.
3.6.10. Necessary nonconductors. Nonconductors in the SSW shall be lim-

ited to those which are not easily or cost-effectively replaced. Examples are:
1. Painted, grounded metals

2. Plastic-coated, grounded metals with plastic less than 20 mils thick and
metal more than 60 mils thick
3. Latex gloves (see 3.6.12.11)
4. Floors
5. Chairs
6. Walls
7. Structural members within the workstation
8. Personal apparel including I.D. badges
9. Insulated (non conductive) tool handles
10. Dispensing syringes and bottles
As necessary, these shall be treated with topical antistat to make them anti-
static so that their field will not violate the CD Rule (3.6.11). If corrosion is feared,
a chloride-free topical antistat may be selected. In the above list, Nos. 1 and 2
will not need antistat treatment because of voltage suppression, No. 3 will be
used without antistat treatment but subject to the conditions of 3.6.12.11, Nos. 4-8
probably will not need antistat treatment (for badges, see Examples 4 in Chapter
8), and Nos. 9 and 10 shall be treated with topical antistat and so labeled. The
treatment shall be renewed every three months (Note: or more often, as indicated
by experience), and its effectiveness shall be monitored by field-meter checks.
3.6.11. CD Rule. The CD Rule shall never be violated, the requirement being
that ESDS items must be kept at least a safe distance d from a surface with appar-
ent charge V in volts according to the equation: d =yo.s/l.8. The apparent charge
is the charge in volts of a surface or object as read by a hand-held field meter
according to the meter manufacturer's instructions; this is an uncorrected reading
and represents the field strength at the point in space where the meter is held.
(Note: See the discussion of the CD Rule in Chapter 4.) For convenience, some
solutions of the equation are:
V, VOLTS d, INCHES V, VOLTS d, INCHES
300 to 6,000 43
500 12 8,000 50
1,000 18 to,OOO 56
2,000 25 15,000 68
3,000 30 20,000 79
4,000 35 30,000 96
3.6.12. Operator disciplines.

3.6.12.1. Maintenance oJSSZ. The operator shall maintain an SSZ and shall
always know its boundaries so that ESDS items are kept within it. In general, an
SSZ will end at an operator's wrists, at least 12 inches from the points of an ion-
ized air blower or 3 inches from a nozzle (3.6.8), and at distance d from any charged
surface according to the CD Rule (3.6.11).
3.6.12.2. Parts handling. The operator shall not unnecessarily touch ESDS
leads, terminals, or contacts even though he or she is grounded. Parts shall be
handled by their cases as size permits. An ESDS item shall never be held beyond
the front edge of the work surface because of possible fields from the chair or
operator's lower garments.
3.6.12.3. Touching 0/ surfaces. The operator shall not allow ESDS items to
touch surfaces or objects other than the approved work surface, approved materi-
als and equipment, or the assembly to which the item is being attached. For ex-
ample, clothing shall not be touched.
3.6.12.4. Triboelectric charging by body movement. The operator shall not
fidget, squirm, or shuffle his or her feet unnecessarily. The object of this discipline
is to minimize triboelectric charging of the operator's body or apparel.
3.6.12.5. Voltage surges by capacitance drop. The operator shall not touch
ESDS leads, terminals, or contacts while arising from a chair or lifting his or her
feet. This prohibition protects ESDS items from surges in V when C drops in the
equation Q = CV.
3.6.12.6. Hair. Hair shall be tied back as necessary so that it is within 2 inches
from the skin in order to limit the apparent charge by voltage suppression. (Note:
See Example 2 in Chapter 8. This requirement on hair is an example of the incor-
poration of the results of troubleshooting studies into the specification.)
3.6.12.7. Shorting clips and protective containers. The operator shall not re-
move shorting clips or protective containers until ready to use an ESDS device.
3.6.12.8. Exclusion o/nonconductors. Static-generating nonconductors, e.g.,
ordinary, untreated tote boxes, plastic envelopes or notebooks (for work instruc-
tions, etc.), glazed paper, or ordinary plastic packaging material, shall be excluded
from the SSW. Only necessary nonconductors (3.6.10) shall be allowed.
3.6.12.9. Cleanliness. A clean and orderly SSW shall be maintained. Lunches,
coffee cups, candy wrappers, etc., shall be excluded; many of these materials are
unnecessary nonconductors.
3.6.12.10. Wearing apparel. Operators shall wear (1) close-fitting blouses or
shirts with short sleeves (above the elbow), (2) close-fitting blouses or shirts with
sleeves rolled up above the elbow, or (3) ESD-protective smocks. (Note: In assem-
bly areas, these smocks are an attractive "uniform" that raises ESD-awareness and
may impress visitors, but periodic cleaning and replacement of the smocks may
prove relatively expensive.)
3.6.12.11. Gloves and finger cots. Cotton or other antistatic gloves and finger
cots shall be worn, with the exception that nonconductive gloves or cots, e.g.,
latex gloves for cleanroom work where antistatic additives might rub off to cause
contamination, may be used when they are tightly fitting and are judged to be
necessary nonconductors (3.6.10(3». Of course, nonconductive gloves shall not
be put on or removed at the SSW in such a way that their field, uncontrolled by
voltage suppression as it is when they are being worn, violates the CD Rule
(3.6.11). In general, nonconductive latex gloves shall be put on or removed at
least 5 feet from the boundary of the SSW. (Note: For precautions on using these
gloves, see Example 15 in Chapter 8.)
3.6.12.12. Working near chain link/ences. The SSZ (3.5) shall be at least 18
inches from a chain link fence bounding the SSW when people in ordinary clothes
may pass near the other side of the fence. (Note: This paragraph would not be
included in most specifications. It's an example of a special provision based on an
unusual situation, namely, the presence of fences in assembly areas; see Example
29 in Chapter 8 and the real need for a fence provision in Chapter 11.)
3.6.12.13. Transient personnel. Operators shall ensure that transient person-
nel, e.g., supervisors or visitors, do not enter SSWs unless such personnel are
grounded and closely supervised by the operator.
3.6.13. Certification and training 0/personnel. Operators and other personnel
who specify, acquire, design, assemble, inspect, test, package, repair, rework, in-
stall, or maintain ESDS items shall be certified in accordance with Reference
Document 6 by completing a training course in ESD control (at least 8 hours in-
struction and one test), then recertified annually (at least 2 hours brush-up instruc-
tion and one test). Records of certification and recertification shall be maintained.
(Note: Operator training should correspond to the specification; for example, if
this model specification were used, our book with appropriate explanations and
enlargements would serve as the text or basis of course material. The need for
consistency of specification and training material seems obvious, but in practice
there may be contradictions because the ESD-control program was developed piece-
meal and the training course and the evolving specification went in different direc-
tions. Another point we wish to emphasize is that operators should not be kept
ignorant by a management jealously guarding facts because "knowledge is power."
Rather, in regard to ESD control, the operator should be a "senior employee" able
to act independently and intelligently like an airplane pilot or police officer.)
3.6.14. Testing.
3.6.14.1 Dielectric strength or insulation resistance tests. Equipment contain-
ing ESDS items shall not be subjected to dielectric strength or insulation resis-
tance tests.
3.6.14.2. Applied voltage. ESDS items shall not be subjected to applied volt-
age, e.g., from an ohmmeter, at an energy level that could create a damaging
current.
3.6.14.3. Removal from test sockets. ESDS items shall not be removed from
test sockets with power applied.
3.6.14.4. Power to connectors. All power shall be turned off to the connectors
in a system before printed circuit boards or other assemblies containing ESDS
items are inserted into or removed from the connectors.
3.6.14.5. Proper polarity o/voltage. Test equipment setups shall be arranged
for proper polarity of voltage before conducting parametric or functional testing.
3.6.14.6. Supply voltages and signal voltages. Direct-current (DC) supply
voltages shall always be applied before signal voltages except for memory subas-
semblies. Signal voltages shall always be removed before DC supply voltages
unless the function removes signal voltages simultaneously with DC supply
voltages.
3.6.15. Installation site. The following minimum installation procedures shall
be enforced:
a. Protective coverings shall be maintained intact when ESDS units are to be stored
prior to installation.
b. Protective caps, shunts, etc., on unit receptacles shall not be removed during
unit installation until the completion of required interconnecting cable and
connector assemblies.
c. Connector and receptacle surfaces shall be momentarily contacted to ground in
order to discharge any electrostatic potentials on the cable or installer's body
just prior to engaging a cable connector with a mating receptacle and prior to
removing protective covering.
3.6.16. Caution marking.

3.6.16.1. Labeling. ESDS items or their containers shall be identified with an
ESD caution label for all in-plant operations. The label may be on ordinary paper
or latex-impregnated paper for cleanroom use and shall be backed with pressure-
sensitive adhesive. The label shall include the MIL-STD-129 symbol or the Elec-
tronics Industries Association (EIA) triangle (see note in Model Specification 2,
para. 3.14.3.1) and shall be readily visible to personnel when marked assemblies
are mounted on the next higher assembly.
3.6.16.2. Caution tag. When labels are not practical, an ESD caution tag shall
be used. For cleanroom use, the tag shall be sealed in an envelope of transparent
antistatic polyethylene or, better, nylon. Specific instructions may be recorded on
the reverse side of the tag. A sample tag design follows.
CAUTION
OBSERVE PRECAUTIONS
FOR HANDLING
ELECTROSTATIC
SENSITIVE
DEVICES
3.7. Audits and reviews. Audits and reviews of the ESD-control program shall
be conducted in accordance with contractual requirements. (Note: A thorough au-
dit can be a disturbing revelation. The ARTG may be found to be exceeded in a
majority of cases if continuous wrist-strap monitors, as required in this specifi-
cation, are not used, and many SSWs may be found to have faulty grounding. An
adverse audit could result in much assembled hardware being considered suspect
and perhaps rejected.)
3.8. Packaging for delivery. Packaging for delivery shall be in accordance

with Model Specification 2.
3.9. Automated processes. Automated processes shall be certified ESD-safe,

and periodically recertified, by means of a coupon. (Note: The particular coupon
design and its use must be detailed in your own specification. See Paper No.8 in
the Appendix for a full discussion of the coupon technique.)
3.10. Disposition of mishandled hardware. As soon as possible after the in-

fraction, the ten data items listed in Chapter 13 shall be recorded. Disposition of the
hardware then shall be decided by Quality Assurance in conjunction with Engineering.
4. Quality Assurance Provisions

4.1. Responsibility for quality assurance. The Quality Assurance organiza-
tion shall be responsible for compliance with the requirements specified.
4.2. Certification. SSWs, ionization equipment, and personnel shall be certi-

fied in accordance with 3.6.4, 3.6.8, and 3.6.13, respectively.
4.3. General inspection of SSWs. SSWs shall be inspected monthly, as a

minimum, for compliance with 3.6.3, 3.6.4, 3.6.5, 3.6.7, 3.6.8, 3.6.9, 3.6.10, and
3.6.11.
4.4. Inspection of materials and equipment at SSWs. SSWs shall be in-

spected weekly for compliance with 3.6.6.
4.5. Inspection of operator disciplines. Inspection for compliance with 3.6.12

shall be conducted daily at 10%, minimum, of the SSWs, randomly selected.
4.6. Inspection of testing procedures. Inspection for compliance with 3.6.14

shall be conducted weekly.
4.7. Inspection of installation site practices. Inspection for compliance with

3.6.15 shall be conducted as often as practicable, preferably daily.
4.8. Inspection of packaging for delivery. Inspection for compliance with

3.8 shall be weekly.
4.9. Recertification of automated processes. Recertification of automated

processes (3.9) shall be done monthly. (Note: Frequency of inspection depends on
circumstances and is debatable. The above frequencies seem reasonable to us but

are meant to be illustrative only.)
5. Notes
5.1 Definitions. Definitions of terms used herein are as follows:
Components: Electrical and electronic components.
Assemblies: Electrical and electronic subassemblies and assemblies up to but

not including the unit level.
Units: Enclosed electrical and electronic higher assemblies.
Items: Components, assemblies, and units referred to collectively.
5.2. Precautions pertaining to ionized air blowers.

5.2.1. Ozone generation. All corona-discharge-type ionized air blowers gen-
erate ozone to some extent. The OSHA standard is 0.1 ppm for an 8-hour expo-
sure. It is reported that 0.015 ppm ozone is "barely detectable by some sensitive
people" while 1.0 ppm gives a "very noxious odor." Therefore, the rule ofthumb is
that if ozone can be smelled easily, the concentration is too high for prolonged
exposure.
5.2.2. Shadowing effect. Charged surfaces protected from the airstream by ob-
structions (the "shadowing effect") will not be neutralized effectively.
5.3. Precautions pertaining to antistatic polyethylene. This material is used

for bags (typically 6-mil film), parts carriers, tote boxes, etc. It achieves its anti·
static property by a weakly conductive "sweat layer" which is formed when an
organic liquid antistat, added before extrusion, migrates or blooms to the surface
and combines with water vapor from the ambient air (but some antistats can func-
tion in dry air as seen in Paper No. 17 in the Appendix). This invisible layer allows
incipient static charges to bleed off before they can cause ESD damage. The anti-
stat can cause contamination and, in the case of one product, corrosion occurred.
(Note: See Paper No.9 in the Appendix.) The following considerations must be
kept in mind when using antistatic polyethylene.
5.3.t. Fading of pink color. The identifying pink color will gradually fade to
faint amber or off-white under fluorescent light or sunlight due to the ultraviolet
portion of the spectrum. This fading does not affect the antistatic behavior; only
the color changes.
5.3.2. Cleaning. The material should not be exposed to solvents. Water, alcohols,
hydrocarbons, Freons, or other solvents used in electronics manufacturing all re-
move the "sweat layer" and destroy the antistatic property, though new antistat
will bloom to the surface and regenerate the sweat layer a few times until the
internal reservoir of antistat is exhausted. The material must be cleaned only with
an antis tat solution, without drying; this procedure leaves an antistatic film on the
surface. Other products such as permanently antistatic or static-dissipative film
with a radiation-cured coating may also be solvent-sensitive and may lose their
effectiveness after a single brief exposure.
(Note: Paragraphs 5.2 and 5.3 illustrate the kind of tutorial information that may
be included in Section 5. Even though personnel have been taught these facts,
caution notes are desirable to heighten the awareness of hazards that may damage
hardware and, even more importantly, may endanger health or cause personal
injury.)
Reference Document 1. General Design,

Construction, and Maintenance Guidelines
for ESD-Protected Areas
1. Introduction. This document gives general guidelines for SSWs in com-
pliance with Model Specification 1. The term "ESD-protected area" means a loca-
tion, e.g., a room or portion of a factory floor, containing a number of SSWs
separated by aisles or other non-SSW zones.
2. SSW designation. The SSW shall be a clearly defined volume of space

with caution signs, station identification numbering, and, when possible, borders
marked with yellow tape. The "buffer zone" within the SSW but beyond the bound-
aries of the expected SSZ shall be as large as can be afforded in terms of space
usage.
3. Housekeeping. The SSW shall be kept clean, orderly, and free of unneces-
sary nonconductors and unapproved materials and equipment.
4. Work surface. The SSW work surface shall be of an approved static-dissi-

pative or antistatic material. Grounding of the work surface shall be certified
annually in accordance with Reference Document 2, and the SSW shall have a
current certification sticker.
5. Necessary nonconductors. Existing floors, utility rails, building structures,

and standard seating (chairs and shop stools) may be tolerated within the volume
in space of the SSW subject to the CD Rule in accordance with Model Specifica-
tion 1. Plastic windows and fabric-covered walls are likely to violate the CD Rule,
and even glass windows may be a problem. (Note: See Example 8 in Chapter 8.) If
necessary, conductive chairs or chair covers may be procured from Reference
Document 4.
6. Construction review. All plant layouts and construction design drawings

shall be reviewed by Facilities and Industrial Engineering.
7. General guidelines. The following general guidelines shall be considered

in planning and designing and ESD-protected area. These guidelines provide in-
formation beyond the scope of Model Specification 1, which is intended to give
requirements with only a modicum of "how-to" detail. Thus, as new equipment
becomes available, the present document will be updated to utilize the equipment
while Model Specification 1 will not need to be revised.
7.1. Nonconductors. Plain plastic or synthetic fabric (such as wall panels)
shall be excluded if possible. Otherwise, necessary nonconductors shall be treated
with topical antistat or kept at a safe distance from ESDS items by the CD Rule.
7.2. Walls and other elevated surfaces. Walls and utility rails constructed of gyp-
sum board and painted with low-gloss latex are generally not an apparent-charge
problem. Problems can come from plastic pipes, signs, windows, and special wall
coverings such as high-gloss epoxy paint. The CD Rule shall be applied to these,
and the opposite side of the wall shall be checked for power transformers, electric
motors, or other machines which may produce fields that can penetrate the wall.
7.3. Floors. Tile or concrete floors must be treated with static-limiting noor
finish in accordance with Model Specification 1. Floors with ordinary finish ("wax")
or carpeting are an ESD hazard insofar as they fail to provide the desired safety
net. Antistat-treated carpet, or carpet sold as "antistatic," may be acceptable sub-
ject to the CD Rule, but ESD-protected areas should not be designed with carpet
when concrete or tile would serve.
7.4. Conductive floors. These may be needed where the mobility of workers
precludes the use of wrist straps, even with long cords and an overhead sliding
trolley system. Conductive floors without conductive footwear or shoe straps are
ineffective, and rubber-composition shoe soles may generate higher charges on
the wearer when he walks on conductive floors than when he walks on noncon-
ductive vinyl tile. Also, the resistance to ground may fall below the ARTG lower
limit of 1 megohm (paragraph 3.6.5.3.2 of Model Specification 1) with wet leather
shoes on a conductive floor, and we have recorded a case where the resistance was
below 100,000 ohms so that a person touching a 110 V hot lead would have carried
a possibly lethal current of more than 1.0 rnA. This danger must be considered
before buying and installing conductive floors and is a good reason, along with
expense, to avoid them if possible. (Note: Our approach calls for ordinary tile with
static-limiting noor finish.)
7.5 Work surfaces. Work surfaces must be an approved material from Reference
Document 4. In areas where processes such as conformal coating, potting, or chemi-
cal processing are carried out, other work-surface materials such as stainless steel
may be used after they have been added to Reference Document 4 with notes limit-
ing their use to special situations. A conductor such as stainless steel is not a novel
hazard at SSWs - hand tools, for instance, are often steel- but our philosophy
is to minimize conductive surfaces and maximize static-dissipative and anti-

static surfaces in order to best protect both personnel and ESDS items.
7.6. Ionization. Ionization can be used to reduce the ESD hazard in almost
any area. However, the general use of ionization is not recommended because it
will not allow any disciplines to be relaxed nor will it eliminate the requirement
for any ESD-protective materials or permit the casual use of plain plastics. Ioniza-
tion cannot limit the triboelectric charging process, as humidification can do,
but does control the build-up of charges on nonconductors; thus the hazard of
high apparent charges on some necessary items such as shields and laminar-flow
booth side panels is reduced. In other words, standing charges are controlled by
ionization of the room type, i.e., with overhead emitters. However, space charging
could make room ionization part of the problem that it purports to solve, so this
method should be considered only for special areas such as cleanrooms when the
advantages are well defined and the space-charging hazard is definitely controlled.
Localized ionization is necessary in grit-blasting, spray-coating, and operation of
some automated process machinery where triboelectric charging cannot be
eliminated.
7.7. Humidification. Humidity control does limit the triboelectric charging
process but does not eliminate any of the conventional safeguards; it is strictly a
backup or "safety net" measure. Also, humidity control may give personnel a false
sense of security and cause relaxation of operator disciplines, thus lowering ESD
safety. Humidity control also is expensive and can cause corrosion or other ad-
verse side effects. In conclusion, humidification is a backup that should be imple-
mented only after careful consideration of benefits versus cost and hazards.
7.8. Lights and light fixtures. Light fixtures with plastic covers or diffusers are
potential hazards because of fields from the plastic as well as from the lighting
element. Ultraviolet (UV) lights used in some inspection stations are safe when
ESDS items are exposed for limited times as in normal inspection. (Note: See
Example 27 in Chapter 8.)
7.9. Maintenance. ESD-protected areas shall be maintained under the close
supervision of ESD-certified personnel. For example, only static-limiting floor
finish, antistatic plastic trash-can liners, and approved cleaning agents must be
used. No cleaning or janitorial work shall be done while ESDS items are out of
their Faraday cages and not under the control of certified personnel.
(Note: This document is merely a model or sample and could be far more de-
tailed in practice. Guidelines for special areas such as cleanrooms could run to
many pages.)
Reference Document 2. Certification

of Static-Safe Workstations (SSWs)
1. Introduction. This document gives procedures for certifying the grounding

and the ground connections of certain critical, permanently fixed surfaces of SSWs.
2. Detinition of SSW. Model Specification 1 defines an SSW as a volume in

space which is designated and equipped to maintain an SSZ, which is a volume in
space at every point of which the Basic Rule (paragraph 3.5.1 of Model Specifica-
tion 1) is followed. The volume in space of the SSW includes assembly benches,
ovens, desiccators, wave-solder machines, test equipment, and other equipment in
which ESDS items are stored, processed, or handled.
3. Certitication procedures.
3.1. Certification decal. Check that each SSW has a current certification decal
to the requirements herein.
3.2. Electrical resistance from work surface to ground.
3.2.1. Benches with static-dissipative or antistatic work sUrfaces with a buried
conductive layer. Measure the resistance from work surface to ground with a
megohmmeter and an NFPA 99 electrode. Connect the negative ("low") lead of
the megohmmeter to the electrode and the positive ("high") lead to the grounding
lug. Place the electrode on the bench surface 18 plus or minus 2 inches from the
lug. Measure the resistance from the electrode to the lug at SOO V. The resistance
shall be between 1 X lOS and I x 1011 ohms. Alternatively, a Voyager RTG-210
Resistance to Ground meter or equivalent can be used. (When the resistance is
under about 7 megohms using the palm of the hand as the electrode, a Zapflash
will light. This is a useful quick check when the expected resistance to ground is
toward the low end of the required range. If the Zapflash doesn't light, the resis-
tance might still be acceptable, of course, by being in the middle or high end of the
required range.)
3.2.2. Benches with conductive work surfaces. Using a multimeter at 0.5-S0
V, probe between the work surface and the distribution wire (static ground). The
resistance shall be 1 megohm, provided by a current-limiting resistor as done for
static-dissipative or antistatic work surfaces. Laminar-flow booths, exhaust booths,
and similar facilities where the conductive surface is electrically common with the
structure shall have a resistance of O-S ohms, verified by probing between the sur-
face and commercial power ground. Note that conductive bench tops (work sur-
faces) are used in many cleanroom and special-process applications where the nature
of the task requires highly conductive materials such as stainless steel or aluminum.
3.2.3. Processing, storage, and test equipment. Using a multimeter at O.S-SO
V, probe between the commercial power ground connection on the chassis and
bare metal on the front (or on the inside for ovens and desiccators) of the equip-
ment. The resistance shall be O-S ohms. Verify the grounding of baffles in ovens
and environmental chambers by probing in a similar way; again, the requirement
is O-S ohms.
3.3. Potential difference between static ground and commercial power ground.
Static ground shall be used for assembly benches (3.2.1 and 3.2.2). Other SSWs
(3.2.3) may use commercial power ground. These two grounds shall not differ by
more than S V dc or rms. Continuous wrist-strap monitors use both these grounds
and will detect a gross failure in either.
3.4. Wrist-strap and ground wire connection. Check that each SSW has wrist-
strap and ground wire connections in good condition. Alligator clips and twisted
wire connections shall not be certified. All wrist straps must have continuous
wrist-strap monitors.
3.5. Resistance between the grounding lug ofthe work surface and static ground
distribution wire. Check that this resistance is 1 megohm, provided by a series
resistor in the lead wire which connects the grounding lug and the distribution
wire. Make sure that the lead wire is a minimum of 18 AWG and the distribution
wire is a minimum of 12 AWG. ("Minimum" refers to size; the higher the AWG,
the smaller the conductor.)
3.6. Resistance between static ground distribution wire and commercial power
ground. Check that this resistance is less than 25 ohms.
4. Identification of SSWs. Upon satisfactory completion of the above proce-

dures, a certification decal shall be applied to the SSW to record compliance of
grounding and ground connections with paragraph 3.6.5 of Model Specification 1.
Reference Document 3. Procurement

Requirements for Continuous
Wrist-Strap Monitor Unit
1. Scope. This document establishes requirements for a resistive (two-conductor)

continuous wrist-strap monitor unit, including an alarm device and special wrist
strap.
2. Applicable Documents. Model Specification 1, Handling and Assembly of

Unpackaged ESDS Items; Reference Document 4, Approved ESD-Control Mate-
rials and Equipment.
3. Requirements.
3.1. General requirements.
3.1.1. Overall design. The unit shall consist of a monitor (alarm device) and a
special wrist strap. The wrist strap shall consist of a two-snap wrist band and a two-
conductor coil cord for connection to the monitor. The unit shall provide two points
of skin contact and two ground connections. An alert shall be indicated in the
event of an "open" (high resistance) or "short" (low resistance) in either circuit.
3.1.2. Workmanship. Workmanship shall equal or exceed industry standard,
and fabrication shall be from good-quality components.
3.1.3. Provision for calibration. The monitor shall have provision for calibra-
tion and calibration seals.
3.1.4. Documentation. Service instructions, calibration procedures, and draw-
ings, including a schematic of the circuitry, shall be provided by the supplier.
3.2. Specific requirements.

3.2.1. Skin-to-ground resistance. Total skin-to-ground resistance shall be con-
tinuously measured at 0.1-5.0 V and shall be greater than 1 ± 0.1 megohm and less
than 10 ± 1.0 megohms with both circuits active.
3.2.2. Wrist band. The wrist band shall be (1) 3/4-inch-wide Velcro with D-
ring fastener or (2) adjustable, stretchable cloth. There shall be no conductive
material in the band other than two metal skin contacts attached to the snaps. The
contacts (plates or washers) shall be 0.1-0.6 inch apart. A wrist band shall be
supplied as part of the standard unit. (Note: This is our preference based on tests
for operator comfort. You may choose to allow stainless-steel expansion bands of
the Spidel type.)
3.2.3. Coil cord. The coil cord shall be resilient plastic, e.g., polyurethane or
polypropylene, with flexible conductors and shall be a minimum of 6 feet in length
when stretched to its practical limit. Banana plugs shall be provided for plugging
into the monitor.
3.2.4. Cord connection. Cord connections shall be of the quick-disconnect
spring type, which twists in any direction and allows release without damage to
either the springs or banana plugs. This is a safety feature allowing operators to
pull free easily.
3.2.5. Power supply. The monitor shall operate on standard 110 V, 6O-Hz house
power.
3.2.6. Ground connections. The system shall have two ground connections,
one for each point of skin contact. One connection shall be to commercial power
ground, e.g., via a three-pronged plug on the unit power supply, and the other
connection shall be to a static-ground wire system in accordance with paragraph
3.6.5.2.3. of Model Specification 1.
3.2.7. Alert signals. Alert signals shall be (1) an amber light to indicate high
resistance in either skin-to-ground circuit, (2) a green light to indicate that both
circuits are in the specified resistance limits, (3) a red light to indicate low resis-
tance (danger to operator), and (4) an audible alarm to accompany the amber or
red light. (Note: You may choose to omit the low-resistance warning light.)
3.2.8. Alert delay. The monitor shall have an alert delay of 0.25-1.0 second to
prevent annoying and unnecessary alert signals during a momentary discontinuity
due to normal body movement.
3.2.9. On-off switch. The on-off switch, if any, shall be recessed or otherwise
protected from being inadvertently turned on or off, e.g., by being struck by a
knee.
3.2.10. Provisionfor mounting. The monitor shall have provisions for mount-
ing either on top of or under the workbench surface.
3.2.11. Safety. The system shall be designed fail-safe for shock hazard.
4. Approved Source List. Units for purchase shall be selected from among
those listed in Reference Document 4, paragraph 3.3.
Reference Document 4. Approved ESD-Control

Materials and Equipment
(Note: This document lists materials and equipment selected according to the cri-
teria of Reference Document 5 as well as other criteria such as cost. The list gives
the name of the item, supplier's part number, supplier's location (city and state),
and the basis for approval of the item. Approval may be by standard test, special
test, or analysis, and the basis-for-approval column cites internal notebook page
references, letters, and test reports, as well as suppliers' data. See Reference Docu-
ment 5 for categories of materials and equipment listed.
This document is a "living list" which is constantly being added to and updated.
Listing approved items directly in the specification has proved impractical in our
experience because of the need for frequent revisions of the specification to keep
the list current, each revision requiring a time-consuming coordination with the
Engineering, Manufacturing, and Quality Assurance organizations.
Since ESD control is still a "young" field, novel products are always appearing.
For example, a digital desoldering apparatus meeting all the requirements of 000-
STD-2000-1B was advertised in late 1988 as featuring "the only static-dissipa-
tive desoldering handpiece on the market." Since the need for treating a plastic
handle with topical antistat would be avoided, this equipment is a good example
of a candidate worth evaluating. (Of course, we can't list commercial products in
this book, so now we'll proceed to Reference Document 5, which gives the criteria
for choosing them.)
Reference Document 5. Criteria for

Selection of Approved ESD-Control
Table o/Contents
1. Introduction
2. Publications cited
3. Materials/equipment and criteria
3.1. Bag materials and plastic films/sheets
3.2. Foam and bubble-wrap materials for cushioning
3.3. Apparel and operator grounding equipment
3.4. Tote boxes, bins trays, boxes, containers
3.5. Workbench tops and mats
3.6. Chairs, floor mats for chairs, seat covers
3.7. Floor materials (mats, tiles, carpets, static-limiting floor finish)
3.8. Static meters
3.9. Air ionizing (ionization) equipment

3.10. Desiccator cabinets
3.11. Topical antistats
3.12. Conductive tape
3.13. Shunt materials and lead-shorting devices
3.14. Metals
3.15. Test equipment
3.16. Miscellaneous
1. Introduction. The purpose of this document is to define criteria for selec-

tion of materials and equipment for handling unpackaged ESDS items in accor-
dance with Model Specification 1. The materials and equipment which have been
selected by these criteria are listed in Reference Document 4.
(Note: This document is merely illustrative. The criteria can and should be more
extensive than those shown. Some will be standard tests cited by suppliers, and
others will be special tests developed by the user.
Remember that many of the items listed are optional, as emphasized in some
cases by asterisks and notes; this is not a "shopping list" of necessities but an all-
inclusive compendium of items which might be used. In other words, the presence
of an item on the list does not imply the desirability of the item for most applica-
tions. An example is room ionization systems, which we feel are applicable only
to special areas such as cleanrooms if advantages such as an air-cleaning effect can
be demonstrated and the space-charging hazard is controlled.)
2. Publications cited.
2.1. Government documents.
SPECIFICATIONS (Military):
MIL-B-81705 Barrier Materials, Flexible, Electrostatic-

Free Heat-Sealable
(Other government activity):
KCS-C-123 John F. Kennedy Space Center,

NASA: Specification for Surface
Cleanliness of Fluid Systems
STANDARDS
Federal
FED-STD-l 0 1 Preservation, Packaging, and
Packing Materials: Test Procedures
Military

Electrical and Electronic
Parts, Assemblies, and
Equipment (Excluding Electrically
Initiated Explosive Devices)

Handbook for Protection of
Electrical and Electronic
Parts, Assemblies, and
2.2. Non-government documents.

SPECIFICATIONS
Model Specification 1 Handling and Assembly of Un-

Packaged ESDS Items
STANDARDS
American Society for Testing and Materials
Modified ASTM D 257 Tests for DC Resistance or

Conductance of Insulating
Materials. (The modification is
that a Voyager SRM-ll 0
Surface Resistivity Meter or
equivalent is used.)
National Fire Protection Association
NFPA99 Standard for Health Care Facilities
OTHER PUBLICATIONS
Reference Document 3 Procurement Requirements for

Continuous Wrist-Strap
Monitor Unit
EOS/ESD Association and /IT Research Institute
EOS/ESD Symposia, 1979- Electrical Overstress/Electro-

current static Discharge Symposium
Proceedings (Annual)
Government-Industry Data Exchange Program (GIDEP)
GIDEPAlert MX-A-82-02 Manufacturing/Process Tech-

(March 21, 1983) niques, Wrist Strap, ESD
Prevention
GIDEP Alert D5-A-84-0l Protective Clothing and Equip-
(May 25,1984) ment, Antistatic Smock
(Note: Many other documents could be referenced. The ones given are merely
illustrative.)
3. MateriaisiEquipment and Criteria. In the following subparagraphs, ma-

terials/equipment are listed in the left-hand column and criteria in the right-hand
column.
3.1. Bag materials and plastic films/sheets.
a. Black, conductive carbon- Volume resistivity greater than

loaded polyolefin 1000 ohm-cm at low voltage,
calculated from surface
resistivity (ASTM D 257) and
thickness.
b. Antistatic polyethylene Must be on QPL for MIL-B-
81705, Type II. Antistatic
agent must be a neutral, non-
corrosive compound. Shelf
life at R. T. for closed bags
should exceed 8 years (EOS/
ESD Symposium 1983; Paper
No.2 in the Appendix).
c. Antistatic clean nylon Bleed-off less than 1 second for
5000 V to 50 V at 700 P and
50% R.H. Cleanliness level
50A by Kennedy Space
Center KCS-C-123.
d. Faraday-cage foil laminate All ESD-prevention require-
ments of MIL-B-81705, Type I,
as well as an antistatic
polyethylene liner with sur-
face resistivity less than 1012
ohms/square. (MIL-B-81705B
had an unsatisfactory liner.)
3.2. Foam and bubble-wrap materials/or cushioning.
Antistatic foam and bubble-wrap FED-STD-I01, Method 4046: 2

seconds maximum charge decay
time. Surface resistivity less
than 1012 ohms/square.
3.3. Apparel and operator grounding equipment.
a. Smocks, antistatic Surface resistivity less than 10 12

ohms/square. (Successful
smocks have been 6 x 1()6, 3
X 107 , and 8 x 107 ohms/
square.) Must not shed steel
fibers (GIDEP Alert D5-A-84-
01,05-25-84).
b. Gloves and finger cots, Less than 50 V apparent charge
antistatic when stroked with polyester
fabric. Fingertip wall thick-
ness less than 0.003 inch.
Resistant to inorganic acids
and alkalis.
c. Wrist straps Part of continuous monitor unit
in accordance with Reference
Document 3.
d. Continuous wrist-strap Gives alarm by light and sound
monitor when the ARTG exceeds 10
megohms, in accordance with
Reference Document 3.
e. Heel strap, conductive Surface resistivity less than lOS
ohms/square.
f. Heel-grounder, with ankle Resistance from heel-grounder
strap to end of strap 1-2 megohms
at IOV.
g. Ground cord Built-in I-megohm resistor.
Sturdy construction.
3.4. Tote boxes. bins. trays. boxes. containers.
a. Static-dissipative tote box Surface resistivity at least 1()5

but less than 109 ohms/square
b. Portable bin system, anti- Less than 50 V apparent charge
static when rubbed with polyester
fabric. Antistat must be inter-
nal (extruded into the plastic).
Surface resistivity less than
1012 ohms/square.
c. Trays, conductive carbon- Surface resistivity less than lOs
loaded polyolefin ohms/square.
d. Box, transparent, antistatic Less than 50 V apparent charge
when rubbed with polyester
10 12 ohms/square.
e. Cardboard box, aluminum When closed, must be electri-
foil lined cally continuous over all
internal surfaces.
f. Box, conductive, double- Surface resistivity less than 1()5
walled carbon-loaded ohms/square.
polypropylene
g. Cubical antistatic container Less than 50 V apparent charge
when rubbed with polyester
10 12 ohms/square.
3.5. Workbench tops and mats.
a. Workbench tops (laminates) Resistance less than 1011 ohms

between grounding lug and a
5-lb., 2.5-inch-diameter elec-
trode (NFPA 99) on the sur-
face. Surface resistivity less
than 10 12 ohms/square.
b. Table mat, soft, static- Same as for workbench tops
dissipative (laminates).
3.6. Chairs,floor mats/or chairs seat covers.*
a. Conductive chairs Surface resistivity of upholstery

less than lOS ohms/square.
When chair sits on conductive
surface, less than I megohm
resistance from upholstery to
surface.
b. Hard conductive floor mats Surface resistivity less than lOs
for conductive chairs ohms/square.
c. Conductive seat covers Surface resistivity less than lOs
ohms/square.
3.7. Floor materials (mats, tiles, carpet, static-limiting floor finish).
a. Floor mat, conductive or Surface resistivity less than 109

static-dissipative* ohms/square.
b. Floor tile, conductive* Electrical continuity between upper
and lower surfaces.
c. Carpet, polyolefin, containing Surface resistivity less than 108
black, conductive fibers* ohms/square (based on testing
successful products). Electrical
continuity between black
fibers and conductive backing.
d. Static-limiting floor finish Topaka slip 0.20 minimum,
James slip 0.50 minimum,
gloss 30 minimum, surface
resistivity 10 10 ohms/square
maximum, when tested in
accordance with the unabridged
version of Paper No.6 in the
Appendix.
3.8. Static meters.
Hand-held field meters Favorable review in the literature,

e.g., EOSIESD Symposium
1983 (Ref. 4-8). Easily
calibrated by metrology.
*Optional items.
3.9. Air ionizing (ionization) equipment.
a. Ionized air blowers Neutralization rate for charged

plastic: at 2 feet from the
blower, time to reduce an
apparent charge on polyethylene
bubble-wrap from -3300 V
to -2000 V shall not exceed
30 seconds. Minimum safe
distance from corona-discharge
points no more than 12 inches
by special test with MOSFETs
(see under "Air Ionizers" in
Paper No.3 in the Appendix).
b. Ionized air nozzles, electrical Neutralization rate for charged
surfaces: to be determined by
experience. Minimum safe
distance from corona-discharge
points no more than 3 inches
(Paper No.3 in the Appendix).
c. Ionized air nozzle, nuclear Neutralization rate for charged
plastics: to be determined by
experience. (There is little field
from this equipment.)
d. Ionizer, pulsed, DC Frequency must be at least 5 Hz
(Example 21 in Chapter 8).
e. Grid, static-neutralizing Minimum safe working distance
(determined as for ionized air
blowers) no more than 12 inches.
f. Room ionization systems* Apparent charge on polyethylene
film at bench level must decay
from -5000 V to -500 V in less
than 3 minutes; similarly, the
apparent charge on polystyrene
foam must decay from +5000 V
to +500 V in less than 3 minutes.
Isolated conductors at bench
level must not be charged to
more than 50 V. Balancing control
must be automatic. Installation
must be attractive in appearance.
*Optional items.
3.10. Desiccator cabinets. FR-4 circuit-board laminate must

develop an apparent charge of
no more than 300 V in normal
operation (see Example 28 in
Chapter 8). The cabinet must
protect ESDS items from external
fields as demonstrated by analysis
or special test.
3.11. Topical antistats. Chloride level less than 100 ppm for
"chloride-free" type. Good wetting
of polyethylene surface for chloride-
containing type. Treated surface
must be antistatic.
3.12. Conductive tape. Surface resistivity less than lOS
ohms/square.
3.13. Shunt materials and lead-shorting devices.

a. Foam, antistatic or Must pass high-humidity test for
conductive, shunt. corrosivity to device leads (see
Example 14 in Chapter 8).
b. Connector shunt, antistatic Made from material (Type II) on
the QPL for MIL-B-81705.
(Note: The user must satisfy
himself that the feeble conductivity
of an antistatic material has
sufficient shunting effect; this is
generally doubted.)
c. Connector shunt, conductive, High-humidity exposure with
carbon-loaded plastic gold-plated connector pins;
see Example 14 in Chapter 8.
d. Lead-shorting devices Surface resistivity less than 105
ohms/square. Volume resistivity
less than 100 ohm-cm.
3.14. Metals. Low electrical resistance.

3.15. Test equipment.
Example: Zapnash multi- Lights dimly at 7 megohms but
tester not at 10 megohms or above.
(Note: This section includes surface resistivity meters, resistance-to-ground

meters, megohmmeters, etc. See the checklist of test equipment in Chapter 9.)
3.16. Miscellaneous (examples).
Voltage-suppressing screens for Special test for apparent charge

cathode ray tubes (CRTs) of using a field meter.
computers or oscilloscopes
Scrub brushes Antistatic or conductive handle.
May have nonconductive
plastic portion as allowed by
special test; see Example 11 in
Chapter 8.
Antistatic hand lotion No lanolin or silicones. Acceptable
ARTG when applied to dry skin.
Reference Document 6. Certification

of Personnel for ESD Control
In accordance with paragraph 3.6.13 of Model Specification 1, personnel shall

receive a minimum of 8 hours of training and pass a test for certification, and they
shall be recertified annually by taking a minimum of 2 hours of brush-up training
and passing a test.
The training course shall emphasize (1) compliance with Model Specification 1,
(2) "how-to" procedures, and (3) the "why" of procedures and rules being fol-
lowed. (See the sample test at the end of this chapter.)
(1) Experience has shown that training course material and the specification
being used may be poorly related and even contradictory! The objective should be
to teach people to comply with and implement the specification, not to question it,
ignore it, or violate it because training gave them other ideas.
(2) Among "how-to" procedures, operator disciplines should be heavily stressed.
Remember that an ESD "guru" could do assembly work safely at a wooden table
with no wrist strap, because knowledge and techniques are more important than
the best ESD-control materials and the most expensive, sophisticated equipment.
(3) The logical derivation of all ESD-control procedures from the Basic Rule is
the essence of our "A-to-Z" method, and operators and other non-engineers must
be given a simple conceptual picture of ESD damage models that makes every
precaution understandable. Fields may be pictured as a "toxic cloud" whose di-
minishing density is judged by apparent charge and distance, so that the prin-
ciple of the CD Rule is made obvious. Discharges are easily understood as sparks
or else subtle, invisible "sparks" that can "zap" a device; thus an operator will see
the imperativeness of grounding conductors to maintain the same potential and
avoid the flow of electrons like a liquid from a higher level to a lower. Electrons
can be pictured as "static dust" that rubs off of one surface and contaminates another
during triboelectric charging. Concepts like the ARTG are a little more difficult
to explain; in this case, a momentary surge of voltage must be envisioned, requir-
ing limited resistance (enough conductivity) to drain off the electrons so quickly
that they can't accumulate to the danger level.
Visual aids such as videotapes are valuable if they are tailor-made to fit the
specification. A demonstration with a curve-tracer, showing a MOSFET being
damaged by an ungrounded person's touch, is a vivid demonstration of proven
value in convincing the audience that ESD damage really happens. We remember
one videotape which showed an airplane crashing in a ball of flame because of
latent failure. This may be a bit exaggerated, but horror stories of ESD destruc-
tion of expensive assemblies (see Example 25 in Chapter 8) are documented and
cogent. No employee would want to be a guilty party in such an infraction, and no
one needs to be if he follows the rule: "If there's doubt about ESD protection in
this process, ask Engineering for advice." As in any field, the operator or engineer
on the shop floor must know the limits of his knowledge so that he won't overstep
them and risk either overt damage or the undetected "wounding" of devices that
can cause those hard-to-prove but always possible latent failures.
A suitable test is 20 questions with multiple-choice answers and a passing grade
of, say, 70%. The test should reveal whether the student understands how precau-
tions and rules flow from and implement the Basic Rule. People tend to follow
instructions when good reasons are given but to ignore "senseless" orders. See the
sample test at the end of this chapter.
We have found that decals applied to identification badges are useful in screen-
ing unqualified people from ESD-protected areas. Also, the decals raise the ESD-
awareness of their wearers and, like miniature diplomas, are a source of pride.
Thorough records of certification and recertification are required by MIL-STD-
1686 and should be kept in any event. If hardware should be mishandled, the tes-
timony of certified personnel will be more credible when their training is
documented. For example, an operator might say, "When my wrist strap pulled
loose, I was so absorbed in my work, which involved touching ESDS leads sensi-
tive to 300 V, that I ignored the alarm of the continuous monitor for a minute. But
as soon as I realized what had happened, I measured my voltage, by pointing a
field meter at the grounded work surface, and found it to be less than 100 V even
when I shuffled my feet (on static-limiting floor finish), which I hadn't been
doing during the infraction." Based on this report by an astute and certified person,
if the hardware tested O.K., it could be assumed to be safe from latent failure.
We emphasize once again that an operator is analogous to an airplane pilot who
can either "fly" or "crash" the most costly, well-equipped SSW. Since the operator's
skills are vital, he or she should have considerable autonomy and, as Dan Ander-
son says, should be "vice-president" of the SSW. An authoritarian management
system which withholds information because "knowledge is power" will not only
keep operators ignorant but make them indifferent and even hostile. In ESD con-
trol, personnel must be trusted as well as trained.
Model Specification 2. Packaging

of ESDS Items for Shipment, Storage,
and In-Plant Transfer
I. Scope
This specification establishes requirements for packaging constructions which meet
the definition (3.2, below) for a static-safe package (SSP). An SSP protects the
packaged ESDS item from both fields and discharges.
2. Applicable Documents
The following documents, of the latest issue in effect except as otherwise indi-
cated, form a part of this specification to the extent specified herein. In the event of
conflict between documents referenced herein and the contents of this specifica-
tion, the contents of this specification shall govern.
2.1 Government Documents.

SPECIFICATIONS
Federal
RR-W-365 Wire Fabric (Insect Screening)
PPP-C-795 Cushioning Material, Packaging
(Flexible Cellular, Plastic
Film) for Packaging Applications
PPP-C-1752 Cushioning Material, Unicellular
Polyethylene Foam, Flexible
(for Packaging Purposes)
PPC-C-1797 Cushioning Material, Resilient,
Low Density, Unicellular,
Polypropylene Foam
PPP-C-1842 Cushioning Material, Plastic; Open
Cell for Packaging Applications
Military
MIL-B-131 Barrier Material, Waterproof,
Greaseproof, Flexible, Heat-Sealable
MIL-S-19491 Semiconductor Devices, Packaging of
MIL-M-38510 Microcircuits, General
Specification for
MIL-R-39032 Resistors, Packaging of
MIL-B-81705 Barrier Materials, Flexible, Electrostatic-
Free, Heat-Sealable
MIL-P-81997 Pouches, Cushioned, Flexible,

Electrostatic-Free, Reclosable,
Transparent
STANDARDS
Military
MIL-STD-129 Marking for Shipment and Storage
MIL-STD-280 Definitions of Item Levels, Item
Exchangeability, Models, and
Related Terms
MIL-STD-794 Parts and Equipment, Procedures
for Packaging of
MIL-STD-1285 Marking of Electrical and
Electronic Parts
Explosive Devices)
2.2. Non-government documents.

SPECIFICATIONS
Model Specification 1 Handling and Assembly of

Unpackaged ESDS Items
STANDARDS
American Society for Testing and Materials
ASTM D 257 Tests for DC Resistance or Con-

ductance of Insulating Materials
ASTMD999 Vibration Testing of Shipping
Containers
3. Requirements
3.1. Order of precedence. In the event of conflict between the requirements

of this specification, a program contract, and drawings calling out this specifica-
tion, the requirements of the contract and the drawings shall take precedence in
that order.
3.2. Static-Safe Package (SSP). Packaging shall achieve an SSP, which is

defmed as an SSZ (3.2.1) within a Faraday cage (3.2.2) or within material providing
MODEL SPECIFICAnONS 109
sufficient shielding to pass a discharge test (4.3.3) on the particular configuration

of shipped/stored ESDS item and packaging. The Basic Rule (3.2.3) is assumed
to be followed in regard to triboelectric charging when antistatic materials known
to limit charging by the "lubricity effect" are used as the liner. (Note: This exem-
plifies the kind of assumptions necessarily made in any specification for ESD con-
trol. Special tests for every combination of ESDS item and package would give
high confidence in protection but are impractical.)
3.2.1. Static-Safe Zone (SSZ). An SSZ shall always be maintained around an
ESDS item, with an SSZ being defined as a volume in space at every point of
which the Basic Rule (3.2.3) is followed.
3.2.2. Faraday cage. The Faraday cage which is part of an SSP (3.2) is de-
fined as a container whose wall is electrically continuous and has at least the elec-
trical conductivity ofO.OOO25-inch (0.25-mil) aluminum foil. Small holes, such as
made by staples, are allowed.
3.2.3. Basic Rule. ESDS items shall always be protected by the Basic Rule,
which is: An ESDS item shall never, even for a nanosecond, be exposed to an E
field in violation of the CD Rule (paragraph 3.6.11 of Model Specification 1) or
have its sensitive leads or terminals touched to, or receive a discharge from, any
surface at more than 50 V.
3.3. Materials. The following materials shall be used subject to the require-
ments of this specification.
3.3.1. Specified materials.
a. Bag, foil laminate, antistatic version ofMIL-B-131, Class 1

b. Bag, MIL-B-SI705, Type I
c. Bag or film wrap, MIL-B-S1705, Type II
d. Pouch, MIL-P-SI997, Type II
e. Cushioning, PPP-C-1752, Type VII, Class 4
f. Cushioning, PPP-C-1S42, Type VIII, Style A or B
g. Cushioning, PPP-C-1797, Type II
h. Cushioning, PPP-C-795, Class 2
i. Wire fabric, RR-W-365, Type VII, IS x 16 regular
j. Bag, laminate with see-through metallization covered by an antistatic
film (exposed metallization, or metallization protected only by a coating
instead of a film, is undesirable). MIL-B-SI705C, Type III, with polyester
over the metallization, is suitable.
k. Bag, cellulosic, transparent, permanently antistatic, noncorrosive, for
intimate wrap
3.3.2. Other materials. The materials shall conform to the following require-
ments.
3.3.2.1. Conductive material. Noncorrosive conductive ESD-protective ma-
terial shall have a surface resistivity of less than lOS ohms/square.
3.3.2.2. Static-dissipative material. Static-dissipative material shall have a

surface resistivity of at least 1()5 ohms/square but less than 109 ohms/square when
measured at 700 P and 50% relative humidity.
3.3.2.3. Antistatic material. Antistatic material shall have a surface resis-
tivity of at least 109 ohms/square when measured at 70 0 P and 50% relative humidity.
3.3.2.4. Nonconductive material. Nonconductive material is defined as hav-
ing a surface resistivity of at least 1012 ohms/square when measured at 700 P and
50% relative humidity, minimum.
3.4. Conformance of package to an SSP (3.2). The package shall provide

Faraday-cage or equivalent protection (3.4.1) and shall limit triboelectric charging
(3.4.2) within its interior.
3.4.1. Faraday-cage or equivalent protection. Protection from fields and dis-
charges shall be provided by materials conforming to the definition of a Faraday
cage (3.2.2), e.g., materials 3.3.1.a, b, or i, or else the discharge test of 4.3.3 shall
be passed. (Note: In practice it may be necessary to allow the use of 3.3.1.j bags to
accommodate suppliers, and these bags may pass option 1 of paragraph 4.3.3 but
not option 2. However, we discourage the use of option 1 because 100 pP is low
for people and 1500 ohms is high; 185 pP and 150 ohms as in option 2 provide a
realistic worst case. See "8BM" in Chapter 4.)
3.4.2. Limiting oftriboelectric charging. Because of lubricity (3.2), the ma-
terials in 3.3.1, except band i, are satisfactory. Other materials (3.3.2) shall pass
the triboelectric charging test of 4.3.4, or else the ESDS item(s) in the package
shall be mounted securely so that sliding cannot occur. Even with approved liner
materials, sliding shall be minimized, e.g., by the pressure of cushioning material.
(Note: Material3.3.1.b, Revision B, had a nonconductive liner which caused charg-
ing that damaged MOSFETs in a published test (Ref. 4-7). However, this bag may
be lined with a bag of 3.3.1.c or 3.3.1.k in the technique of "double-bagging" (Ref.
9-1 and "Conclusions on Bag Materials" in Paper No.1 in the Appendix). Alterna-
tively, the packaged ESDS item could be prevented from sliding inside the 3.3.1.b,
Rev. B bag.)
3.4.3. Surface resistivity. Both interior and exterior surfaces shall be conduc-
tive, static-dissipative, or antistatic, or they may be nonconductive if the appar-
ent charge never exceeds 50 V because of voltage suppression. Examples of the
latter are carrying cases made of anodized aluminum or painted steel. Other ex-
amples are foil bags: 3.3.1.b (Rev. B) or MIL-B-131, Class 1, which is noncon-
ductive both inside and out. However, the safest policy is to double-bag with 3.3.I.b
(Rev. B) (see Note in 3.4.2), to use 3.3.1.b (Rev. C), or to use the antistatic version
ofMIL-B-131, Class 1, which is 3.3.1.a.
3.5. Corrosivity. The interior of the package shall be noncorrosive when tested
in accordance with 4.3.6.
3.6. Contamination. The interior of the package shall be noncontaminating

when tested in accordance with 4.3.7.
3.7. Rails and carriers. Rails shall be conductive, static-dissipative, or an-

tistatic and shall meet the requirements of 3.4.2,3.5, and 3.6. For supplier packag-
ing, carriers may be nonconductive. (Note: Nonconductive carriers are allowed
for two reasons. First, their small surface area limits their apparent charge and
their field. Second, their acceptance is a matter of practicality because so many
devices are provided in them.)
3.S. Shunting. ESDS leads, terminals, or contacts shall be protected by

shunting whenever possible.
3.9. Cardboard. Glazed cardboard shall not be used unless the glaze is a spe-
cial antistatic composition.
3.10. Physical protection. ESDS items shall be packaged in a manner that

will prevent damage and contamination (3.6) during handling and transportation.
3.11. Cleanliness. Prior to packaging, ESDS items shall be clean and free
from foreign material in accordance with the applicable component or assembly
specification.
3.12. Uniformity. Containers shall be uniform in size and shape for shipment
but not necessarily for storage or in-plant transfer.
3.13. Closure. Closure shall be secure, as needed, and according to instructions.
3.14. Marking. Marking shall be in accordance with MIL-STD-129. Also,

MIL-STD-1285 will be applicable as tailored for individual programs.
3.14.1 Description of contents. Each package shall be marked with the fol-
lowing information as a minimum:
a. Part No. or other identifying number of ESDS items inside package

b. Supplier's trademark or name
c. Quantity in package
d. Lot or date code (when applicable)
3.14.2. Transparency. Transparent containers need not be marked when

markings inside can be read easily. (Note: For transparency as well as superla-
tive shielding, use 3.3.1.i sandwiched between layers of 3.3.1.c. or use 3.3.1.,
which is MIL-B-81705C, Type III, with "metal in" (see Paper No. 15 in the
Appendix).
3.14.3. Markingfor ESD sensitivity

3.14.3.1. Exterior packs. Exterior packs shall be marked with a yellow cau-
tion label having black lettering. The MIL-STD-1285 symbol (beneath this para-
graph) or the EIA symbol shall be used, and the lettering shall say: "CAUTION
NOTE: This equipment contains parts and assemblies sensitive to damage by elec-
trostatic discharge (ESD). Use ESD precautionary procedures when touching, re-
moving, or inserting." A 2 x 2 inch label shall be placed on one side of each
intermediate container. Two 4 x 4 inch labels shall be placed on each exterior
container-one on the identification marking side (or surface) and one on the op-
posite side of each shipping container exceeding I/Z cubic foot. Smaller shipping
containers shall be marked in the same manner except that the 2 x 2 inch label shall
be used as an alternate to the larger one.
Note on symbols:
The ESD Association (Rome, NY) states that items that are
susceptible to ESD, such as ICs and assembled boards, should
use the triangle, hand, and slash symbol (Fig. a). ESD-protective
(a) products, such as static-shielding bags, should use the triangle,
hand, and arc symbol (Fig. b).
Some manufacturers use a lightningbolt symbol to indicate
ESD susceptibility, but that symbol should not be used because
it can be confused with one for electrical safety. Likewise, a
circle with three arrows entering it (Fig. c), which is often used
(b) for ESD protection, should not be used because it is generic
for electromagnetic, magnetic radioactive, or electrostatic fields.
3.14.3.2. Interior (unit) packs and containers for storage or

in-plant transfer. Unit packs or containers for storage or in-
plant transfer shall be marked with the MIL-STD-1285 sym-
bol (above), or the EIA symbol (paragraph 3.6.16.2 of Model
(c) Specification 1), and the statement: "Do Not Open Except at a
Static-Safe Workstation." Minimum size of the symbol shall
be one-third of an inch measured vertically. The symbol shall be printed in black
or in the same color as the identification marking if this is other than black. When
available marking space permits, the caution note of 3.14.3.1 may be used.
3.1S. Other packaging requirements. (Note: Many special requirements are pos-
sible, especially for physical protection, but the purpose of this model specification
is to emphasize ESD-protection requirements. On military contracts, documents such
as MIL-STD-794, MIL-S-19491, MIL-M-3851O, and MIL-R-39032 may be cited.)
3.16. Containers for storage or in-plant transfer. It is emphasized that con-

tainers, such as tote boxes or bins, for storage or in-plant transfer of ESDS items
shall meet all the requirements of this specification.
4. Quality Assurance Provisions
4.1. Responsibility for inspection. Unless otherwise specified in the contract

or order, the packager of the ESDS items shall be responsible for the perfonnance
of all testing required herein except for first-article inspections which are not qual-
ity-confonnance inspections; the buyer shall be responsible for the latter.
4.2. Materials of construction of package.

4.2.1. Specified materials. The packager of the ESDS items shall verify by
means of certification from the material supplier that materials procured in accor-
dance with 3.3.1 are in compliance with the specifications.
4.2.2. Other materials. The packager of the ESDS items shall verify by means
of certification from the supplier or by means of tests in accordance with 4.3.2 that
materials procured in accordance with 3.3.2 are in compliance with either 3.3.2.1,
3.3.2.2, or 3.3.2.3.
4.3. Methods of examination and test.

4.3.1. Certifications. Certification shall be by virtue of acceptance and ship-
ment of the order.
4.3.2. Surface resistivity. Surface resistivity shall be measured with a Voy-
ager SRM-l10 Surface Resistivity Meter, or equivalent, in accordance with the
intent but not necessarily the electrode configurations of ASTM 0257.
4.3.3. Discharge test. (Note: Options are as follows:
(1) High-resistance discharge test. The test apparatus is composed of (1) a steel
probe with resistor and capacitor and (2) a pair of disks to be placed inside the
package. The package shall be arranged so that both disks are pressed tightly against
its inside surface. The probe is 3/8 ± 1/16 inch in diameter and 3 ± 0.25 inches
long, and it has a hemispherical tip. It is connected by a 45-inch long, 16-AWG
insulated lead to a 1500-ohm resistor, which is connected by a 4-inch-Iong, 16-
AWG lead to a l00-pF capacitor. The capacitor is charged at 10,000 V (positive,
but either polarity might be used). Then, starting at 12 ± 2 inches from the pack-
age, the probe is moved toward the package at a speed of 5 ± 2 feet/second until
contact is made at the place where one of the disks (metallic, 0.75 inch diameter,
0.050-0.10 inch thick) touches the inside of the package. This upper disk is con-
nected to the gate lead of a 2N4351 or equivalent MOSFET with and HBM rating
of about 100 V, and the lower disk is connected to the substrate-case lead. The
package rests on a grounded metal plate so that the lower disk is separated from
the plate by the wall of the package. The disks are joined in a "capacitive sensor"
which separates them by 0.5 inch of nonconductive plastic (see Fig. 2 of Paper
No.3 in the Appendix). Alternatively, the disks may be separated from each other
by a minimum of 1/2 inch of air inside the package. Five consecutive MOSFETs
are tested, and protection is judged inadequate if one or more of the devices shows
a shift of more than 0.04 V in the gate-source threshold voltage (Vosem), if the
device is shorted, or if the current-voltage curve has changed in shape. Data are
obtained with a curve-tracer, Tektronix Model 576 or equivalent. Or a Static Event

Detector™ can be substituted for the MOSFET; see Paper No. 20 in the Appendix.
(2) Low-resistance discharge test. The test conditions and criteria for protec-
tion are exactly the same as for the above high-resistance test except that the ca-
pacitance is 185 pF and the resistor is 150 ohms. This is Test 5 in Table 2 of Paper
No.4 in the Appendix. We advise using this option unless expedience necessitates
option 1. A variation on this test using square electrodes attached to the inside of
the walls is suitable for tote boxes; see Paper No.7 in the Appendix.)
4.3.4. Triboelectric charging test. (Note: A standard test using quartz and
Teflon (Appendix G of EIA 541) is available, but the reader is advised to develop
his own test with the most common surfaces he plans to package. For example,
FR-4 circuit-board material was used in Papers No.2 and 7 in the Appendix, and
aluminum and an acrylic coating were also tested in Paper No.7. Use a Faraday
cup, and either shake coupons in a bag (as in the EIA 541 test), shake a box lined
with the material in question so that coupons slide back and forth over it (Paper
No.7 in the Appendix), or use the roller test developed by 3M and applied to floor
finishes in Paper No.6 in the Appendix. The allowable limit of charging expressed
in nC (nanocoulombs) must be set based on experience.)
4.3.5. Shunting. The item shall be visually inspected, with magnification as
required, for proper shunting.
4.3.6. Corrosivity. (Note: The user must select or develop appropriate tests for
his packaged items. A possibility is TAPPI T-406 (not referenced in this specifica-
tion) for tarnishing of silver, solder, and copper by reducible sulfur. Another is a
high-humidity, elevated-temperature test as used to indicate galvanic corrosion in
Example 14 in Chapter 8 or to detect corrosion of solder by transferred antistat in
Paper No.9 in the Appendix. The tests must be realistic so that corrosion will be
avoided but cost-effective packaging materials won't be ruled out; in contrast,
unrealistic laboratory tests might either pass an inadequate material or require such
unnecessarily high performance that a cost-effective candidate could fail.)
4.3.7. Contamination. (Note: Again, the user must select or develop pertinent,
realistic tests. See Paper No.9 in the Appendix for potential contamination prob-
lems from transferred antistats. Other materials may also cause trouble, e.g., car-
bon-loaded polyolefin which sloughs conductive particles or a fabric or other
composition containing stainless-steel fibers which might shed a fiber to short out
two circuit lines of a printed wiring board not yet conformally coated.)
4.3.8. Physical protection. The packaged item shall be visually examined with
the unaided eye for damage to the item as well as damage to the package. When
required, the completed package shall be subjected to vibration testing in accor-
dance with ASTM D 999, after which the ESDS item shall be functionally tested
in accordance with the applicable specification for item performance.
4.3.9. Cleanliness. The item shall be visually examined to determine that re-
quired cleanliness has been accomplished.
4.3.10. Uniformity. Packages shall be visually examined for adequate
uniformity.
4.3.11. Closure. The closure shall be visually examined.

4.3.12. Prescribed packaging. The package shall be examined and tested to
determine that all elements of the prescribed construction have been used and that
cardboard and paper are unglazed or have a special antistatic glaze. Foil thickness
shall be verified with a micrometer.
4.3.13. Marking. Visual examination of the completed package shall verify
the proper application of the specified marking.
5. Notes
5.1. EMIIRFI shielding. EMI/RFI shielding is outside the scope of this model
specification. However, protection is afforded by foil or screen, e.g., 3.3.1.a, or by
3.3.1.i, or by opaque, vapor-deposited metal (MIL-B-81705C, Type I).
5.2. Magnetic shielding. Shielding from H fields is outside the scope of this
specification. Iron foil may be used.
5.3. Definitions. MIL-STD-280 defines packaging terms such as "items

levels" and "models."
Multiple-Choice Test
for Personnel Certification
The following twenty questions are designed to test comprehension of Model Speci-
fications 1 and 2 and a course in ESD principles. A suggested passing grade is 70%
(no more than six wrong answers). Correct answers are given at the end of the
questions.
1. An ESD event that may cause device damage is: (a) always a surge (electron
flow), which may include a spark, (b) always a spark (air discharge), (c) always a
discharge from a person's finger or a tool, (d) always the result of high voltage
(over 3 kV).
2. Damage from the Human Body Model (8BM) is: (a) prevented by ground-
ing the skin, (b) the sole cause of ESD damage when people handle devices, (c)
caused by charges residing on the devices themselves, (d) prevented by operators
wearing conductive footwear on nonconductive floors.
3. The Charged Device Model (CDM) is prevented by: (a) grounding the skin
or tools that may touch a charged device, (b) grounding plastic or ceramic cases
of devices, (c) neutralizing charges on ESDS items with ionized air, when pos-
sible, and touching the leads only with antistatic materials, (d) touching leads
only with bare fingers, never a metallic tool.
4. An antistatic material is: (a) always a flexible plastic film, (b) electrically
nonconductive so that charges can't flow on or off it, (c) a material with low
surface conductivity but high volume conductivity, (d) a CDM-safe surface but a
poor barrier to external discharges when used for packaging.
5. For our purposes, a Faraday cage is: (a) an antistatic plastic bag or box, (b)
a carbon-loaded plastic bag or box, (c) a laminate containing aluminum foil, (d) an
electrically isolated open cup used for measuring static charges.
6. A static field or E field is: (a) a force field in space, (b) measured by a field
meter pointed at a charged surface, (c) the cause of ESD damage by the FIM, (d)
all of the above.
7. Relative humidity: (a) prevents all ESD events if high enough, (b) is not a
factor in ESD events, (c) affects the performance of most antistatic materials, (d)
must be controlled in all assembly areas and warehouses.
8. Air ionization: (a) is effective when used for specific reasons, (b) is a cure-
all making skin-grounding unnecessary, (c) is effective only at high relative hu-
midity, (d) neutralizes charges on conductors but not on nonconductors.
9. Allowable resistance to ground (ARTG) pertains to: (a) tote boxes lying on
workstations, (b) operators wearing wrist straps, (c) static-dissipative bench tops,
(d) conductive surfaces such as stainless-steel oven shelves.
to. The CD Rule is: (a) applied using measurements taken with a field meter,
(b) a technique of static-safe packaging, (c) calculated on the basis of relative
humidity, (d) never violated when continuous wrist-strap monitors are worn.
II. Recertification of operators is required: (a) when operators are transferred

to new work areas, (b) when new disciplines are introduced, (c) at regular intervals
with brush-up instruction and a test, (d) at irregular intervals with a test only.
12. A static-safe package (SSP): (a) uses antistatic film as a shield against
external discharges, (b) is opened only at a static-safe workstation (SSW), (c)
must be a rigid, metallic enclosure, (d) must be an opaque, foil-containing
laminate.
13. When your continuous wrist-strap monitor alarms, you must: (a) cease
work and determine the reason for the alarm, such as the need for applying anti-
static lotion to your wrist, (b) disconnect the strap to silence the alarm and finish
working, (c) ignore the alarm while you return the ESDS item to its SSP, (d)
continue work until your supervisor hears the alarm and comes to advise you.
14. If you detect an apparent charge of 5 kY on a wall bounding your SSW,

you must: (a) work as near the wall as you wish but don't touch leads to it, (b) stop
work and notify your supervisor so that the wall might be antistat-treated or the
SSW moved, (c) complete the work in hand and then calculate a safe working
distance by the CD Rule for future work, (d) work 5 inches from the wall by the
rule of 1 inch/1 kY.
15. You may allow a visitor to enter your SSW if he: (a) is ungrounded but
refrains from touching ESDS items, (b) wears an antistatic smock, (c) is grounded
and closely supervised by you, (d) keeps his feet outside the SSW boundary while
he leans in to observe.
16. If your hair is long, you must: (a) tie it back so it is within 2 inches of your
skin, (b) let it hang loose but don't let it brush ESDS items, (c) brush it to the left
side of your head and work to the right, (d) let it hang loose but keep ESDS items
at least 15 inches from it.
17. An ESDS item must always be in an SSP or SSW unless: (a) personnel
remain at least 5 feet from the item, (b) the relative humidity exceeds 50%, (c) the
item is under the continuous supervision of certified personnel while being trans-
ferred between SSWs, (d) the item is lying on antistatic plastic.
18. When a plastic tool handle is found to have an apparent charge of7 kY: (a)
the tool must be excluded from the SSW, (b) the handle should be given a one-
time treatment with topical antistat, (c) the handle should be treated every three
months with topical antistat unless field-meter checks show need for more fre-
quent treatment, (d) the handle is left untreated but not allowed to touch the leads
of ESDS items.
19. Pink antistatic bags: (a) should be washed with water when soiled, (b) should
be cleaned only with an antistat solution, (c) should be excluded from the SSW at
relative humidities below 30%, (d) are examples of Faraday-cage packaging.
20. A static-dissipative mat must be grounded by: (a) an alligator clip on the
snap with a lead to a metallic water pipe or building structural steel, (b) an ap-
proved snap connection with a lead to commercial power ground, (c) an approved
snap connection with a lead wire, containing a resistor, connected to the static
ground distribution wire, (d) an approved snap connection with a lead to a plastic
water pipe.
Answers: 1. a, 2. a, 3. c, 4. d, 5. c, 6. d, 7. c, 8. a, 9. b, 10. a, 11. c, 12. b,

13. a, 14. b, 15. c, 16. a, 17. c, 18. c, 19. b, 20. c.
Chapter 10
Industry Standards
Industry standards are becoming more important as government standards fade

away. As part of the Secretary of Defense's Specifications and Standards Refonn
Initiative, the Department of Defense Industry Review Panel on Specifications
and Standards is reviewing documents with the intention of canceling "non-value-
added cost drivers." Among the "endangered species" are the following:

handbook for Protection of Electrical
and Electronic Parts, Assemblies and
Equipment
MIL-STD-883 Electrostatic Discharge Sensitivity
Classification
Program for Protection of Electrical
and Electronic Parts, Assemblies and
MIL-B-81705 Barrier Materials, Flexible, Electrostatic
Protective, Heat-Sealable
FED-STD-I0l, Method 4046 Electrostatic Properties of Materials
Fortunately, there is a variety of industry ESD-control standards, and more are

in preparation. Unfortunately, not all are realistic, but you can decide which ones
to use and then reference them in tailored versions of the model specifications in
Chapter 9. When in doubt, use special tests, including the unofficial but realistic
tests described in Papers No. 15, 16, and 20 in the Appendix.
Some industry standards follow.
118
INDUSTRY STANDARDS 119
ASTM 0 257 Resistance or Conductance of

Insulating Materials
ASTM0991 Rubber Property-Volume Resistivity of
Electrically Conductive and Antistatic
Products
ASTM 0 2679 Standard Test Method for Electrostatic
Charge
ASTM F 150 Electrical Resistance of Conductive
Resilient Flooring
AATCC 134 Electrostatic Propensity of Carpets
CECCOOO15 Protection of Electrostatic Sensitive
Devices (Europe)
EIA 541 Packaging Material Standards for ESO
Sensitive Items
IEC 801-2 Electromagnetic Compatibility for
Industrial Process Measurement and
Control Equipment. Part 2, Electro-
static Oischarge Requirements
In recent years, the ESO Association, formerly called the EOS/ESO Associa-
tion, in Rome, New York, has created several more standards (Ref. 10-1 and 10-2).
In the list below, "s" means Standard, and "OS" means Draft Standard. The Oraft
Standards have been released to industry for comments and suggestions. "AOV"
means Advisory Program. Note that some standards are called "EOS/ESO" and
others "ESO."The "EOS/ESO" designation is to be changed to "ESO" in the course
of time when the standards are revised. In the right-hand column are Model Speci-
fication I paragraphs to which the standard might be applied. The standard marked
"N/A" is inappropriate in our view because it does not involve a spark discharge
(see "Discharge Test" in Chapter 4).
EOS/ESO-S 1.0-1987 EOS/ESO Standard No. I Standard 3.6.5.3.3

for Protection of Electrostatic
Oischarge, Susceptible Items:
Personal Grounding Wrist-Straps
EOS/ESO-S3.1-1991 ANSI EOS/ESO S.3.1-1991, Standard 3.6.8
Oischarge Susceptible Items-
Ionization
EOS/ESO-S4.1-1990 ANSI EOS/ESO S4.1-1990, Standard 3.6.4
Oischarge Susceptible Items-
Work Surfaces-Resistive
Characterization
EOS/ESD-S5.1-1993 ANSI EOS/ESD S5.1-1993, Standard 3.2

for Electrostatic Discharge (ESD)-
Sensitivity Testing-Human Body
Model (HBM)-Component Level
ESD-S5.2-1994 ESD S5.2-1994, Standard for Electro- 3.2
static Discharge (ESD)-Sensitivity
Testing Machine Model (MM)-
Component Level
EOS/ESD-DS5.3-1993 EOS/ESD DS5.3-1993, Draft Standard 3.2
for Electrostatic Discharge (ESD)-
Sensitivity Testing Charged Device
Model (CDM)-Component Level
EOS/ESD-S6.1-1991 ANSI EOS/ESD S6.1-1991, Standard 3.6.5.2
for Protection of Electrostatic Dis-
charge Susceptible Items-Ground-
ing-Recommended Practice
ESD-S7.1-1994 ESD S7.1-1994, Standard for Protection 3.6.9
of Electrostatic Discharge Susceptible
Items-Floor Materials-Resistive
Characterization of Materials
EOS/ESD-S8.1-1993 EOS/ESD S8.1-1993, Standard for 3.6.16
Protection of Electrostatic Discharge
Susceptible Items-ESD Awareness
Symbols
ESO-DS9.1-1994 ESO OS9.1-1994, Draft Standard for 3.6.5.3.5
Protection of Electrostatic Discharge
Susceptible Items-Footwear-
Resistive Characterization (not to
include heel straps and toe grounders)
EOS/ESD-S 11.11-1993 ESD S11.11-1993, Standard for Pro- 3.6.6.1
tection of Electrostatic Discharge
Susceptible Items-Surface Resistance
Measurement of Static Dissipative
Planar Materials
ESD-S 11.31-1994 ESD S 11.31-1994, Standard for Eval- N/A
uating the Performance of Electrostatic
Discharge Shielding Materials: Bags
EOS/ESD-ADV1.0-1991 EOS/ESD Association Advisory- 5.1
Glossary of Terms
The following Advisory Programs are also active as of this writing.
EOS/ESD ADV2.0 Considerations for ESD (Handbook)

EOS/ESD ADV3.0 International Standards
INDUSTRY STANDARDS 121
EOS/ESD ADV 4.0 ANSI Accreditation and Procedures

EOS/ESD ADV5.0 Military Standards and Communication
Further ESD Association standards are planned. As of this writing, the Associa-
tion is considering preliminary drafts on component handling equipment, solder-
ing equipment, volume resistance of materials, and measurement systems for
waveforms, radiation effects, and equipment effects. Also, a committee will focus
on ESD issues in cleanrooms.
ISO 9000 standards are becoming important in ESD control. They promote pro-
tection measures such as continuous wrist-strap monitoring (Ref. 10-3), which
is a mainstay of our recommended approach.
In conclusion, you should (1) review the government and industry documents
available on CD-ROM to make sure you have the latest versions and (2) contact
the ESD Association for a list of their available standards, handbooks, and advi-
sory programs. If the test methods seem practical and meaningful in terms of your
program, work them into your handling and packaging specifications.
Chapter 11
Living Without MIL-STD-1686
As of this writing, MIL-STD-1686, like many other government specifications

and standards, may be on the way out (see Chapter 10). Some government con-
tracts already exclude MIL-STD-1686 and, of course, it has never controlled com-
mercial work. But watchdogs such as government agencies, not to mention critically
minded customers, will continue to demand a well-documented ESD-control pro-
gram. The purpose of this chapter is to sketch the paperwork required.
The top document will be a statement of policy, indicating how lower-tier docu-
ments are to be used. An example follows.
Electrostatic Discharge (ESD)-Control

Policy for X Corporation
1. Scope
This document establishes X Corporations's policy and ESD-control program for
ensuring product quality by preventing damage or degradation by ESD during
handling, assembly, testing, shipment, storage, and in-plant transfer.
2. Applicability
This document pertains to all ESD-sensitive assemblies manufactured by all of X
Corporation's divisions and subsidiaries. Procedures required by this policy re-
main applicable to all ESD-sensitive components and subassemblies until the prod-
uct has been completely assembled and becomes a Faraday cage (shielded
enclosure) immune to ESD damage. All items shall be handled with the same
precautions as the most sensitive item.
122
LIVING WITHOUT MIL·STD-1686 123
3. Compliance with Industry Standards

This policy is in accordance with textbooks and handbooks widely consulted in
industry and universities. Accepted industry standards are imposed when applicable.
The principal elements of this policy are in effect at major corporations manufac-
turing critical, high-reliability equipment. All electronic items covered by this policy
are assumed to be of the highest ESD sensitivity level, thus ensuring rigorous
protection.
4. Implementation
The ESD-control policy is implemented by the following documents:
4.1 Procedures for Handling, Assembly, and Testing of Electrostatic Discharge
(ESD)-Sensitive Items
4.2 Procedures for Packaging of Electrostatic Discharge (ESD)-Sensitive Items
for Shipment, Storage, and In-Plant Transfer
4.3 Program Bulletins for Government Contracts
5. Order of Precedence of Documents

Specific ESD protection requirements noted on drawings shall take precedence,
except for government contracts, in which the program bulletin takes precedence.
6. Responsibilities and Actions

All personnel who design, assemble, inspect, test, package, repair, rework, install,
or maintain ESD-sensitive items shall be directly responsible for following all the
required ESD-control procedures. The Statistical Process Control (SPC) function
will monitor rates of ESD damage to reflect the success of the program. Manage-
ment will be responsible for providing the requisite training and equipment as well
as emphasizing the importance of ESD control and maintaining a high level of
employee awareness.
Note that this policy document serves as the ESD-control program plan. Also
note that there is no Quality Assurance (QA) organization. The workers monitor
their own quality, including ESD control, as part of a Total Quality Management
(TQM) approach. Another approach, to be seen below, is to have department man-
agers play the QA role.
As examples of 4.1 and 4.2 from the above policy document, the following Case
1 and Case 2 are tailored versions of the model specifications in Chapter 9. The
tailoring was done by the authors acting as consultants to two medium-sized com-
panies (approximately 500 employees each), and so these are real-life examples.
In both cases, all notes at the ends of paragraphs in the models were deleted.
Case 1
Model Specification 1. Retitled "Procedures for Handling, Assembly,

and Testing of Electrostatic Discharge (ESD) Sensitive Items"
PARAGRAPH CHANGES
First sentence changed to: "This document establishes in-plant

procedures for implementing the X Corporation ESD-Control
Program Plan."
2 "Specification" changed to "document."
2.2 Reference Documents 4 and 5 deleted.
3.2 Changed to: "ESDS items shall be considered sensitive to 100 V by
the Human Body Model."
3.6.5.2 Changed to: "Commercial power ground shall be used."
3.6.5.2.1. Deleted.
3.6.5.2.2.
3.6.5.3.5
Figure 1 "Static ground" changed to "Commercial power ground." "Carbon
paper layer" changed to "Static-dissipative mat." "Lug" changed to
"Snap."
3.6.6.1 Second sentence deleted.
3.6.8 Ref. Doc. 4 deleted.
3.6.6.2. Deleted.
3.6.7.
3.6.10(3).
3.6.12.1 0(3).
3.6.12.11.
3.6.12.12.
3.9.3.10
4.1 Changed to: "The department managers shall be responsible ... "
4.10 Deleted.
Ref. Doc. 1. Last sentence deleted.
para. 7
Ref. Doc. 1. Deleted.
para. 7.3. 7.4.
7.5.7.7.7.9
Ref. Doc. 2. In heading. "with buried conductive layer" deleted.
para. 3.2.1
para. 3.2.2.
3.3. and 3.6
Ref. Doc. 3. Ref. Doc. 4 deleted.
para. 2
para. 4
Ref. Docs. 4 Deleted.
and 5
UVING WITHOUT MIL-8TD-1886 125
Model Specification 2. Retitled "Procedures for Packaging of

Electrostatic Discharge (ESD)-Sensltlve Items for Shipment,
Storage, and In-Plant Transfer"
PARAGRAPH CHANGES
Changed to: ''This document establishes procedures for

implementing the X Corporation ESD-Control Program Plan."
2.1 All documents except MIL-B-SI70S deleted.
2.2 ASTM D 999 deleted.
3.2 Only "Faraday cage" allowed. Reference to discharge test (4.3.3)
deleted.
3.3.2 After "electrically continuous," new wording is: "and has at least the
shielding capability of 3.3.1.b."
3.3.1.a Changed to: "Bag, MIL-B-SI70S, Type I."
3.3.1.b Changed to: "Bag, MIL-B-SI70S, Type Ill, with polyester film over
metallization."
3.3.1.c- Deleted.
3.3.1.j
3.4 The words "or equivalent" deleted.
3.4.1 In heading, "or equivalent" deleted. Only "Faraday cage" allowed.
Reference to discharge test (4.3.3) deleted.
3.4.2 Changed to: "Because of lubricity (3.2), the materials in 3.3.1 shall
be satisfactory."
3.4.3 Changed to: "Both interior and exterior surfaces shall be static-
dissipative or antistatic."
3.S Changed to: "The interior of the package shall be noncorrosive."
3.6 Changed to: "The interior of the package shall be noncontaminating."
3.14 Changed to: "Marking shall be in accordance with Electronic
Industries Association (EIA) requirements."
3.14.3.1 Illustration changed to EIA symbol (reaching hand).
3.IS Deleted.
4.1 Changed to: "Department managers shall be responsible for
compliance."
4.2,4.2.1, Deleted.
4.2.2,4.3.1,
4.3.3,4.3.4,
4.3.6
4.3.7 Changed to: "The item shall be visually examined for
contamination."
S.I "Model" and "or screen" deleted.
Case 2
Model Specification 1. Retitled as In Case 1.
Paragraph Changes
First sentence changed to: "This document establishes in-plant
procedures for implementing the X Corporation ESD-Control
Program Plan."
2.1.2.2 Deleted.
3.1 Changed to: "In the event of a conflict between the requirements of
this document and drawings calling out this document. the
requirements of the drawings shall take precedence."
3.2 Changed to: "ESDS items shall all be considered sensitive to lOOV
by the Human Body Model."
3.3 Changed to: "The need for design protection shall be determined in
collaboration with component suppliers."
3.3.\ Changed to: "The least sensitive component that will meet
performance requirements shall be used. When necessary. protective
circuitry shall be incorporated at the lowest practical level of
assembly to limit the sensitivity of that assembly."
3.3.2 Changed to: "Unit terminals. including test points. located on the
surface of the unit enclosure shall contain protective circuitry."
3.6.2 Deleted.
3.6.4 "Reference Document 2" changed to "Note 5.3."
3.6.5.2 Added: "Also. voltages surges from equipment may occur on
commercial power ground."
Figure I "Carbon paper layer" changed to "Static-dissipative mat." "Lug"
changed to "Snap."
3.6.5.3.5 Deleted.
3.6.6.1 "Surface resistivity meter" made generic. Reference to Ref. Doc. 4
and 5 deleted.
3.6.6.2. Deleted.
3.6.7
3.6.8 Changed to: "Air ionizers shall be used only when absolutely
necessary. e.g .• in peeling nonconductlve tape from rolls. Ionized
air blowers shall be electrical. but nozzles (guns) may be either
electrical or nuclear. The blowers shall have an autobalancing
feature. All Ionization ... of the ionizers."
3.6.9 Changed to: "An approved static-limiting floor finish shall be
applied to all tile floors and maintained so that appearance is
attractive. slip resistance is adequate. and the surface resistivity never
exceeds 10" ohms as measured between two NFPA 99 electrodes
1 inch apart."
3.6.10.4 Changed to: "TIle floors (carpets are unacceptable)."
3.6.10.11 New item added: "Plastic panels on equipment."
3.6.12.6 Changed to: "Hair shall not hang in front of the operator so that it
touches or closely approaches ESDS items."
3.6.12.12 Added: ..... other side of the fence or nonconductlve plastic is stored
there."
LIVING WITHOUT MIL-STD-1686 127
3.6.13 "Ref. Doc. 6" changed to "Note 5.4." "Annually" changed to "every
two years."
3.6.15 Deleted.
3.6.16.1 MIL-STD-129 deleted.
3.6.16.2 Sealing of tags in envelopes deleted.
3.7 Changed to: "Audits and reviews of the ESD-control program shall
be conducted as required."
3.9 Changed to: "Automated processes shall be periodically reviewed
for ESD safety."
3.10 Changed to: "As soon as possible after the infraction, the following
ten data items shall be recorded. Disposition of the hardware then
shall be decided by a Material Review Board." (The ten items are
listed from Chapter 13.)
4.1 Changed to: "All personnel handling ESDS items shall be
responsible for compliance with the requirements specified."
4.2 Changed to: "SSWs, ionization equipment, and personnel shall be
certified. "
4.5,4.6, Deleted.
4.7,4.8
4.10 Changed to: "Verification of automated processes. Verification of
automated processes shall be done monthly."
5.3,5.3.1, Deleted.
5.3.2, Ref.
Doc. 1
Ref. Doc. 2 Relabeled "Para. 5.3."
New 5.3, "Document" changed to "note."
subpara. 1
New 5.3, Changed to: "AN SSW is ... "
subpara.2
Ref. Docs. 3, Deleted.
4,5, and 6
New 5.4 Added: "Operator training."
New 5.4.1 Added: "Certification. Operators shall receive eight hours of
instruction, including a 30-minute videotape reviewing the basic
reasons for ESD control and illustrating the key procedures required
by this document, and operators shall be given a test with a
minimum of 20 multiple-choice questions (passing score 70%)."
New 5.4.2 Added: "Recertification. Operators shall receive two hours of
instruction, including a 15-30 minute video illustrating key
procedures, as well as a test as in 5.4.1."
New 5.4.3 Added: "Badge decal. Certified operators shall have a distinctive
decal applied to their company badges."
Model Specification 2. Retitled as In Case 1.
PARAGRAPH CHANGES
Changed to: "This document establishes procedures for implement

ing the X Corporation ESD-Control Program Plan."
2.1 Deleted.
2.2 AlI documents except ASTM D 257 deleted.
3.1 Changed to: "In the event of a conflict between the requirements of
this document and drawings calling out this document, the
requirements of the drawings shalI take precedence."
3.2 Changed to: "Packaging shall achieve an SSP, which is defined as an
SSZ within a Faraday cage, for in-plant handling. Supplier
packaging shall be judged acceptable based on the ESD sensitivity
of the packaged parts."
3.3 Changed to: "Material. The bag material shalI be a foil laminate
consisting of antistatic spun-bonded polyethylene on the outside,
O.25-mil (O.OOO25-inch) aluminum foil in the middle, and antistatic
polyethylene on the inside. Foam for cushioning shalI be antistatic."
3.3.2, Deleted.
3.3.2.1,
3.3.2.2
3.3.2.3 Added: " ... in accordance with ASTM D 257."
3.3.2.4 Deleted.
3.4.1 Changed to: "Protection from fields and discharges shall be provided
by material (3.3) conforming to the definition of a Faraday cage
(3.2.2)."
3.4.2 Changed to: "Both interior and exterior surfaces shall be antistatic."
3.5 Deleted.
3.6 Changed to: "The interior of the package shall be free of
contamination such as dust or fibers."
3.7 Changed to: "Rails shall be conductive, statlc-dlsslpatlve, or
antistatic. "
3.11 Changed to: "Prior to packaging, ESDS items shalI be clean and free
from foreign material."
3.14 Changed to: "Each package shall be marked with the following
information as a minimum."
3.14.b,3.14.c, Deleted.
3.14.2
3.14.3 "Exterior packs" changed to "packages." "MlL-STD-129" changed
to "EIA. "l11ustration changed to EIA symbol (reaching hand).
3.14.3.2, Deleted.
3.15,3.16
4.1 Changed to: "Unless otherwise specified in the contract or order, the
packager of the ESDS item shalI be responsible for the performance
of all testing required herein."
4.2,4.2.1, Deleted.
4.2.2,4.3.1,
4.3.2,4.3.3,
4.3.4,4.3.6
LIVING WITHOUT MIL-STD-1686 129
4.3.7 Changed to: "The item shall be visually inspected."

4.3.8 Changed to: ''The packaged item shall be visually examined with the
unaided eye for damage to the item as well as damage to the
package."
4.3.12 Changed to: "The package shall be examined to determine that all
elements of the prescribed construction have been used and that
cardboard and paper are unglazed."
5,5.1, Deleted.
5.2,5.3
In Case 1, Model Specification I is much simplified. Commercial power ground

is used for static ground. Soft mats instead of hard laminates are used on ESD-
control benches. Materials such as floor mats and static-limiting floor finish are
deleted. There is no list of required materials/equipment and no provision for dis-
position of mishandled hardware. Department managers are responsible for QA.
In Case 2, Model Specification 1 is altered differently. Static-limiting floor
finish is retained and, in fact, only tile floors are allowed. The chain link fence
provision is applicable, with non conductive plastics being stored behind the fence.
Details of operator certification, and recertification every two years, are given.
Operators are their own QA.
Model Specification 2 is also much simplified in both cases. Again, Case 1 as-
signs QA to department managers, while Case 2 makes the "packager" (operator
doing the packaging) responsible. A major difference in the tailoring is that Case 2
uses opaque foil-laminate bags because maximum ESD protection was desired
and transparency was felt not to be needed. Also, very durable bags were desired
to avoid the impractical task of periodically collecting bags and testing them for
effectiveness. But, in Case 1, transparency was a need, for reading serial numbers
of packaged items, and the bags were assumed to be effective until they appeared
wrinkled enough to be discarded.
These bag selections illustrate the quite different approaches taken by organiza-
tions, and you must make your own unique choices of ESD-control materials and
procedures based on ESDS item sensitivity, operator skill level, budget constraints,
and many other factors.
The above examples illustrate the minimal paperwork of a policy document,
also serving as an ESD-control program plan, with handling and packaging proce-
dural documents to implement it. Many satellite documents can be added, of course,
such as manuals on shop practices, reporting test data, operator training, and equip-
ment selection. All this documentation can be, and may have to be, created without
the basic reference of MIL-STD-1686. The main requisites for success are that
documents be complete and lucid and that responsibilities for adhering to them be
clearly assigned. See the capsule discussion and flow diagram on documentation!
organization in Chapter 3.
Chapter 12
Program Organization
and Implementation
This chapter includes several topics: reasons for an ESD-control program; history
of ESD control and purpose of this chapter; objectives and overview of program;
what? when? and how?; cost-effectiveness; plan of action; the enforcement prob-
lem; advice to small companies; the role of the program coordinator; and safety.
Reasons for an ESD-Control Program

One reason for an ESD-control program is that it is required by a government
contract. In this case, compliance with MIL-STD-1686 (formerly DoD-STD-1686)
will be necessary, and three Data Item Descriptions (DIDs) are involved: DI-RELI-
80669 for an ESD-control program plan, DI-RELI-80670 for reporting results of
ESD-sensitivity tests of electrical and electronic parts, and DI-RELI-80671 for
handling procedures for ESDS items. These will be Contract Data Requirements
List (CDRL) items to be submitted for approval by the procuring agency.
A second reason is that an effective ESD-control program saves money by lower-
ing rejection rates of products. Many articles and papers have attested to large sav-
ings. For example, one company spent $200,000 on ESD control beginning in 1980.
Before that, ESD damage was estimated at $800,000 to $1,200,000 per year; after-
ward, it fell to under $100,000 per year. Also, warranty repairs fell from $270,000
per year to $180,000 per year (Ref. 12-1). Another large corporation achieved sav-
ings in remanufacturing and repair that were anticipated to reach $21,000,000 in 5
years (Ref. 12-2). Incidentally, selection of the right materials and equipment can
save much money within the ESD-control program itself. Reference 12-2 cites the
purchase of wrist straps and charge-draining materials costing about $6,000 instead
of $50,000 worth of air-ionizing blowers, which would have been inadequate by
themselves anyway. We'll say more about prudent expenditures later in this chapter.
A third reason for an ESD-control program is to prevent latent failures, although
instances of these are difficult to prove (see "Latent Failure" in Chapter 4). The
130
PROGRAM ORGANIZATION AND IMPLEMENTATION 131
possibility of latent failure could be enough to cause mishandled hardware to be

declared suspect or even quite unacceptable by a government customer after an
unfavorable audit of manufacturing practices. MIL-STD-1686 is sketchy enough
for people to rationalize away the requirements in a token effort, but the govern-
ment is beginning to realize that latent failures are possible in mission-critical
equipment. Therefore, the war against ESD damage must be fought even though
the enemy is invisible-a process that might be likened to exorcism, but with
ghosts which could be real. Don't let disciplines become lax with the excuse that
yields are high. Besides yields, the reliability of hardware is at stake, and deficien-
cies can become evident only over the long term.
Remember that every ESD event does not cause measurable damage to hard-
ware. In fact, many ESD experts have concluded that 90% of ESD damage is less
than catastrophic and will not be detected until the degradation develops into a full
functional failure, or intermittent malfunction, after the product is in use. These
devices are walking wounded or candidates for latent failure, and they may cause
unnecessary rework/retest hours, field returns, and a decrease in MTBF and MTBR.
Minimizing the cost of dealing with the walking wounded will put an effective
ESD-control program on a paying basis, though savings aren't obvious as they are
for products killed by ESD during manufacturing.
History of ESD Control

and Purpose of This Chapter
During the past several years, ESD-control measures have been implemented in
most major electronics manufacturing firms. At first a few simple controls were
imposed on assembly operators. Then the addition of more complex procedures
caused turmoil while the organizations adjusted. As time passed, uncongenial ad-
ditions were eliminated or ignored until they disappeared. With the advent of DoD-
STD-1686, the industry was challenged with converting existing control measures
into formal programs. The commercial sector, which would ordinarily not be con-
cerned with government documents, is finding this standard (now MIL-STD-1686)
useful as a guide for establishing a technically sound, cost-effective ESD-control
program. Unfortunately, the measures used by many organizations have grown in
an unplanned way. The purpose of this chapter is to suggest a systematic approach
to formalizing a program with flexibility to grow with ever-changing ESD-control
technology and products. Readers who already have formalized programs will find
suggestions for reevaluating and upgrading program elements.
Objectives and Overview of Program

Simply stated, the objective of an ESD-control program is to protect ESDS hardware
from damage by ESD. This requires a coordinated effort by every function from
the chief executive to the new hire in the floor-maintenance crew. All too often the
ESD-control program is viewed as a set of rules inflicted on assemblers and tech-
nicians who process hardware. This distorted viewpoint destines a program to fail,
in the sense that the program costs more that it saves. Remember that operators
stand in the front line of the ESD battle, but without the logistics of the rear ech-
elon providing materials, equipment, and direction, the battle will be lost. There-
fore, a total management system must be developed to support the front lines.
A realistic and systematic approach is needed. This requires information on the
flow of product through the plant, from receipt and inspection of raw materials to
quality control and shipment of finished hardware. The factory is a complex, deli-
cately balanced structure whose operation is almost impossible to diagram be-
cause many support functions are not clearly defined. However, overlooking one
seemingly obscure function can jeopardize an ESD-control program. Engineering
organizations may be equally bewildering, but none of their elements can be slighted.
Only a total, integrated effort can succeed.
The beginning of the development of an ESD-control program is an informa-
tion-gathering process in which the following questions must be answered:
I. What is the lowest V-zap level (see "HBM" in Chapter 4) in each process?
2. What control methods are best suited to each step in the process? For example,
are wrist straps preferable to conductive floors combined with conductive foot-
wear? (We recommend wrist straps, with long cords and overhead trolleys if nec-
essary, whenever possible. Conductive floors can be hazardous, as pointed out in
Reference Document 1 in Chapter 9.)
3. What level of personnel training will be required?
4. Are existing policies supportive of an ESD-control program?
5. What new company policies will be required to enforce ESD-control
disciplines?
6. What special skills will be required in each area, e.g., assembly versus testing?
More questions will become apparent as the information-gathering process proceeds.

Now comes the real challenge: arranging the information in a meaningful for-
mat. One practical method is to construct a functional profile delineating the inter-
action of each function with other activities. This profile should identify key
individuals in the management structure. The profile should include a matrix tabu-
lating elements such as workstation types, worker skills, special process require-
ments, environmental conditions and requirements, and other conditions impacting
ESD-control disciplines.
What? When? and How?

What? when? and how? are the questions for program implementation. What areas
and processes require ESD control will be easily identified from the functional
profile. "How?" is the tough question, so let's look at "when?" first. It is impracti-
cal to implement total ESO control all at once. The first control measures should
be introduced concurrently with personnel training in the areas of most vulnerabil-
ity to ESO. These initial measures are easiest to sell because their cost-effective-
ness is readily demonstrated. After the first disciplines are in place, refinements
and expansion of ESO controls to the balance of the factory will commence. With
refinements, diligent attention to detail will be needed to prevent reverting to the
old way. Workers will have to break old habits and establish new ones, and
management's failure to recognize and support the habit-changing aspects of the
program will cause it to be ineffective and fail to produce the desired cost savings.
"How" involves the most controversy. Each salesperson selling ESO-protective
materials and equipment claims to have the best "how" and will quickly tell you
why competitors' "hows" will not work. The only practical solution to this di-
lemma is to buy only products that clearly implement the Basic Rule. Remember
that three main damage mechanisms are at work. First is DI, of which the HBM is
the imperfect simulation (too high resistance, for one thing) of the most common
source of current flow: touching by people. Second is the CDM, in which the
device is charged and subsequently is damaged by DI when an ESDS lead touches
a surface with some conductivity, especially a truly conductive one. Third is the
FIM, in which a field causes a difference in potential to be induced in the device
so that the device "zaps" itself from within. A special case of this is the FFB, in
which a field develops on the board holding the devices; when the damage occurs
during subsequent discharge, the FFB resembles the CDM. The FIM is the most
insidious of the damage models because of the complex nature of internal circuitry
and the unknown locus of electrical energy imparted by a field. Repeatability in
FIM testing is difficult, and data can't predict the outcome of a subsequent test
where one of the parameters such as field strength is changed. There is no stan-
dard test established by government or industry for establishing reliable tolerance
thresholds. However, a conservative safe-distance range has been empirically de-
veloped and over the years has proven safe and reliable; this is our own CD Rule.
With the above three damage models in mind, materials and equipment should be
selected to combat specific threats. For example, antistatic materials, as opposed
to conductive, will be chosen to discourage the CDM by lessening the rapidity of
discharges.
Cost-Effectiveness
Special ESO-control program design objectives should be developed around facili-
ties and equipment as configured in the shops and laboratories engaged in producing
ESDS hardware. Therefore, no one program will fit all occasions. Cost-effective-
ness, which is the driving force with top management, is attained by accurately de-
fining control requirements and fitting them to existing conditions. Factors in selecting
these requirements are: (1) lowest level of ESO-sensitivity, (2) mission criticality of
the product, (3) cost of ESDS components, (4) cost and feasibility of rework, (5)
schedule impact of ESD failures, and (6) willingness and ability of workers to accept
and implement more or less complex disciplines of ESD control.
In this book we have made the following assumptions about the program illus-
trated by our model specifications in Chapter 9:
I. Lowest level of sensitivity: 100 V as determined by the OBM, corresponding

to a maximum safe apparent charge of 300 V on nonconductors.
2. Mission criticality: high, e.g., a life-support system.
3. Cost of sensitive components (ESDS items): moderate to high.
4. Cost and feasibility of rework: cost moderate to high, feasibility varied.
5. Schedule impact of failures requiring rework: manageable.
6. Worker willingness and ability: good (but may require union cooperation).
Operator disciplines are the cornerstone of a cost-effective program emphasiz-

ing skillful handling rather that an attempt to "buy one's way out of the problem" by
"throwing money at it." Indeed, such an attempt would fail, because the most expen-
sive materials and equipment are useless if not used intelligently. Some key opera-
tor disciplines (see paragraph 3.6.12 of Model Specification 1 in Chapter 9) are:
1. Never touch an ESDS device unless absolutely necessary.

2. When a device must be touched, avoid the leads and handle it by the case.
3. Touch devices only to approved materials which are grounded. Antistatic
materials are safer to touch than are conductive materials (Chapter 5 and Papers
No. 3,7, 13, and 14 in in the Appendix).
4. Never subject a device to a field of an unknown safe level. The safe level
(safe distance for a given apparent charge) is determined by the CD Rule.
5. Before a device leaves the SSW, place the device in an SSP or closely super-
vise it till it enters another SSW.
In achieving cost-effectiveness, don't cut comers by accepting myths. Myths

which will weaken an ESD-control program are discussed in Paper No. 11 in the
Appendix. Some of the more dangerous are:
1. "A wrist strap will render the human skin harmless." The truth is that ESD is
first and foremost a people problem, and a grounded operator is only the lesser of
two evils versus an ungrounded operator because of the CDM. Furthermore, the
mere wearing of a wrist strap doesn't guarantee that the ARTG is being met; con-
tinuous monitoring is necessary.
2. "Higher humidity will solve all ESD handling problems." The truth is that
increasing the humidity is far from a panacea and will only help. It would be more
correct to say that low humidity will aggravate problems when ESD-control
measures are marginal.
3. "Components are safe once they are mounted on a printed circuit board."
The surprising truth is that they can be at greater risk, because circuit lines in-
crease the target area for DI and CDM events and act as antennas for FlM events.
At best the board can provide protective networks which will reduce the ESD
hazard but not eliminate it.
4. "Faraday cages are constructed of carbon-loaded plastic." The truth is that
carbon-loaded conductive plastic provides some field attenuation but insufficient
protection from DI. See "Faraday Cage" in Chapter 4.
5. "Ionization will allow the use of plain, nonconductive plastics in an ESD-
controlled area." The truth is that ionization will neutralize charges on ungrounded
nonconductors but will not prevent triboelectric charging in the first place. This
charging can occur in an ionized atmosphere and zap a part before ionization can
neutralize the charge! Therefore, ionization cannot eliminate requirements for
passive control measures.
6. "Ionization is a substitute for the use of wrist straps or other grounding of
personnel." This is an egregious myth that has been proposed in the past by at least
one equipment supplier. The truth is that grounding is necessary whether the at-
mosphere is ionized or not, because ionization takes time to neutralize a charge
and ESDS items can be damaged before the ions have done their work. Remember
that the Basic Rule requires protection "even for a nanosecond."
7. "Devices with protective networks (designed and built in) are not ESDS."
The unfortunate truth is that protective networks make devices less sensitive but
can't provide total protection.
There is no one right to control ESD and, in developing a cost-effective pro-

gram, one must innovate while keeping the ground rules intact. Remember that
operator disciplines must be compatible with the paraphernalia used in the work-
place. For example, if smocks are the norm, make them ESD-protective smocks;
then personal clothing will be of much less concern and sleeves needn't be short
or rolled up (paragraph 3.6.12.10 of Model Specification 1 in Chapter 9). If
reusable packaging is employed, the procedure for handling it will differ from
that for throwaway materials. Using conductive flooring with conductive foot-
wear to ground the skin, as opposed to wrist straps, will modify disciplines; for
example, shoe straps or soles will have to be kept clean to maintain conductivity
of their surfaces. The near-term cost compared to the long-term cost of owner-
ship of equipment must be evaluated up front, to avoid later changes which might
impact productivity. For example, nuclear ionizers have an annual cost for the
polonium-210 elements. If this cost is discovered after a few years and electrical
ionizers are substituted, rules must be imposed for avoiding FIM damage from
fields created by these new ionizers, and failure to add these rules could cause
loss of ESDS hardware (see paragraph 3.6.8 in Model Specification 1 in Chapter
9). Control measures must be tailored to the work area and keep up with changes
in it.
Plan of Action
A plan of action can now be fonnulated to assign the responsibilities and actions
required to put the ESD-control program into place. This plan should include, but
not be limited to, the following elements:
1. Proposals. ESD-control requirements must be considered when a program

is still in the proposal phase. Expected levels of sensitivity should be identified
throughout the in-house production process as well as repair and maintenance ac-
tivities. Failure to identify these requirements in the proposal phase will cause
unexpected costs to occur throughout the life of the program and may have a nega-
tive impact on the company's reputation for years to come.
2. Contracting. The negotiators for both the contractor and the customer must
have a meeting of the minds on ESD control before the contract is signed. An
imprudent contractor can give away the store with overkill that imposes unneces-
sary, excessive costs on himself or he can underkill and leave himself open to
quality and reliability problems. Contract boilerplate drafted in the vacuum-tube
era is no longer appropriate.
3. Program Management. Program managers can make or break an ESD-

control program. A realistic, cost-effective program will be accepted and supported
only if it is well defined and properly communicated to program management.
Involving program managers in the planning and development phase of the pro-
gram will expedite its future implementation in all functional areas.
4. Purchasing. Buyers of ESD-protective materials and equipment must have

a degree of expertise in every facet of the ESD-control program for the purpose of
estimating the cost of coordinating the requirements with suppliers and subcontrac-
tors. The supplier and subcontractor ESD-control program is generally the responsi-
bility of the prime contractor by Department of Defense mandate. The purchasing
function should be established as the key interface between the supplier and subcon-
tractor ESD-control program and the program of the prime contractor.
S. Logistics. All documentation, including technical publications such as

service and repair manuals, and lists of spare parts, must be in agreement with the
ESD-control program. The field service organization must train its engineers and
technicians in a wide variety of ESD-control methods. Field environments can be
radically different from that of a factory and may require a much higher level of
expertise.
6. Facilities and Industrial Engineering (F &IE). Personnel in F&IE will

be called upon to provide facilities and equipment to establish and maintain ESD-
controlled areas. The ability to debunk the myths about humidification,
ionization, and the like, is a must for all F&IE personnel. Facilitization of an
ESD-protected area can be one of the more costly elements of the program and
therefore should be specified and designed by people well-trained in electrostatics.
7. Plant Services. Maintenance personnel will be required to service and re-

pair some rather unusual equipment. Electricians must understand the special
grounding requirements of ESD control and the need to suppress fields around
some motors, transformers, and switching equipment. Janitors need to know about
application of static-limiting floor finish, the hazards of nonconductive trash
bags, and so forth. All personnel working in ESD-controlled areas must be aware
of the threat of ESD to company products and profits.
8. Engineering. Engineering will determine and specify the V-zap levels (see
"HBM" in Chapter 4) of the product in every configuration from component
through module, subsystem, and system levels. Testing procedures, failure analy-
sis, and designing to meet customer ESD requirements are only the tip of the ice-
berg. The engineer who takes the easy way out when specifying ESD characteristics
can send production costs out of sight. Few engineers have had formal training in
electrostatics, and even fewer have experienced the unpredictable behavior of elec-
trostatic phenomena, e.g., tribolelectric charging, at assembly stations. There-
fore, educating engineers in ESD is an absolute must. Even then, standards,
handbooks, and specifications will not be understood in the same way by all engi-
neers, and the burden of specification interpretation will fall on the shoulders of
Engineering management.
9. Manufacturing. Manufacturing management must make a total commit-

ment to the ESD-control program. A basic system of ESD controls in the factory is
required in the absence of detailed drawings or specification call-outs and will
become a way of life in the shops. Since ESD-control rules can be circumvented
with no tangible evidence and possibly with no harm done, the effectiveness of the
program depends on the integrity of the workers. This unique characteristic of
ESD control makes program maintenance of the utmost importance, and this main-
tenance in the factory might be described as a continuous selling job.
10. Quality Assurance (QA). QA, with its many subset activities, will inter-
face with every part of the ESD-control program from procurement to final test
and customer acceptance. The ESD training of inspectors and first-line manage-
ment should be more extensive than that of manufacturing operators. As in manu-
facturing, the integrity of personnel is primary to success of ESD control.
Program development should be a team effort of experts representing each of

the ten functions noted above. The plan of action should deal not only with
implementation but with long-term maintenance of the program. The electronics
industry is changing daily, and the ESD-control program must change with it. The
need for minor changes will be obvious only to persons close to the program, and
they must ensure that it grows as required to maintain high yields of product and
prevent latent failures.
In conclusion, an ESD-control program is unlike anything else experienced in
the electronics industry because of the latent failure aspect: the product cannot be
routinely inspected to determine if the program is effective. Thus the program
depends on good practice, not testing, and requires integrity by everyone from top
to bottom. It is expensive and not easy to justify. If the program is realistic and
clearly presented to the people who can make or break it, it will be successful.
Also, it must be cost-effective, which is the key feature of the approach we are
outlining in this book. Unnecessary overprotection or overkill tends to result in an
unprofitable project-but remember that inexpensive overkill, as with foillami-
nate bags as Faraday cages, is all to the good. No matter how effective in theory,
the program will receive multilateral support only if it is comfortable to imple-
ment. That is, it must be rational, understandable, and practical, and this is the kind
of program we are attempting to sketch for you.
The Enforcement Problem

ESD control is simple in principle-just follow the Basic Rule and some corollar-
ies such as our CD Rule-but a program requires endless judgments in practice.
Some of these concern enforcement. Can rules ever be waived? If so, when?
For example, can the Faraday cage definition ever be violated? Consider a
metal-cased electronic box which is itself a Faraday Cage except for the connec-
tors (Example 30 in Chapter 8). Should the connectors be covered with metal caps
or with aluminum foil to complete the Faraday cage as "electrically continuous"
by our definition? Or should the box without connector caps be packaged in a
Faraday cage? By the absolute or "Ten-Commandments" approach, such a pack-
age would be foil or better or at least meet the discharge test of paragraph 4.3.3 of
Model Specification 2 in Chapter 9. However, in Example 30 of Chapter 8, we
applied the "situation-ethics" approach in the form of a special test which wasn't
the specified discharge test. Thus we bent the rules and allowed a see-through
metallized bag as packaging. Incidentally, another special test might have involved
measuring the attenuation of a worst-case expected field by a metal plate with a
hole representing the connector opening beneath which the pins are set; high at-
tenuation would justify low shielding by the package. The point is that judicious
rule-bending is allowable and even necessary for a cost-effective program. In the
above example we were avoiding the expense of opening and closing an opaque
package to verify the code numbers on the box.
But what about the enforcement of rules such as operator disciplines that may
not be safely bent? In a milieu where engineers and even assemblers are individualistic,
shortcuts will be taken without consulting authority. Besides willfulness, the reason
for such disobedience is twofold: (1) the need for the rule is not understood and (2)
compliance is onerous. Therefore, part of the remedy for noncompliance is thorough

training of all personnel so that they grasp the concepts underlying all rules. Another
part of the remedy is to design rules which are comfortable, and operators' opinions
must be considered seriously in this design process. Indeed, this is part of the Total
Quality Management (TQM) approach. For example, metal expansion bands for
wrist straps serve their purpose of providing good skin contact only if they are worn
correctly. In practice, our operators complained that these bands pinched their skin
and pulled hairs, so some individuals stretched their bands to deform them until they
hung loose, used oversized bands or, in extreme cases, worked without a wrist strap
when the coast was clear. (This was in the days before continuous wrist-strap moni-
tors.) We should have evaluated various band designs in limited numbers and then
ordered the Velcro-fastening cloth or adjustable stretch-cloth bands which our op-
erators preferred and would have worn willingly.
In conclusion, bend rules only when there is strong incentive as well as justifi-
cation by analysis or special test. Get workers' opinions, and if a rule is consid-
ered impractical, try to replace it. If a rule is troublesome to follow but there is no
alternative, explain its purpose to all personnel involved and insist that they com-
ply with it unless they can suggest an alternative. An assembler or technician who
is taught the reasons for disciplines and is allowed to help select or change them is
likely to be cooperative. As in politics, autocracy breeds rebellion. What is wanted
is a measure of democracy but not so much that the management chain of com-
mand is disrupted. Also, ESD control should be included among company sugges-
tion programs giving certificates and awards.
Advice to Small Companies

Small companies with limited ESD-control budgets don't have enough money to
buy their way out of ESD problems, but that approach fails anyway. When funds
are low, our motto "Buy the basics, forget the frills" is especially apropos. Here's
an example of two shopping lists for equipment and materials for simple assembly
work where triboelectric charging, as caused by processes such as grit-blasting,
doesn't necessitate the use of ionization:
List A
NEEDED FOR EACH OPERATOR COST, $ (ILLUSTRATIVE ONLY)
Conductive floor mal 25

Conductive footwear 25
Bench-model air-ionizing blower 400
Conductive chair 100
Antistatic smock 50 (including cleaning cost
for a year)
Ordinary one-conduclor wrist strap 20
Total: 620
List B
NEEDED FOR EACH OPERATOR COST, $ (ILLUSTRATIVE ONLY)
Static-limiting floor finish 20 (excess in cost over

standard finish for a year)
Continuous wrist-strap monitor unit, 100
including a special strap and
two-conductor cord)
Topical antistat 10
Field meter 300
Zapflash 25
Total: 455
What does list A buy in ESD safety over list B? List A includes a conductive
mat and footwear as a backup to the unmonitored wrist strap, but dirt on the mat
and/or shoe soles can cause resistance to ground to exceed the ARTG, and the
operator might lift both feet and be completely ungrounded. Use of a continuous
wrist-strap monitor is the only positive method of enforcing the Basic Rule,
which does not allow the ARTG to be exceeded "even for a nanosecond."
The personal field meter in List B puts the operator in command of the FIM
danger and makes an antistatic smock and a conductive chair unnecessary. Sleeves
will be rolled up, and if necessary, as shown by apparent charge, the ordinary
chair upholstery could be treated with topical antistat to meet the CD Rule. List
B gives the means to meet the requirements of Model Specification 1 in Chapter 9.
Furthermore, list B includes static-limiting floor finish as a low-cost, unobtru-
sive backup or "safety net." ("Do it right, but keep the safety net tight.") For ex-
ample, if a supervisor with ordinary footwear entered the SSW and touched an
ESDS lead, damage would be far less likely with the special floor finish than with
a conductive mat, which tends to produce high charges on rubber-soled shoes by
triboelectric charging.
The ionizer in list A has no clearly defined purpose and is needless in this case;
see "Ionization" in Chapter 4.
Finally, list B includes a versatile Zapflash for checking grounding and, among
other things, proving that black plastic items are conductive (highly loaded with
graphitic carbon) and not merely pigmented black (with about 1% of nonconduc-
tive carbon).
In conclusion, list B is far superior to A but costs 27% less! This is a good
example of getting more for your money with basics instead of frills.
An even better example is seen in Lists C and D, based on the real-life case of
Corporation X, Case 2 (see specification tailoring in Chapter 11). Costs are in
1994 dollars. List C was being contemplated, and our consultation reduced the
expense by 54% by trimming superfluous or excessive equipment in favor of con-
tinuous wrist-strap monitors and some important but economical items in List D.
The principle of cost-effectiveness illustrated here is built into this entire book,
which we have written with the needs of small companies in mind. But remember
that an optimum choice of materials and equipment is the minor part of the battle
against ESD. The major part is the training and encouraging of operators. For ex-
ample, that hypothetical $300 for a field meter is well spent only if the meter is used
properly! See "Field Meter" in Chapter 4. Thus, more important than "Buy the
basics, forget the frills" is another motto: "Work smart." A small company may have
an advantage here, because there are fewer people to be educated and perhaps some-
thing of a patriarchal system that encourages personnel to identify themselves with
the company and its policies and to be craftsmen rather than bored, alienated hirelings.
Incidentally, the work surface is most economically a static.dissipative table
mat or, if solvent- and solder-resistance is required, a hard laminate retrofitted to a
standard (but grounded) workbench. Much money can be saved by converting
LlstC
APPROXIMATE NUMBER APPROXIMATE
EQUIPMENT COST, S/UNIT OF UNITS TOTAL COST, $
Bench-top 600 30 18,000

ionizer
Conductive 100 76 7,600
floor mat
footwear
chair
Smocks 100 40 4,000
(additional) (including cleaning
for one year)
Total: 41.000
List 0
APPROXIMATE NUMBER APPROXIMATE
EQUIPMENT COST, $/UNIT OF UNITS TOTAL COST, $
Continuous 140 76 10,600

wrist-strap
monitor
Static meter 300 10 3,000
Zapflash 30 7 200
Bench-top 600 5 3,000
ionizer
NFPA99 50 3 150
(or ASTM F 150)
electrode set
screens for
oscilloscopes
and CRTs
Total: 19,000
existing benches this way rather than buying new ESD-control benches as a big
company would tend to do.
A final example of cost-saving on equipment is the use of unglazed cardboard
or hard vulcanized fiber tote boxes. Assuming that tote boxes are allowed to go in
and out of SSWs, the boxes must be SSPs (Faraday cages) if unprotected ESDS
items are placed in them but need be only antistatic if the items are safely inside
Faraday-cage packages such as foil laminate bags. In the latter case, unglazed
cardboard (paragraph 3.6.6.1 of Model Specification I, Chapter 9) or hard vulca-
nized fiber (Paper No.7 in the Appendix) is satisfactory. As of early 1989, ap-
proximate prices for tote boxes 11 x 17 inches by 6 inches deep in a quantity of
100 were:
MATERIAL PRICE EACH. $
Corrugated cardboard (antistatic by nature) 1.60

Nonconductlve high-density polyethylene 4.50
Hard vulcanized fiber (antistatic by nature) 5.60
Antistatic polypropylene 5.60
Black conductive (carbon-loaded) polyolefin 11.50
The cardboard-foil Corshield box described in Paper No.7 in the Appendix is

potentially very economical but couldn't be included in the above comparisons
because it wasn't routinely available in a suitable stock size.
The point is that small companies can cut costs, with no sacrifice in program
effectiveness, by the shrewd selection of ESD-control products. The best needn't
cost the most.
The Role of the Program Coordinator

In large companies, an ESD program coordinator for each division should be ap-
pointed to manage the program and be accountable for compliance with MIL-
STD-1686 (a good ruling document while it lasts (see Chapter 11) even if no
contracts are military). The coordinator should:
1. Serve as leader of a working group made up of representatives from each

functional organization required by Table 1 of MIL-STD-1686 to participate in
ESD control.
2. Conduct monthly meetings of the working group to (1) exchange technical
information, (2) present educational/training seminars, and (3) maintain a high
degree of ESD awareness.
3. Establish various ESD-awareness projects to motivate workers to be diligent
in maintaining disciplines.
4. Respond to all audit deficiency notices on ESD-control matters.
PROGRAM ORGANIZATION AND IMPLEMENTAnON 143
5. Conduct informal surveys to identify potential problem areas and to assist in

remedial action.
6. Maintain a list of approved ESD-control materials and equipment (Refer-
ence Document 4 in Chapter 9).
7. Evaluate new ESD-control materials and equipment for possible addition to
the approved list (using the criteria of Reference Document 5 in Chapter 9).
8. Support new business proposals by providing ESD cost impact data.
9. Support Procurement Quality Assurance (PQA) in the qualification of sup-
pliers and subcontractors to assure comprehensive ESD control from the device
level to the completed system.
10. Review company and government standards and specifications which might
impact ESD control.
11. Provide reports and status briefings as required by upper management.
12. Participate in ESD Association symposia to keep the company program
up-to-date.
13. Review employee suggestions on ESD-related matters. This is important in
winning workers' cooperation.
14. Coordinate with the materials and processes laboratory on special tests in
support of ESD-control disciplines. This includes troubleshooting (Chapter 8).
15. Support the development and revision of operator work instructions. (Work
instructions give detailed how-to methods of meeting specification requirements.)
16. Support Quality Assurance (QA) in audits and reviews as required by the
company's ESD-control program specifications.
17. Participate in design reviews of ESDS equipment as required by customer
contracts.
We mention safety last but, of course, it is not least. Be sure to rigidly enforce the
safety provisions of Model Specification 1, including use of ground fault circuit
interruptors (GFCls) if necessary, so that personnel cannot be electrocuted by con-
nection to ground as a result of ESD precautions; accidents can happen, of course,
whenever a hot lead and a hard ground can be touched simultaneously. Remember
to provide springs on the continuous wrist-strap monitors for plugging in the
wrist-strap cords, as required in Reference Document 3 in Chapter 9, so that the
operator can pull free from any direction. Discourage the use of conductive mate-
rials in favor of antistatic or static-dissipative ones that cannot carry a lethal
current (above 1.0 rnA). If conductive floors are used-which we don't recom-
mend unless absolutely necessary-beware of low resistance to ground that might
allow a person with wet, leather-soled shoes to be electrocuted; we found an ex-
ample of this condition (Reference Document 1 in Chapter 9).
In conclusion, when you design a safe system, consider every possible path to
ground and every worst-case scenario, e.g., wet leather-soled shoes on a conductive
floor as mentioned above. If I-megohm resistors are installed in every path to ground
and GFCIs are used when high voltage is being handled, electrocution as a result
ofESD control is virtually impossible-but this is a big "if' that requires vigilance
and enforcement. Therefore, QA and the company Safety function must be deeply
involved in the ESD-control program.
Finally, we must mention that many marginal safety questions will arise. Ion-
ization, for example, is considered somewhat dangerous when polonium-210 is
the ion source; see "Ionization" in Chapter 4. Many would call this concern far-
fetched, and an even more unlikely danger is the low voltage (0.1-5.0 V, Refer-
ence Document 3 in Chapter 9) imposed on the wearer of a wrist strap with a
continuous wrist-strap monitor; in this case, the voltage is not only low but it
is imposed across two skin-contact points on the wrist so that current will flow
only between these and cannot pass through vital organs. Concern over such
matters may seem ridiculous, but some nonengineers are easily alarmed by any
new "hazard" in the workplace, and Safety personnel must take their concerns
seriously and provide reassurance. Genuine hazards, such as those noted in
Government-Industry Data Exchange Program (GIDEP) Alerts, must be promptly
removed.
Chapter 13
Disposition of Mishandled Hardware
This is a vital subject when expensive products, e.g., modules worth $100,000
each, might have to be scrapped because of ESD damage that is measured or even
only suspected.
The general definition of "mishandling" includes overstress by dropping, bump-
ing, temperature excursions (heating or cooling), vibration, or other mechanical or
environmental events, but we're concerned here only with ESD.
It is assumed that the attending employee, or another employee who observes
the ESD overstress, will take personal responsibility and promptly report the inci-
dent to a Quality Assurance (QA) representative, who will fill out the appropriate
paperwork ("squawk sheet"). Then a Reliability Engineer will review the data and
make a judgment. These data are:
1. Identification including serial number of mishandled ESDS item.

2. Violated specification requirement(s) with paragraph number(s).
3. Date and time of day of the infraction.
4. Relative humidity and temperature.
5. Nature and arrangement of nearby objects during the infraction, including
distances of charged surfaces from the mishandled item.
6. Apparent charges on these charged surfaces measured as soon as possible
after the infraction.
7. Surface resistivity of workbench top, floor, and surfaces of nearby objects.
8. In a reenactment of the infraction, apparent charges on objects and voltage
on the operator.
9. Identification of parts of the ESDS item, e.g., terminals or circuit lines, which
were touched by the operator during the infraction.
10. Functional test data on the item after (and, if possible, before) mishandling.
These are the basic data to be recorded in the squawk. They mayor may not be
sufficient for Reliability, with advice from ESD specialists in other departments,
to make a decision on disposition of the item.
145
In the case of the module in Example 25 of Chapter 8, functional test showed

that all 83 CMOS devices on the board had been damaged by electrical overstress
because of the field generated by grit-blasting. These components made up 51 %
of all component leads, so rework was judged impractical and the module was
scrapped. This is an example of analysis; no testing beyond the functional test was
required.
On the other hand, all components of the mishandled ESDS item might pass the
functional test, or a limited number might fail so that rework could be feasible.
Now the question is: Might those components which passed the functional test
have been subtly damaged by the ESD event so that latent failure is possible? The
next step is to run more thorough tests, e.g., with a curve-tracer, to look for damage
not seen in the functional test. Probably those components showing even subtle
damage should be replaced, but the decision will depend on the reliability level of
the product, rework or replacement costs, etc. Finally, the most difficult question
must be asked: Though parts were not damaged measurably, even though thor-
oughly tested, might they still have been subtly "wounded" by ESD? A judgment
that they were not so damaged might be based on analysis. For example, if the
measured voltage on the operator and the apparent charges on all charged objects
in the vicinity were less than 300 V, and the ESDS item was sensitive to 500 V, the
item could be assumed to be undamaged. But suppose the voltage on the operator
reached 2000 V, as measured by a Personnel Voltage Tester Model PVT-300 by
Voyager (Paper No.1 0 in the Appendix), and he had touched a sensitive lead. Then
the item might have to be scrapped even though no damage had been found by
curve-tracer or any other test method. Of course, considerations such as the reli-
ability level of the item would enter the decision as mentioned above.
On the other hand, suppose that the operator was grounded but accidentally
dropped a film of common nonconductive polyethylene at 2000 V apparent charge
onto the 500-V-sensitive item. Now the picture in terms of the 2000 V is unclear
because charges on nonconductors are indirectly and incompletely manifested on
ESDS items via fields (the FIM). In this case, a special test, specifically a test
using a coupon like those in Paper No. 8 in the Appendix, would be useful. No
damage to five MOSFETs in a row by dropping similar polyethylene film at 2000
V apparent charge onto the coupon would give confidence that the real ESDS
item hadn't been subtly damaged. Or a Static Event Detector™ from 3M could be
used to represent a device. The necessary assumption, is, of course, that lack of
measurable damage to the very sensitive MOSFETs or detector means lack of
unmeasurable damage, which could lead to latent failure, to the less-sensitive
ESDS item; the item in this example is sensitive to 500 V, whereas the MOSFET
is sensitive to 100 V and its sensitivity is heightened still more by the large anten-
nas and lower voltage suppression of the coupon.
A background which is relatively benign in terms of ESD supports arguments
to save mishandled hardware. Therefore, safety nets such as humidification,
ionization, or static-limiting floor finish are desirable. But remember that
DlsPosmON OF MISHANDLED HARDWARE 147
humidification, and especially ionization, must be used with caution; see the
discussion of these subjects in Chapter 4. For example, the reenactment of an in-
fraction in which the operator touching ESDS leads was accidentally ungrounded
might show 100 V on his skin with static-limiting floor finish versus 800 V with
standard floor finish. Then, if the mishandled item was sensitive to 500 V, damage
by DI from the operator's fingers could be discounted for the static-limiting floor
finish but not the standard one; the safety net proved useful.
Obviously, judgments often will have to be based on incomplete data, just as
business decisions are made, but safety nets and worst-case special tests, e.g.,
with coupons as described above, will help greatly. The more critical and high-
reliability the ESDS item, the less leniently should infractions be treated. Remem-
ber that latent failures are real though they may be rare (see "Latent Failure" in
Chapter 4). Ethics forbids the use of sophistry to excuse any and every infraction
not causing measurable damage, and some mishandled hardware will have to be
scrapped even though it passes all its functional tests.
Chapter 14
Checklists, What to Buy and Do.

Conclusion
Checklists, General
Failure Mechanisms (MIL-HDBK-263)

Voltage-dependent:
dielectric breakdown
gaseous arc discharge
surface breakdown
Power-dependent:
thermal secondary breakdown
metallization melt
bulk breakdown
Damage Mechanisms (Chapter 4)

Discharge type:
DI (imperfectly modeled by 8BM and including MM)
CDM (involves DI)
Field type:
FIM
FFB
Greatest ESD Hazards

Personnel (DI by touching by fingers, MM by tool)
CDM especially by leads touching conductors
Fields from dynamic processes such as grit-blasting or spray-coating (FIM)
148
CHECKLISTS, WHAT TO BUY AND DO. CONCLUSION 149
Fields from charged nonconductors such as common plastics (FIM)

Fields from packing material such as polystyrene foam "peanuts" (FIM)
Key Elements of Our Approach

Basic Rule
SSZ concept
CO Rule
Safety nets such as static-limiting floor finish
Operator disciplines
Troubleshooting with analysis or worst-case special tests
Cost-effective materials and equipment
Techniques such as voltage suppression
Conservative ARTG (10 megohms, maximum)
Frequent use of field meter, Zapflash, topical antistat
Minimal SSW (Chapters 7 and 9)

Static-dissipative table mat
Lead wire with resistor
Distribution wire with clamp to ground
Field meter
Continuous wrist-strap monitor with strap (cord and band)
Topical antistat
Static-limiting floor finish
Zapflash
Well-trained operator!
SSP Requirements (Chapters 6 and 9)

Adequate shielding, preferably by Faraday cage
Intimate wrap preferably antistatic
Secure mounting of ESOS item in package
Cushioning by antistatic bubble-wrap or resilient foam
Compliance with triboelectric charging test (if necessary)
Compliance with special test with MOSFETs or Static Event Detector™
(for highest confidence)
Freedom from contamination or corrosion of packaged item
Documents for Reference

MIL-STD-1686
MIL-HDBK-263
MIL-HDBK-773
ESD Controls Study, Final Report, prepared for NASA by Reliability Analysis
Center, Rome Air Development Center, September 1981
NAVSEA SE 003-AA-TRN-OtO: Electrostatic Discharge Training Manual
Mil-STD-129: Marking for Shipment and Storage
EIA 541 : Packaging Material Standards for ESD-Sensitive Items (was Interim Standard
5-A or 5-A-l) (order from Electronic Industries Association, Washington, DC)
NFPA 99: Standard for Health Care Facilities (order from National Fire Protection
Association, Quincy, MA) (formerly NFPA 56A; defines electrode for measur-
ing resistance of surfaces to ground)
ASTM F 150 (defines same electrode as NFPA 99)
DoD-SID-2000-lB (soldering) (replaced by MIL-STD-2000 in Jamuary 1989)
MIL-M-385to (microcircuits)
MIL-B-81705 (packaging)
FED-SID-tO 1 (test procedures)
ESD Association Standards (Chapter to)
Checklists, What to Buy

Workbenches with antistatic or static-dissipative surface and buried conductive
layer (for economy and portability, use a static-dissipative table mat)
Wires and resistors for grounding
Continuous wrist-strap monitors with wrist strap (for every operator)
Humidification equipment (if justified)
Ionization equipment (if clearly required)
Antistatic smocks (optional)
Conductive or static-dissipative shoes (optional)
Faraday-cage bags, boxes, and tote boxes
Conductive chairs (optional)
Antistatic bubble-wrap
Topical antistat
Shunting bars or foam
Miscellaneous items (scrub brushes, screens for CRTs, approved desiccator
cabinets, etc.)
Test Equipment
Field meter (preferably one for every operator)
Zapflash (one for every operator)
Surface resistivity meter (Voyager SRM-l1 0 or SRMlRTG, which also measures
resistance to ground, or equivalent) (one for every ESD-protected area)
Multimeter
Megohmmeter and NFPA 99 (or ASTM F 150) electrode for measuring resistance
to ground of surfaces (or use Voyager SRMlRTG Meter)
Faraday cup
Personnel Voltage Tester (Voyager PVT-300 or equivalent)
Discharge test apparatus, including high-voltage power supply
MOSFETs and curve-tracer
Static Event Detector™
Charged plate monitor (for testing ionizers)
Topaka slip test equipment (for static-limiting floor finish)
Gloss meter (for the floor finish)
Samples and ramp for roller test for triboelectric charging (see Paper No.6 in the
Appendix)
Miscellaneous equipment for special tests
Checklists, What to Do
Major Tools for Troubleshooting

Analysis
CD Rule
Coupon
Discharge test
Field meter
Grounding
Ionization
Shielding
Special test
Standard test
Topical antistat
Voltage suppression
Operator Disciplines
Maintain SSZ and know its boundaries at all times.
Don't touch ESDS leads unnecessarily.
Handle parts by cases when size permits
Don't touch ESDS items to unapproved surfaces such as clothing.
Don't fidget or shuffle feet unnecessarily.
Don't touch ESDS leads while rising from chair or lifting foot.
Wear wrist strap with continuous wrist-strap monitor.
Keep hair less than 2 inches from skin and tie back as required.
Don't remove shorting clips until ready to use item.
Exclude nonconductors from SSW.
Keep SSW clean and orderly.

Roll up sleeves or wear an antistatic (ESD-protective) smock.
Follow CO Rule using field meter to monitor SSW for common plastics and
other field sources.
Ground all conductors.
Use topical antistat as needed.
Closely supervise transients in the SSW.
Use of Field Meters

Meter must be calibrated.
Meter must be grounded if it has a ground connection; otherwise, operator must be
grounded.
Meter must be zeroed, e.g., on grounded operator's hand.
An ungrounded operator with an ungrounded meter may roughly estimate the
voltage on himself by reading a grounded surface.
The apparent charge on a surface must be read according to the meter supplier's
instructions, e.g., "Multiply scale reading by 2 when sensor is 12 inches from
surface being measured."
Readings must not be construed as actual voltage, especially of conductors.
Readings must be used only as apparent charge in determining safe working
distance by the CD Rule.
Mottoes
Buy the basics, forget the frills.
Do it right, but keep the safety net tight.
Humidify if you can justify; ionize only when it's wise.
ESD-wax the floor; it doesn't cost much more.
People are the prime ESD problem-and the prime solution.
Work smart.
Don't look for trouble. (Don't touch leads unnecessarily, etc.)
A grounded operator is the lesser of two evils versus an ungrounded operator.
(Because of the COM.)
No charge, no discharge. (Preventing COM.)
When in doubt, try it out. (Run special tests.)
Setting Up a Program
Reasons.
Required by government contract.
Saves money by lowering reject rates.
Prevents latent failures.
Overview.
Use realistic, systematic approach.
Gather information:
V-zap levels (by (HBM).
Best control methods.
Needed levels of personnel training.
Existing policies supportive of program?
New policies needed.
Special personnel skills needed.
Arrange information in meaningful format.
What? When? and How?

"What" determined from functional profile.
"When"-start with most vulnerable areas.
"How"--controversial; buy products to clearly implement the Basic Rule.
Factors in Cost-Effectiveness.
Lowest level of ESO sensitivity.
Mission criticality.
Cost of ESDS components.
Cost and feasibility of rework.
Schedule impact of ESO failures.
Willingness and ability of workers.
Assumptions about Our Sample Program.

Lowest level of sensitivity: 100 V.
Mission criticality: high.
Cost of ESDS items: moderate to high.
Cost of rework: high.
Feasibility of rework: variable.
Schedule impact of rework: manageable.
Worker skill: good (but may require union cooperation).
Key Operator Disciplines.

Touch ESDS devices only when necessary.
Avoid leads and handle by case if possible.
Touch leads only to approved, grounded materials.
Avoid unknown fields; use CD Rule.
Outside SSW. keep ESDS items in an SSP or closely supervise them.
Dangerous Myths.
"Wrist straps render skin harmless."
"High humidity solves all ESO handling problems."
"Components are safe when mounted on boards."
"Carbon-loaded plastic provides a Faraday cage."

"Ionization allows use of plain, nonconductive plastics."
"Ionization is a substitute for grounding (e.g., by wrist straps)."
"Devices with protective networks are not ESDS."
(See Paper No. 11 for myths.)
Elements of Plan of Action.

Proposals
Contracting
Program management
Purchasing
Logistics
Facilities and Industrial Engineering (F&IE)
Plant Services
Engineering
Manufacturing
Quality Assurance (QA)
Enforcement Problem.
Bend lesser rules while still meeting Basic Rule.
Justify rule-bending by analysis or special test.
Teach personnel reasons for rules.
Make rules "comfortable."
Get workers' suggestions.
Use democracy but not too much.
Advice to Small Companies.

Buy the basics, forget the frills.
Work smart: train operators well.
Follow our approach.
Role of the Program Coordinator.

Lead working group.
Conduct monthly meetings.
Establish ESD-awareness projects.
Respond to audit deficiency notices.
Conduct informal surveys to find problem areas.
Maintain list of approved materials and equipment.
Evaluate new materials and equipment to add to list.
Support new business proposals with ESD cost impact data.
Support Product Quality Assurance (PQA) in qualification of suppliers and sub-
contractors.
Review standards and specifications.
Brief upper management on ESD issues.
Attend ESD Association symposia.

Review employee suggestions.
Coordinate with materials and processes laboratories function on troubleshooting.
Help originate and revise work instructions.
Support QA in audits and reviews.
Participate in design reviews of ESDS equipment.
Safety.
Use GFCIs when necessary (high voltage).
Install springs on continuous wrist.strap monitors.
Discourage conductive materials in the SSW.
Beware of conductive floors.
Consider all possible paths to ground and worst-case scenarios for low-resistance
grounding of people.
Be sure the Safety function is deeply involved in the ESD-control program.
Deal with employees' concerns about remote safety hazards of new ESD-control
equipment.
Conclusion
What are we guarding against? Fields and discharges. The SSP fully protects
against both. When out of the SSP, and ESDS item is either in transit under con-
tinuous operator supervision or else is in an SSW.
In the SSW, where operator disciplines are vital, fields are controlled by ground·
ing conductors, by excluding unnecessary nonconductors, and by controlling nec-
essary nonconductors with the CD Rule and topical antistat. The antistat makes
surface resistivity low enough for incipient static charges to bleed off; also, lubric-
ity of the antistat layer on the surface limits the triboelectric charging propensity.
To control static-generating processes such as grit-blasting, the techniques are ground-
ing, voltage suppression, and, if special tests demonstrate the need for it, ionization.
Discharges from people are avoided by using continuous wrist-strap moni·
tors to meet the ARTG and keep skin near ground potential along with other
grounded surfaces. Also, isolated conductors could participate in discharges if
charged by fields, so fields from electrical machinery, CRTs, ionizers, etc., must
be controlled. Of course, there shouldn't be any isolated conductors, but human
fallibility requires redundant (overlapping) precautions.
Backups to the above primary precautions or defenses include static-limiting
floor finish, certain operator disciplines such as prohibitions against touching,
and possibly humidification. We do not recommend ionization as a general backup;
it should be used only for specific purposes when other methods fail. (But, as of
1995, ionization is becoming more popular as a backup; see Chapter 4.)
Overall, the strategy is to be cost-effective by keeping fields weak rather than
eliminating them and by substituting astute operator disciplines for expensive
materials and equipment. In fact, the operator's performance is critical no matter

how excellent the facilities.
We've called this book an "approach" and provided ample concepts, techniques,
and general advice. The closest we could come to a handbook or set of instructions
is the content of Model Specifications 1 and 2 in Chapter 9, but these will have to
be much altered by the user. The reason is that setting up an ESD-control program
is far from as simple as assembling a piece of lawn furniture (and even that may
not be easy). Each program is a special case because of variables such as:
1. Sensitivity classes of ESDS items.

2. Reliability level and mission criticality of the items.
3. Commercial versus military requirements.
4. Nature of facilities (available space, ambient humidity, etc.).
5. Unique ESD-control problem situations.
6. Nature of processes: labor-intensive versus automated.
7. Skill level and attitude of personnel.
8. Complexity of the organization (large versus small companies).
9. Extent of management support of program.
10. Funding available for program.
Given these variables, we can't be more explicit than we've been. Besides, no
matter how much detail we gave, we could only tell you what to do technically, not
contractually or "politically." You must absorb our advice and then cast it in your
own mold, arriving at specifications extrapolated from our models.
This book may serve as a "consultant's bible" because an ingenious ESD-control
specialist could review a customer's specific needs and adapt our approach to them.
In other words, we've provided the cloth to be tailored to fit any program.
In conclusion, our approach is cost-effective yet conservative. By means of strict
operator disciplines and concepts such as the SSZ and the CD Rule, we forgo
purchases of frills such as conductive chairs which might otherwise be bought
"just in case." But we're conservative in the sense of insisting on protection with
high confidence by such means as Faraday-cage foil packages (or at least pack-
ages passing a discharge test), continuous wrist-strap monitors, and a relatively
low ARTG. We are also wary of dangers such as space charging by ionization,
corrosion by humidification, or contamination by antistats or particles or fibers
from conductive bags, boxes, wrist bands, etc. "Buy the basics, forget the frills."
The money saved on frills such as an unjustified room ionization system can
procure not only essentials, e.g., continuous wrist-strap monitors, but highly
recommended options such as a one-to-one ratio of field meters to operators.
Besides a cost-effective program, we also stress troubleshooting to deal with
any situation using the A-to-Z real and conceptual tools in Chapter 4. In summary,
what we're propounding is a logical, commonsense, economical approach based
on data from meaningful, realistic tests-as opposed to unsupported intuition and
data from misleading tests contrived by tendentious suppliers.
CHECKUSTS, WHAT TO BUY AND DO, CONCLUSION 157
We repeat that we are defending against fields and discharges, and every re-
quirement in the model specifications of Chapter 9 has a purpose in this "war," By
way of a summary, Table 14-1, which is a variation on Table 4-1 in Chapter 4,
relates some defenses to hazards and damage mechanisms; Table 14-2 distinguishes
between primary and secondary defenses. (Note: These tables are simplified for
clarity; not all defenses are listed.) Operators as well as engineers should fully
understand these two tables. "Do it right, but keep the safety net tight."
A book on industrial accidents makes the important point that the immediate
cause, e.g., a spark which ignites a flammable vapor mixed with air, is just the last
of a series of causes such as a broken valve letting gas leak, the failure of a detector
to note this leaking gas, etc. By the time the final cause occurred, the accident was
inevitable. Therefore, the question is not "What was the cause?" but "What can be
changed to prevent another similar accident?" Regarding ESD, the immediate cause
of damage by DI might be a spark from an operator's finger, but the underlying
causes could include failure to require continuous wrist-strap monitors, inad-
equate training of personnel, and the use of ordinary floor wax instead of static-
limiting floor finish. Thus, the chain of causes of ESD damage can lead back
Table 14-1. Relation of Defenses to Hazards

and Damage Mechanisms.
HAZARD DAMAGE MECHANISMS DEFENSES
Fields FlM,FFB CO Rule (apparent charge,

field meter)
Ionization (when clearly
needed)
Shielding
Topical antlstat
Voltage suppression
Discharges 01 (imperfectly modeled Less than 50 V on conductors
by HBM for people). ARTG
COM Continuous wrist·strap
monitor
Grounding
Shunting
Zapnash
Static·limiting Door finish
Both fields and FlM, FFB, 01, COM Basic Rule
discharges SSZ
SSW
SSP
Humidification (when
justified)
Approved ESD-control
materials and equipment,
e.g .• surface resistivity meter
Operator disciplines
Table 14-2. Primary and Secondary Defenses.

PRIMARY SECONDARY DEFENSES
CD Rule Humldlncatlon (when justified)

Approved ESO-control materials Exclusion of unnecessary nonconductors
and equipment from whole ESO-protected area as well as
Operator disciplines fromSSWs*
Ionization (when clearly needed) Certain operator disciplines such as not
Continuous wrist-strap monitor unnecessarily touching ESDS device leads
Grounding ("Don't look for trouble. ")
• Not part of our nonna! procedures but could be done.
through the failure of primary defenses as well as lack of secondary defenses (safety
nets) which make the ESD climate more forgiving. When an aerialist falls to his
death, the ftrst cause might be a frayed rope on the flying trapeze while the second
cause might be a tom or missing safety net.
A capsule summary of this book is:
This book is not a conventional text but a cost-effective approach to ESO control based
on extensive data in the authors' twenty published papers reprinted in the Appendix.
Equipment expense is minimized by superior handling techniques which let you "buy
the basics, forget the frills." Chapters include real and conceptual tools from A to Z,
damage mechanisms, troubleshooting (special tests), model specifications, industry stan-
dards, program management, and lists of what to buy and do.
Chapter 15
The Future
Changes in ESD-control materials, equipment, and standard as well as in the ESDS

items themselves are of course inevitable. The following is a quick sketch of some
possibilities.
Materials
As we said earlier in this book, the EIA's single classification "dissipative" (at
least 105 but less than 10 12 ohms/square) is sensible, but we continue to use the
antistatic and static-dissipative categories in this chapter because they appear
in all the old literature on ESD control. Also, the distinction is meaningful in
terms of the CDM; antistatic surfaces are safer for charged devices than are
static-dissipative surfaces.
Antistatic packaging materials which depend on antistats that create a surface
sweat layer have problems with the fugitivity of the antistat, causing imperma-
nence of the antistatic property (Paper No. 2 in the Appendix) and sometimes
corrosion and contamination (Paper No.9 in the Appendix). Another deficiency is
the humidity dependence of the sweat layer, which involves moisture attracted
from the air. Therefore, hoped-for improvements in antistatic or static-dissipative
materials include permanence and effectiveness at very low relative humidity.
In fact, permanently antistatic or static-dissipative plastics have been created.
For example, "static dissipation originates within the polymer molecule rather than
being added" for a new static-dissipative ABS alloy (Ref. 15-1). Various plastics
are made permanently antistatic with polymeric additives (Paper No. 20 in the
Appendix). Also, a bag made of cellophane has the unusual properties of true
permanence and no antistat rub-off (because it contains no antistat) as well as
biodegradability. This material is listed in paragraph 3.3 .1.k of Model Specification
2 in Chapter 9.
159
Conductive plastic materials are filled not only with graphitic carbon but
stainless-steel or other conductive fibers, and there are proprietary compositions
whose conductive additives are unrevealed. Many new products can be expected
in this area, a prime application being tote boxes. (But note that we discourage
conductive surfaces in the SSW; see Paper No.7 in the Appendix.)
Coatings are another active field. Early antistatic or static-dissipative ones we
have tested either blistered or shifted to nonconductive after a I-day water soak,
representing tote boxes being washed or left outdoors in the rain (which has hap-
pened!). Static-dissipative packaging film with a radiation-crosslinked coating
(Ref. 15-2) is noncontaminating and essentially humidity-independent but has the
drawback of becoming nonconductive after brief exposure to water or various
organic solvents; we found this in our work, and the same result was described in
a NASA Materials Science Laboratory report issued in April 1988. Attempts have
been made by the manufacturer to increase solvent resistance, and current ver-
sions must be evaluated. Polyvinyl chloride film with a relatively heavy radiation-
cured coating has been found to resist solvent extraction of the antistat.
In general, materials of the future will be noncontaminating, permanent by an
oven test, e.g., at 160°F, solvent-resistant, and humidity-independent. The user
will be able to pick any surface resistivity he wants. However, triboelectric charg-
ing may be a slight problem. The fugitivity of old-fashioned antistats has been
beneficial in controlling charging by the lubricity effect, in which a microscopic
amount of liquid antistat rubs off onto the other surface and liquid separates from
liquid to cause minimal tearing loose of electrons. Without this purposeful con-
tamination, charging of packaged items will tend to be higher. But whether the
loss of lubricity for new, permanently antistatic packaging materials will make
much practical difference remains to be seen. A compromise between contamina-
tion and triboelectric charging is to minimize the amount of liquid antistat on the
surface, as has been done in the new generation of MIL-B-81705, Type II, films
mentioned in Paper No.9 in the Appendix.
For the future of packaging materials, see Paper No. 20 in the Appendix.
Equipment
Equipment will evolve; for example, ionization will become safer in regard to
space charging (causing charging of ungrounded conductors) and more fully au-
tomatic in operation. Various test equipment will appear as called for in new stan-
dards and specifications. We hope that less expensive versions of popular items
such as field meters and surface resistivity meters will appear on the market so
that these essential tools can be more generously distributed to personnel and
preferably given to every operator.
Among new items of equipment are rigid ESD-protective boxes and cases that
are reusable and recyclable (Ref. 15-3) and a low-cost spring-like shunting device
to protect device leads or the edge contacts of a circuit-board module (Ref. 15-4).
THE FUTURE 161
Standards
New standards are evolving. Unfortunately, some of them may be unrealistic, as
discussed for the shielding/discharge test for bags in Chapter 6. The astute con-
sumer will take salesmen's claims cum grano salis and devise his own special
tests. Consider the case of static-limiting floor finish. If the objective is to limit
charges on personnel wearing their customary shoes, then the "bottom-line" test is
the walk test (Paper No.6 in the Appendix) or measurements with a Personnel
Voltage Tester (Paper No. to in the Appendix), whereas surface resistivity is
useful only insofar as it correlates with the voltage on walking people-and this
correlation holds only for a given floor finish, not for different finishes being com-
pared. The slip test (Paper No.6 in the Appendix) is also a critical bottom-line test,
though unconnected with ESD, because a slippery floor can bring lawsuits. Let's
forget irrelevant tests and concentrate on those which reflect reality.
Among the newer standards are the new verion of Test Method 3015.6 in Notice
7 to MIL-STD-883C (February 1988), EIA 541, MIL-STD-1686 (replacing DoD-
STD-1686), and MIL-HDBK-773. The last will "change the ways things are done
in-plant" (Ref. 15-5); for example, an interesting requirement in paragraph 5 of
this document discourages the use of conductive materials because they can ac-
cept spark discharges.
Also, ESD Association Standards are appearing rapidly (Chapter to).
ESDS Items
Several years ago, there was speculation that protective circuitry might greatly
reduce the burden of ESD control by leaving few ESDS devices sensitive to less
than 2000 V (by the HBM) when installed. However, at the same time, device
geometries were shrinking to less than 0.5 microns, putting some devices into the
12-volt category. On balance, ESD remains a major threat, only partly alleviated
by protective circuitry, and ESD control is difficult indeed for new components
sensitive to well under the old MOSFET level of 100 V.
For one thing, a wrist strap grounded through 5 megohms will not protect a
12-volt device from an operator stroking a garment sleeve (represented by a plastic
film in our test) on a table in such a way that he would generate 3800 V ungrounded;
see the curve for Aclar in Fig. 1 of Paper No. lOin the Appendix. To protect such
a device, operator movement would have to be restricted and a low ARTG, per-
haps 1 megohm, required. Also, antistatic garments, conductive chairs, and
humidification (if justified) might be needed.
What about fields? Making the assumptions of Paper No.3 in the Appendix, the
equation for the CD Rule for a device of any HBM sensitivity level becomes:
d =(100/S)o.s (V)°.s/1.8 =(5.556) (V)°.s/(S)o.s

where d = minimum safe distance in inches from a charged surface, S = sensi-

tivity of the device in volts, and V = apparent charge on the surface in volts.
ForS=12V,
d = (5.556) (V)0.s/02)o.s = (1.60) (V)0.s.
Solutions to the above equation for 12-volt-sensitive devices are:
V, VOLTS d, INCHES d, FEET
1,000 41 4.2
5,000 113 9.4
10,000 160 13
20,000 226 19
Of course, the coeffieient, which we calculated to be 1.60, in the above equation

would be verified by tests with actual 12-volt-sensitive devices,just as we did with
MOSFETs in Paper No.3 in the Appendix,
Fields on the devices themselves, or on the boards holding them, will present a
great CDM menace as discussed in Chapter 5. Antistatic finger cots may be needed
to protect leads from operators' too conductive skin.
To protect against DI, the maximum allowable voltage on conductors would
have to be about 6 V when handling 12-volt-sensitive devices. This provision would
require careful grounding.
In general, the SSW for highly sensitive items will be analogous to a Class 10
cleanroom: a place where extreme, almost fanatical precautions are taken. How-
ever, these precautions can be extrapolated from the approach described in this
book. ESD control will still involve "buying the basics, forgetting the frills," but
more basics will have to be bought. ESD-protective smocks, for example, may
become a "basic" instead of a "frill" because the fields on ordinary clothing may
not be tolerated for 12-volt-sensitive items as they now are for l00-volt-sensitive
items.
A challenging future lies ahead for ESD control.
References
2-1. J. M. Crowley, Fundamentals of Applied Electrostatics, Krieger/Wiley, 1991.

2-2. J. A. Cross, Electrostatics: Principles, Problems and Applications, AlP/Hilger, 1987.
2-3. E. WhitWker, A History of the Theories ofAether and Electricity, Volumes I and n, Dover, 1990.
2-4. W. A. Harrison, Electronic Structure and the Properties of Solids, Dover, 1989.
2-5. Electrostatic Discharge (ESD) Protection Test Handbook, Second Edition, KeyTek Instrument
Corporation, 1986, page 7.
2-6. Mll..-HDBK-263A, paragraph 3.24, Table m of Appendix A.
2-7. Understanding Solid-State Electronics, Second Edition, Radio Shack (Texas Instruments, Inc.),
1972, page 5.
2-8. Electrostatic Discharge Training Manual, NAVSEA SE 003-AA-TRN-OIO, Naval Sea Sys-
tems Command, 1980, pages 31-41.
2-9. S. Wong, "A Piezoelectric Crystal Electrostatic Discharge Generator," Conference Proceedings
of the EMCIESD International, Anaheim, CA, April 12-15, 1994.
2-10. N. Jonassen, "Do Gases Charge?" Proceedings of the 16th Annual EOSIESD Symposium, Las
Vegas, NV, September 27-29,1994.
3-1. 0.1. McAteer, Electrostatic Discharge Control, McGraw-Hill, 1989.
3-2. DoD-HDBK-263, ESD Control Handbook for Protection of Electrical and Electronic Parts,
Assemblies. and Equipment (Excluding Electrically Initiated Explosive Devices), Table IV,
pages 9-11.
3-3. J. M. Kolyer, R. Rushworth, and W. E. Anderson, "ESD Control in an Automated Process,"
EOS/ESD Symposium Proceedings, EOS-9, 1987, pages 41-50.
3-4. J. M. Kolyer, D. E. Watson, and W. E. Anderson, "Controlling Voltage on Personnel," EOS/ESD
Symposium Proceedings, EOS-ll, 1989, pages 23-31.
3-5. 1. M. Kolyer and D. E. Watson, ESD from A to Z: Electrostatic Discharge Control for Electron-
ics, Van Nostrand Reinhold, 1990, page 40.
3-6. Ibid., pages 11, 160, and 161.
3-7. Ibid., pages 193-201.
3-8. T. Dangelmayer, ESD Program Management, Van Nostrand Reinhold, 1990.
3-9. Kolyer and Watson, op. cit., pages 3-27, 39-56.
4-1. Comments by J. M. Kolyer in "ShOUld Testing Be Done Beyond the Lab?" edited by Donald
Ford, EOSIESD Technology Magazine, August/September 1987, pages 15 and 22.
4-2. Electrostatic Discharge (ESD) Protection Test Handbook, KeyTek Instrument Corp., 1983,
pages 10, 12, 13, 16, and 18.
4-3. W. Simmons and P. Adamosky, "Meeting the New 883 Test Method of IC Static Discharge
Testing," Evaluation Engineering Magazine, July 1988, page 82.
4-4. D. E. Frank, "ESD Considerations for Electronic Manufacturing," presented to American
Society of Manufacturing Engineers Westec Conference, Los Angeles, CA, March 21-24,
1983 (Douglas Paper 7324).
4-5. R. Moss, "Exploding the Humidity Half-Truth and Other Dangerous Myths," EOSIESD
Technology Magazine, April 1987, page 10.
163
4-6. P. S. Neelakantaswamy and R. I. Turkrnan, "ESD Failures of Board-Mounted Devices," Elec-

tronic Packaging and Production Magazine, February 1987, page 132.
4-7. D. M. Yenni and 1. R. Huntsman, "The Deficiencies in Military Specification M1L-B-8170S:
Considerations and a Simple Model for Static Protection," presented at the Reliability Analysis
Center EOS/ESD Symposium, Denver, CO, 1979.
4-8. J. O. Lonborg, "Static Survey Meters," EOSIESD Symposium Proceedings, EOS-S, 1983,
page 63.
4-9. S. S. Sullivan and D. D. Underwood, "The Automobile Environment: Its Effects on the Human
Body ESD Model," EOSIESD Symposium Proceedings, EOS-7, 1985, page 103.
4-10. D. L. Lin, M. S. Strauss, and T. L. Welsher, "Big Problem Uncovered: Zapper Data Differs,"
EOSIESD Technology Magazine, August/September 1987, page 9.
4-11. R. E. McAteer, G. H. Lucas, and A. McDonald, "A Pragmatic Approach to ESD Problem Solv-
ing in the Manufacturing Environment, a Case History," EOSIESD Symposium Proceedings,
EOS-3, 1981, page 34.
4-12. "Nuclear Air Ionizers Recalled by NRC," Evaluation Engineering Magazine, March 1988,
page 65.
4-13. W. R. Van Pelt, "Polonium-21O Contamination," Chemical and Engineering News, April 11,
1988, page 4.
4-14. N.10nassen, "The Physics of Electrostatics," distributed at the Sixth Annual EOS/ESD Sympo-
sium, Philadelphia, PA, 1984.
4-IS. G. Baumgartner, "Electrostatic Measurement for Process Control," EOSIESD Symposium Pro-
ceedings, EOS-6, 1984, page 2S.
4-16. "Latent ESD Failures: a Reality," Evaluation Engineering Magazine, April 1982, page 80.
4-17. G. T. Dangelmayer, "ESO-How Often Does It Happen?" EOSIESD Symposium Proceedings,
EOS-S, 1983, page I.
4-18. B. Rodgers and W. Tan, "Attacking ESD Where It Lives," Circuits Assembly Magazine, 1une
I99S, page 40.
4-19. B. N. Stevens, "Determining the Surface Resistivity ofESD Protective Cellular Packaging Ma-
terials," EOSIESD Symposium Proceedings, EOS-8, 1986, page 136.
5-1. T. S. Speakman, "A Model for the Failure of Bipolar Silicon Integrated Circuits Subjected to
Electrostatic Discharge," International Reliability Physics Symposium Proceedings, 1974.
5-2. O. J. McAteer, Electrostatic Discharge Control, McGraw-Hill, 1989, page 173.
S-3. D. C. Anderson, "New Approach to Handling Charged Devices Without Causing ESD Dam-
age," Technical Record of the Expo '92 International Conference on Electromagnetic Compat-
ibility, sponsored by EMC Technology Magazine, Reston, VA, May 18-22, 1992, page 174.
6-1. GIDEP Alert H6-A-83-02, "Materials, Plastic, Antistatic," December 27, 1983. (Antistats from
bags contaminated NASA instrument mirrors.)
6-2. G. C. Holmes, P. J. Hubb, and R. L.10hnson, "An Experimental Study of the ESD Screening
Effectiveness of Antistatic Bags," EOSIESD Symposium Proceedings, EOS-6, 1984, page 78.
6-3. S. A. Halperin, "Selecting the Proper Protective Bag: Part 11," EOSIESD Technology Magazine,
October/November 1988, page IS.
6-4. MIL-HDBK-773, "Electrostatic Discharge Protective Packaging," 1 April 1988.
7-1. N.l. Safeer and 1. R. Mileham, "A Material Evaluation Program for Decorative Static Control
Table Top Laminates," EOSIESD Symposium Proceedings, EOS-6, 1984, page 85.
7-2. 1. R. Huntsman and D. M. Yenni, "Charge Drainage vs. Voltage Suppression by Static Control
Table Tops," Evaluation Engineering Magazine, March 1982.
7-3. E. H. Russell, "Safely Grounding Static-Control Work Surfaces," EOSIESD Technology Maga-
zine,1une 1987, page 10.
7-4. R. Kallman, "Comments on Coping," Evaluation Engineering Magazine, August 1988,
page 108.
8-1. "Arm Hair Pin-Pointed as New Hazard," Evaluation Engineering Magazine, May 1984,
page 70.
REFERENCES 165
8-2. Scott's Standard Methods of Chemical Analysis, 6th Edition, Volume I, page 334.
8-3. NAV SEA SE 003-AA-TRN-0IO, "Electrostatic Discharge Training Manual," pages 35 and 36.
8-4. R. D. Anderson, "Alert Error Corrected: How Sweet It Is!" EOSIESD Technology Magazine,
OctoberlNovember 1987, page 8.
8-5. "Designing for Compliance: Immunity to ESD," Application Note 106, Special Supplement,
Compliance Engineering Magazine, 1991.
9-1. N. B. Fuqua and R. C. Walker, "ESD Controls Study, Final Report," prepared for NASA by the
Reliability Analysis Center, Rome Air Development Center, September 1981, page 30.
10-1. S. A. Halperin, "Anything But Static: The EOSIESD Association Takes Charge," Compliance
Engineering Magazine, Winter 1991, pages 13-28.
10-2. S. Weitz, "New Trends in ESDTest Methods," EMC Test and Design Magazine, February 1993,
pages 22-26.
10-3. 1. C. Hoigaard, "ISO 9000 Promotes Automated Continuous Monitoring," Evaluation Engi-
neering Magazine, February 1995, pages 105-108.
12-1. K.1essen and 1. Barto, "Static Control Team Concept-Implementation of ESD Protection in
Manufacturing," Evaluation Engineering Magazine, November/December 1983, page 94.
12-2. 1. R. Giuliano, "SD Program Nets Large Financial Gains," Evaluation Engineering Magazine,
December 1986, page 26.
15-1. "New Plastics Harvest Adds Variety," Machine Design Magazine, 1uly 21,1988, page 12.
15-2. A. H. Keough, "Antistatic Resin Composition," U.S. Patent 4,623,594 (November 18, 1986).
15-3. P. O'Shea, "Totes/Bins/packaging Go Environmental," Evaluation Engineering Magazine, March
1995, pages 92-95.
15-4. D. Cronin, "Cro-Bar: A New Technique for ESD Protection," EMC Test and Design Magazine,
1anuary 1993.
15-5. "MIL-HDBK-773: Sighs of Relief," EOSIESD Technology Magazine, OctoberlNovember 1988,
page 7.
Appendix
Some of the following papers have been condensed to highlight conclusions and
save space, but the majority have been reproduced in their entirety.
Paper No. 1 was reprinted with permission of lIT Research Institute/Reliability
Analysis Center. Paper No.5 was reprinted with permission of Evaluation Engineering
Magazine. Paper No.7 was reprinted with permission of EOSIESD Technology
Magazine. Papers No.2, 3, 4, 6, 8, 9, and 10 were reprinted with permission of the
EOS/ESD Association. Papers No. 11-20 were reprinted with permission as noted
in each case.
For a quick overview, leaf through and read the abstracts.
167
PaperNo. 1
Presented at the 3rd Annual Electrical Overstress

Electrostatic Discharge Symposium, Las Vegas, Nevada,
September 22-24,1981, sponsored by lIT Research Institute, EOSIESD
Symnposium Proceedings, EOS-3, 1981, page 75.
SELECTION OF PACKAGING MATERIALS FOR ELECTROSTATIC

DISCHARGE-SENSITIVE (ESDS) ITEMS
John M. Kolyer and William E. Anderson
Rockwell International Corporation

Autonetics Strategic Systems Division
Electronics Operations
3370 Miraloma Avenue
Anaheim, CA 92803
Note: The following is a condensed version of this paper. See the Symposium
Proceedings for the complete text.
Abstract
Seven materials were evaluated. Properties such as transparency, puncture resistance, and
heat seal strength were measured. Also, ESDS devices in bags were subjected to the exter-
nal static field of a model "human finger" charged to 25,000 volts as well as to the extreme
condition of a high-voltage continuous discharge. The assumption was that if the packaged
devices survived ESD overstressing with no measurable damage they would not be subtly
damaged under actual handling/transit conditions. Such damage could reduce lifetime and
reliability. For the highest confidence in protection from ESD, at an acceptable cost and
with military specification coverage, a double-bagging system was adopted: antistatic poly-
ethylene, which is MIL-B-81705, Type II, as the intimate wrap with a foil bag, MIL-B-
81705, Type I, as the exterior "Faraday cage." Another foil bag, proposed as MIL-B-81705,
Type III, is effective without a Type II inner bag. A brief review is given of the context of
static control plant procedures in which the protective bags are used.
Introduction
The selection of packaging materials for ESDS items is difficult because many products are
on the market and the claims made by rival manufacturers are often questionable and even
contradictory.
The early products included antistatic polyethylene or "pink poly," which is MIL-B-
81705, Type II; conductive (carbon-loaded) polyethylene; and a laminate incorporating alu-
minum foil, which is MIL-B-81705, Type I. Later, variations on "pink poly" and foil bags
169
170 APPENDIX
were developed; these are "blue poly" and a foil laminate bag proposed as MIL-B-81705,
Type III, and so abbreviated in quotes as "Type III." Also, a new concept in "static-protec-
tive" bags appeared: polyester film with an exterior coating of vacuum-deposited nickel,
protected by a thin lacquer coat, and an antistatic polyethylene lining; this construction has
the advantage of partial transparency. The proponents of the latter bag made the industry
aware of the need for "Faraday cage" protection from external static fields/discharges and
the insufficiency of "pink poly" alone for worst-case handling conditions. An aluminum-
coated, partially-transparent bag is now competitive with the original, nickel-coated ver-
sion. In 1980, Department of Defense documents were issued, DOD-STD-1686 and
DOD-HDBK-263, which emphasized that the interior of a bag must be incapable of
triboelectric charging, as true of "pink poly," while the exterior should be a conductive
"Faraday cage" when external static fields might be encountered.
Our evaluation included all seven materials mentioned above. The objective was high
confidence in protection from external fields/discharges, at an acceptable cost, even for
items of the greatest ESD sensitivity. The bags were for general use: they were not for clean
room applications, handling of open devices, or the packaging of items sensitive to trace
contamination. Functional properties such as heat sealability were checked, and packaged
MOSFETs were exposed to external static overstresses to provide assurance of freedom
from damage by realistic stresses. The chosen material(s), of course, had to be cost-effective.
Experimental Methods
Figures I through 4 show some of the experimental methods. Figures 5 and 6 are a typical
example of ESD damage.
Materials Tested
The seven materials are described in Table I. All were received as samples in the form of
bags. "Pink poly" is MIL-B-81705, Type II; by "MIL-B-81705" we are referring, through-
out this paper, to the current Revision B. Two leading brands of carbon-loaded polyethylene
were tested; the volume resistivities, which were calculated from surface resistivity mea-
surements with bar-clampled samples at low voltage, were 235 and 875 ohm-em. Using
bar-clamped samples and a megohmeter at 10 volts, surface resistivities were in the ex-
pected range for the antistatic lining of the nickel-coated material, 2 x 1010 ohms/square,
and the aluminum-coated material, 7 x 1011 ohms/square. "Pink poly" and "blue poly" gave,
respectively, 8 x 1010 and I x 1011 ohms/square. The black lining of the Type I bag gave an
apparent 8 x 107 ohms/square, but edge effects were possible. "Type III" has a "pink poly"
lining for which 2 x 1011 ohms/square was found. Surface resistivity measurements are
neither closely reproducible nor a critical test for antistatic properties; static bleed-off time
(Federal Standard 101, Method 4046) is better for single materials but not for the Type I
laminate, whose foil layer drains the charge in terms of perceived voltage.
Conclusions on Bag Materials

Table I summarizes ratings for the materials in our various tests, notes military specifica-
tion coverage, and gives approximate prices. Blank spaces mean that no testing was done.
PAPER NO.1 171
CAPACITOR
NONCONDUCTING IURFACE
Fig. 1. Charged "finger" test.
-11.aoo OR -.aoo V
co. 1. 5. OR • SECONDS)
.......
NONCOND_ ....UCTING SURFACE
Fig. 2. Tesla coil test.

172 APPENDIX
. . . .ITIIIIMCI
, .. XI., INCH ICY MIlD

HORN ANTINNA
• • KW PEAK lOWER
• ,. GIGAHERTZ
.,0SECONOS
Fig. 3. Radar exposure.
• MlL.sTD-I1OC, METHOD 114,

PARAGRAPH 4.1.12.2 13 HOURS'
Fig. 4. Vehicular bounce test.
If an ESD shielding rating of "good" and a degree of transparency are required, the
metal-coated bags seem adequate. The aluminum version was more transparent than the
nickel version and performed as well in our other tests. However, potential users should
conduct their own evaluations.
In our case, "excellent" ESD shielding was desired. No sacrifice in confidence in ESD
protection was made in order to gain partial transparency. This position left only two candi-
dates: 'JYpe I and "'JYpe m." These foil bags, incidentally, not only give maximum ESD
protection but have superior sturdiness (abrasion and puncture resistance) and are excellent
moisture barriers. Since 'JYpe I has a questionably antistatic lining, as discussed above, only
PAPER NO.1 173
Fig. 5. Punch-through of gate oxide induced by "finger" (25,000 volts, 167 pF) with
MOSFET in black bag (1200><).
Fig. 6. Punch-through of gate oxide, detail from previous slide (22,OOOX).
''Type III" with its "pink poly" lining remained as a single-bagging candidate. ''Type III" is
not yet covered by a military specification, but approval is said to be pending.
Our MX missile module assembly facility, for which a bagging system was being se-
lected, required cushioning material around assembled modules. The choice was "pink poly"
bubble-pack with skins of Richmond RCAS-1200 (the only material thoroughly tested by
us for antistatic properties). Since this material is not a "Faraday cage," double-bagging
was necessary; a Type I bag was chosen as the ESD shield. Type I was preferred to ''Type
III" because the former has military specification coverage and the latter's "safe" lining
was not required. Also, Type I is less expensive (Table 1).
When cushioning is unnecessary, a MIL-B-81705 system comprises a "pink poly" (Type
II) inner bag and a Type I outer bag; both materials are bought from suppliers on QPL-
81705-6 (issued December 30, 1980). The inner "pink poly" bag provides high at-work-
station transparency; an operator can withdraw the pink bag and easily see its contents. If
174 APPENDIX
Table 1. Summary ofT.ata

WA1ER HEAT ANIl-
ESO 11lANS- VAPOll ABRASION SI!AL I'UNC'IUlB STATIC MIL
MA11!RIAL SHJl!LO PAIU!NCY PEDI. RESIST. S1UNG11I RESIST. PROP. SPEC COST-
Pink poly P B P B B P G Yes 13

Blue poly P G P B B P G No 18
Nickel- G P P P P·B G P No 30
coated
A1uminum- G P P P B G P No 28
coated
Black F X P B P B No 19
Type I B X B B G B P Yes 21
"Type IIr' B x B B G B G No 29
-Cenu per 8 x 10 in. bag. Iota of 2000. S181
Code: E = Excellent
G= Good
F=Fair
P= Poor
x= None
the latter feature is not considered important, the "Type DI" bag is effective but does not yet
have military specification coverage.
The above benefits of double-bagging with Type II inside Type I are bought at a cost
penalty of 13-21% over single-bagging with meta1-coated or '7ype III" bags. Using the
prices in Table I, a Type I bag at 21¢ and a Type II bag at 13¢ together cost 34¢, which is 4,
5, or 6¢ more than the nickel-coated. ''Type III," or aluminum-coated bags, respectively.
The above selections are based on present knowledge, but there are unanswered ques-
tions concerning bag materials. For example, how permanent are their ESD shielding or
antistatic properties? Accelerated aging tests combined with real-time exposures should
give reassuring, or disturbing, answers. Also, new materials may be expected to appear on
the market. The "bag of the future" may have most or all of the following properties: low
cost, >75% transparency, high ESD shielding effectiveness (like Type I or "Type III", per-
manent antistatic and ESD shielding properties, no dependence on a minimum relative hu-
midity, no propensity to contaminate packaged items. minimum sloughing (with no possible
shedding of conductive particles), excellent moisture barrier properties, good heat-sealability,
and, last but not least, cost-effectiveness.
PaperNo. 2
Presented at the 5th Annual Electrical OverstresslElectrostatic Discharge

Symposium, Las Vegas, Nevada, September 27-29,1983, Sponsored by I1T
Research Institute. EOSIESD Symposium Proceedings, EOS-5, 1983, page 87.
PERMANENCE OF THE ANTISTATIC PROPERTY

OF COMMERCIAL ANTISTATIC BAGS
AND TOTE BOXES
lohn M. Kolyer and William E. Anderson

Anaheim, CA 92803
Abstract
The shelf life of antistatic materials has been questioned. Therefore, accelerated life tests
were conducted with several commercial bag materials. One criterion for retention of the
antistatic property was ability of the bag lining to triboelectrically charge NEMA FR-4
glass-reinforced epoxy circuit board material. It was found that light rubbing, as occurs on
withdrawing boards from bags in in-plant handling, gave the same result-a low and pre-
sumably tolerable charge-whether or not antistat was present. Contamination caused higher
charging. Thus, shelf life was judged to be unlimited, by this test, if bags are stored closed
and clean. Another criterion was the charge on the bag itself. The static field is suppressed
by foil or metallization, but surface conductivity is required to control the field for antistatic
polyethylene alone (MIL-B-81705, Type II). For a leading brand of the latter, incorporating
a low-volatility antistat, the shelf life was estimated as at least 8 years if bags are used at
above 20 percent relative humidity. Heavy or repeated rubbing, as may be caused by vibra-
tion in shipment, is another matter. Transfer of antistat from bag lining to surface of the
repeatedly rubbed item seems necessary to prevent high charging. Again, a shelf life of over
8 years was assigned to the tested brand of MIL-B-81705, Type II, but data were insuffi-
cient for a prediction for foil or metallized bags. Antistatic polyethylene tote boxes were
also studied. Used, worn boxes can accept a triboelectric charge. Such boxes might be
periodically treated with topical antistat solution. Conductive boxes are an alternative but
have potential problems such as sloughing of conductive particles or triboelectric charging
of ESD-sensitive items being added or removed.
175
176 APPENDIX
Introduction
The pennanence of antistatic plastic packaging materials is in doubt because large increases
in surface resitivity during a I-year shelf life of films or foam sheets have been reported.'
Increased susceptibility to triboelectric charging by polystyrene foam also was found for
aged antistatic materials. The indication is that the antistat on the surface of the plastic
volatizes into the air or is removed by contact with absorbent materials such as cardboard.
Therefore, minimization of exposure to moving air or paper products was recommended.'
In view of the above, our purpose was to establish the pennanence of certain antistatic
materials, in the fonn of bags, which we use or might use. Of particular interest was the
propensity of bag materials to triboelectrically charge a printed circuit board module on
insertion or withdrawal. Therefore, a static charging test was devised in which a bare glass-
reinforced epoxy board was pulled in a reproducible manner from aged bags and the charges
on the board measured. Also, surface resistivities of aged bags were noted. Accelerated
aging was achieved by exposing bags, both closed and propped open, to 160°F in a forced-
draft oven. Bags were also exposed to blotting paper.
Tote boxes were also of concern because used, worn, antistatic polyethylene boxes found
in the plant were observed to take charges of up to a few kilovolts when stroked with a
nylon cloth. Whether a given charge was really a practical danger was judged by dropping
highly ESD-sensitive devices [metal oxide semiconductor field effect transistors (MOSFETs)]
into a nonantistatic tote box charged to various levels and electrically testing the devices for
degradation.
Bag Materials Tested

Nine bag materials were tested (Table 1). Five of these had been characterized by us 2
years earlier.2 All samples except the foam were received in the fonn of approximately
S x 10 in. bags; the foam was stapled into pouches. The abbreviations in Table I, e.g.,
"black" or "grid," are for convenience.
The "Type I" material previously characterized2 was the Brand B material in Table 1.
"Pink poly" was MIL-B-SI705, Type II, Qualified Products List (QPL) material, specifi-
cally RCAS-1200 from Richmond Corporation. The "black" material was the brand for
which a volume resistivity of 235 ohm-cm had been found. 2This bag, of course, is conduc-
tive rather than antistatic.
Figures I and 2 show the principal experimental methods.
Conclusions
Shelf Life of Antistatic Polyethylene Bags

1. In tenns of charging a popular circuit board material, under the specific conditions of
our test, shelf life is unlimited if the bags are kept closed and clean. Antistatic polyethylene
PAPER NO.2 1n
Table 1.
Type I \0 MIL-B-SI705, Type I (vapor-deposited aluminum +
(brands A and B) spun-bonded polyethylene + aluminum foil +
carbon-loaded polyethylene)
Pink Poly 6 MIL-B-S1705, Type n (antistatic polyethylene)
Foil + Antistatic Poly S Antistatic spun-bonded polyethylene + aluminum foil
+ antistatic polyethylene
Nickel-Coated 2.5 100 Angstroms nickel + 1 mil polyester + 1.5 mils
antistatic polyethylene
Stainless-Steel-Coated 3 Polyester + stainless steel + antistatic polyethylene
Black 4 Carbon-loaded polyethylene (conductive)
Grid 4 Antistatic nylon + barrier film + conductive ink grid
+ barrier film + antistatic copolymer
Antistatic Foam 250 Layered, antistat-treated polypropylene foam
3/4-IN. THICK
WOOD(o.IILI'~ ~
1/2·IN. THICK •• ~
ANTISTATIC FOAM (0.031 LI'~ .. '. " , ...
ANTISTATIC lAG. Ix 10 IN. ~
FR ... EPOXY CIRCUIT IOARD (IARE.
NO COI'I'ERI. e x 12 x O.OIM IN.
"
(1.5 IN. EXTENDED INTO IAGI
TA8LESWITH
MElAMINE.
FORMALDEHYDE
lAMINATE SURFACE
CARRIAGE (CARRIAGE PLUS CIRCUIT
BOARD· 1.1 LII
~8~'ilE' WOOD AND FOAM WERE

LIV lAG WITH EPOXY BOARD
INSIDE. BOARD WAS HELD IY FINGER
PRESSURE ON lAG AT POINT X. THEN
FINGER PRESSURE WAS REDUCED TO
LET WEIGHT FALL. AND CHARGE ON
BOARD WAS MEASURED.
Fig. I. Static charging apparatus.

178 APPENDIX
LEADS TO MEGOHMMITE"
Y1:=. ~-
lUIt'ACE OF lAG
NOTE: A U·LI. WEIGHT WAI PLACED
tl1t4N. ALUMINUM
IIlOUNOED EDGI'
ON EACH OF THE ELECTIIODU
Fig. 2. Surface resistivity electrodes.
has a long record of successful use. Also our tote box tests suggest that limited static fields
may be tolerable. Therefore, it is presumed that the relatively low triboelectric charges
detected on the circuit board would not endanger attached devices. In contrast, the high
charges observed when the board was stroked with carbon-loaded polyethylene might well
be damaging. However, voltage suppression would limit the field on a multilayer board
with internal planes of copper.
2. In tenns of surface resistivity, shelf life at usual relative humidities (over 20 percent)
is conservatively estimated as 8 years for 6-mil "pink poly" bags when kept closed. This
relatively good pennanence is probably due to the low vapor pressure of the antistat, which
is extruded into the polymer and gradually bleeds to the surface to fonn a "sweat layer"
with atmospheric moisture. See Fig. 3 and Table 2.
3. Bags should be stored closed to exclude contamination such as dust, kept out of
contact with paper or other absorbent materials, and, of course, never washed with water or
organic solvents.
4. Two extreme situations for using bags are: (I) light, unrepeated rubbing by the pack-
aged item against the bag lining (in-plant handling) and (2) heavy or repeated rubbing due
to vibration (shipment). Antistat depletion is irrelevant for light rubbing, in tenns of our
board-charging test, but may be critical for heavy or repeated rubbing. In the latter case, the
key factor may be neither resistivity nor lubricity but antistat transfer from bag to packaged
item. The role of relative humidity is not clear. Note that the lining of MIL-B-81705, type I,
which is without antistat, has been reported to develop damaging charges on MOSFETs in
a vibration situation.3 Also note that a surface layer of peanut oil, instead of antistat, can
prevent triboelectric charging of DIPs by polyethylene under heavy-rubbing conditions. 4
5. Antistatic polyethylene bags without foil or metallization to control the static field
can develop significant charges on themselves at some combination of antistat depletion
and low moisture content of the "sweat layer." Here, relative humidity could be a critical
factor. Thus our prediction of a long shelf life for "pink poly" (MIL-B-81705, Type II) is
limited to use at >20 percent relative humidity.
PAPER NO.2 179
1I&0Il10" ATUVRI
tOl.L...----,!'---~2---,l-3- - -..!----:!.f----:!.
TIME. MONTHS
Fig. 3. Surface resistivity versus time for aged 6-mil ''pink poly" bags.
ESD Hazard Caused by Static Charges on Tote Boxes

1. Used, worn antistatic polyethylene tote boxes can develop static chuges of a few
kilovolts when stroked with a nylon cloth. The MOSFET tests indicated that up to 4 tv
could be tolerated. but weaker fields could cause ESD damage when "antennas," e.g., cUcuit
lines, are present (see Paper No.3).
2. It is recommended that antistatic tote boxes be treated periodically with topical anti-
stat solution. This is an inexpensive procedure in terms of antistat cost but mayor may not
be practical in view of handling and scheduling complexities.
Table 2. Estimated Shelf LIves for Antistatic Bags

(Used at >20 Percent Relative Humidity).
ESTIMA11!D SHELF LIFE
(CLOSED, CLEAN)
EXTREMBUSE MIL-S-817OS
CONDmON TYPE n FOIL OR METALLIZED
Light rubbing ofFR-4 epoxy/glass 8 years Unlimited

on insertion/withdrawBl (inplant use)
Heavy/repeated rubbing, eg, due to 8 years Not estimated (rate and effect of
vibration (shipment) antistat depletion are unknown)
180 APPENDIX
3. Conductive tote boxes have no permanence problem but may triboelectrically charge
devices being removed from them, slough conductive particles, or cause ESD damage by
rapid discharges as could happen with any conductor.
4. Each used must select a tote box material based on his unique parts-handling situation.
References
I. G. O. Head, "Drastic Losses of Conductivity in Antistatic Plastics," Reliability Analysis Center
EOSIESD Symposium Proceedings, pp. 120-123, Orlando, FL, 1982.
2. J. M. Kolyer and W. E. Anderson, "Selection of Packaging Materials for Electrostatic Dis-
charge-Sensitive Items," Reliability Analysis Center EOSIESD Symposium Proceedings, pp.
75-84, Las Vegas, NY, 1981.
3. D. M. Yenni, Jr., and J. R. Huntsman, "The Deficiencies in Military Specification MIL-B-81705:
Center EOSIESD Symposium, Denver, CO, 1979.
4. J. R. Huntsman and D. M. Yenni, Jr., "Test Methods for Static Control Products," Reliability
Analysis Center EOSIESD Symposium Proceedings, pp. 94-109, Orlando, FL, 1982.
PaperNo. 3

Symposium, Philadelphia, Pennsylvania, October 2-4, 1984, Sponsored by EOSIESD
Association and lIT Research Institute. EOSIESD Symposium Proceedings,
EOS-6, 1984, page 7.
HAZARDS OF STATIC CHARGES AND FIELDS AT THE WORK STATION
John M. Kolyer, William E. Anderson, and Donald E. Watson

Anaheim, CA 92803
Abstract
Tests were conducted to determine under what practical conditions electrostatic discharge-
sensitive (ESDS) items can be damaged at the work station by static charges and fields.
Various materials and equipment were evaluated for possible hazards (e.g., static fields
created by electrical air ionizers) which must be avoided by careful operating techniques.
The data led to important conclusions and guidelines for an ESD control program in accor-
dance with DoD-STD-1686. Examples are given of equipment, materials, and techniques
which complement one another, and basic rules for electronic assemblers are suggested.
Introduction
A charged surface always creates a static field (E-field). However, in discussing "hazards of
charges" we ignore the field and are concerned with discharges between conductors, for ex-
ample from a person's finger to the copper line on a circuit-board module. In discussing
"hazards of fields" we are concemed with ESD damage caused by induction. The effects of
RF (radio-frequency radiation) or of magnetic fields (H-fields) are not considered in this paper.
Following a description of the test methods used, we evaluate charge and field hazards,
including those from electrical equipment, and discuss some methods of coping with them,
Finally we briefly describe a synthesis of materials/equipment and operator techniques for
handling ESDS items in accordance with DoD-STD-1686.
Test Methods
Static Meter Measurements

Measurements were made with a Simco Electrostatic Locator, Type SS-2 (Simco Co.,
Inc.), calibrated with a 12-inch-square aluminum sheet charged to 1,000-10,000 volts by
181
182 APPENDIX
a current-limiting high-voltage dc power supply (TARI Central Control Module, Static

Control Services). The charges reported in this paper are apparent charges; for example, a
4-inch-square aluminum sheet charged to 10,000 volts showed an apparent charge of about
5,000 volts. The apparent charge is a measure of the strength of the field, which can damage
ESDS items by induction.
Surface Resistivity
The electrodes have been illustrated. I A 5-lb weight rested on each electrode, and strips of soft
carbon-filled polyolefin were placed under the blades to improve contact with static-dissipa-
tive or antistatic surfaces. With this arrangement, a steel surface read lOS ohms/square, so lOS
was subtracted as a correction factor. Readings were made with a Beckman Model L-1O
megohmmeter at 500 volts unless otherwise noted. Multimeters were used for lower voltages.
MOSFET Damage Tests

The N-channel metal oxide semiconductor field effect transistor (MOSFET) used in
our tests was a Motorola 2N4351 with the shorting bar removed. The metal TO-72 type
package had four leads (source, gate, drain, and substrate-case). This device contains no
input/output protective diodes or circuitry. Using a Tektronix 576 curve tracer, gate-source
threshold voltages (VOS(11I) were read at Vns = 10 volts and In = 10 ~A in accordance with
MIL-STD-750B, Method 3403. The operator was, of course, grounded with a wrist strap
when making measurements.
For many experiments a "MOSFET board" was used; see Fig. 1. The substrate-case lead
of a MOSFET was clipped to the "antenna" (circuit line) touched by the operator, while the
gate lead was clipped to the circuit line with the projecting lead. The choice of substrate-
case and gate leads, rather than another combination of leads, maximized the oxide layer
area exposed to an ESD. Each circuit line was 4 inches long, and the lead projected 1.7
inches beyond the end of the board. The circuit-line antennas increased the likelihood of
COPPER
CIRCUIT
LINE
6 x 9 x 0.06 INCH
PR-4 GLASS-EPOXY
LAMINATE
Fig. 1. MOSFET board.

PAPER NO.3 183
ESD damage to the MOSFETs by static fields. 2 Similar printed circuit boards with MOSFETs
and circuit lines for antennas have been used in other investigations3•4 but for different
purposes. Our board, held by the operator as shown in Fig. I, was intended as a worst-case
simulation of ESDS modules being handled at the work station.
MOSFETs were often shorted, as noted in the data, but they were also considered damaged
if the current-voltage curve changed and/or VOS(1lI) shifted by more than 0.1 volt (VOS(1ll) read-
ings for undamaged MOSFETs were reproducible to ±O.02 volt). Since latent ESD failures
are a reality,S a VOS(1ll) shift of only 0.1 volt is considered significant. Damage is reported as a
fraction; for example, "2/5" means that two MOSFETs were damaged out of five tested.
Room Ionization System

The power supply was mentioned above. Two emitters (Static Control Services), each with
four corona-discharge points and a set of plastic reflectors, were positioned 3 ft apart on the
ceiling of an open-fronted booth 95 inches high, 71 inches wide, and 47 inches deep. This
booth contained a bench 29 inches high, 60 inches wide, and 30 inches deep. The distance
from the emitter points to the work surface was 59 inches. One emitter was positive and the
other negative.
Shielding/Discharge Test
This test for evaluating packaging materials was conducted with apparatus in accordance
with published reports. 6•7 In brief, a capacitive sensor, comprising two 0.75-inch-diameter
aluminum disks (0.06 inch thick) separated by 0.5 inch of acrylic plastic, was placed inside
a 3 x 3 inch pouch of packaging material clamped between flat aluminum electrodes. See
Fig. 2. In accordance with a published procedure,8 a pulse of 1200 volts was applied from a
200 pF capacitor through a 400 kilohm resistor to the upper electrode, the lower electrode
being grounded. The pulse picked up by the capacitive sensor was delivered to a storage
oscilloscope. Results were conveniently expressed as "percent attenuation;" for example, if
the pulse recorded by the sensor was 300 volts, the percent attenuation was (1200 - 3(0)
(100)/1200 = 75. One edge of the pouch under test was always continous so that the mate-
rial had electrical continuity from side to side.
Modified Shielding/Discharge Test

The capacitive sensor described above was provided with 4-inch leads ending in clips.
MOSFETs, described above, were clipped with the substrate-case lead connected to the
upper capacitor plate and the gate lead connected to the lower plate. The sensor, including
the leads and the MOSFET, was placed inside an 8 x 10 inch pouch of packaging material
resting on a grounded aluminum plate, and the charged probe of an Electro-Metrics Model
EDS-200 discharge unit (Model D-25, 150 ohms, 150 pF) was touched to the upper surface
of the pouch (see Fig. 2). To ensure a worst-case condition, the packaging material was
pulled tight against the sensor by taping down the edges of the pouch.
The probe of the discharge unit was intended to represent a statically-charged person's
finger touching a bag containing ESDS items and lying on a conductive surface. In a few
cases an actual finger was used.
184 APPENDIX
SHIELDIMG/DISCHAIGI TEST
Fig. 2. Diagram of bag-testing procedures.
Hazards of Charges at the Work Station
Bench Tops
In discussing bench tops we are concerned not with the top holding a charge but with its
ability to drain charges from objects placed on it.
Static-dissipative tops (surface resistivity 1()5-109ohms/square) are favorably discussed
in DoD-HDBK-263 and have been suggested as ideal. 9Antistatic tops may drain charges
too slowly, as will be seen below, while conductive tops are an electrical safety hazard and
might damage charged ESDS items by discharging them too rapidly.9.10 (See also DoD-
HDBK-263 and the discussion below.)
A top may have a conductive sublayer without being satisfactory, because without suffi-
cient surface conductivity the charge will be merely voltage-suppressed and not drained."
In our tests, conductive or antistatic tote boxes charged to ±1O,OOO volts drained incom-
pletely (e.g., 8000 volts remained on an antistatic box) after resting 1 minute on an antistatic
top (lOll ohms/square) but drained completely «50 volts remained) on a static-dissipative
top (109 ohms/square). One successful construction for a static-dissipative top is, in our
experience, a high-pressure laminate with linen fabric in the surface layer and conductive
carbon paper (grounded through 1 megohm) as the sublayer.12 The sol vent resistance of this
laminate is reported to be very good,I2 and in our tests 50 daily rinses with methyl ethyl
ketone had no effect on surface resistivity or appearance.
Static-dissipative tops are safer than conductive tops for operators, but can static-dissipative
tops also be safer for devices as suggested above? To answer this question, tests were con-
ducted with the MOSFET board (Fig. 1). An operator was charged to various levels by
PAPER NO.3 185
contacting the dc power supply, and a second after releasing the supply he touched the lead
of the MOSFET board to either a static-dissipative bench top or a conductive work surface
represented by an aluminum sheet. Both surfaces were grounded through 1 megohm.
Results are given in Table 1. Both surfaces allowed damage to occur, but the static dissi-
pative surface was less destructive. That a more resistive surface can be less damaging will
be seen again in Table 2.
In conclusion, an antistatic top will not itself hold a charge, but it drains charges too
slowly from objects placed on it. The best choice seems a static-dissipative top with a bur-
ied conductive layer and a securely mounted grounding lug.
Tote Boxes
Tote box selection is a controversial subject. Antistatic tote boxes can lose their "sweat
layer" and should be periodically treated with an antistat solution. \ Carbon-loaded plastic
boxes, on the other hand, are permanently conductive but can slough conductive carbon
particles. If sloughing is not considered a problem, one question remains: Are carbon-loaded
boxes too conductive? The fear is that a charge on a conductive box might "zap" an ESDS
item, whereas under the same conditions the charge would bleed off harmlessly from an
antistatic box. Or, conversely, a charged ESDS item might discharge too rapidly to the
"sparking surface" of a conductive box. 10,\3
We attempted to settle this question using the MOSFET board (Fig. 1). Either the opera-
tor or the tote box was charged, and the lead of the MOSFET board was touched to the box.
Two boxes were tested: an antistatic box (surface resistivity 2 x 109 ohms/square at 48
percent relative humidity and 72°F) and a black conductive tote box (2.2 x 10" ohms/square
at 32 volts). Both boxes were injection-molded by the same manufacturer and had the same
dimensions (approx. 7 x 10 x 3 inches deep, with a 0.08-inch wall). Results are given in
Tables 2 and 3, which also include an aluminum sheet (12 x 12 x 0.06 inch) as a reference.
Table 2 also includes a static-shielding bag, discussed below. When the operator was charged,
the box or sheet lay on a static-dissipative bench top grounded through 1 megohm; when
the box or sheet was charged, it stood on insulating stand-offs (plastic breakers) and the
Table 1. MOSFET Damage Caused by Charged Operator

Touching Lead of MOSFET Board (Fig. 1)
to Work Surfaces Grounded through 1 Megohm.
MOSFET DAMAGE
STATIC-DISSIPATIVE
SURFACE ALUMINUM
VOLTAGE (8 x 108 OHMS/SQUARE) SHEET
-100 0/5 1/5

-200 0/5 3/5 (2 shorted)
-300 1/5 (l shorted) 3/5 (l shorted)
-400 0/5 2/5 (0 shorted)
-500 2/5 (l shorted) 3/5 (2 shorted)
-1000 2/5 (1 shorted) 5/5 (4 shorted)
Totals: 5/30 (3 shorted) 17/30 (9 shorted)
186 APPENDIX
Table 2. MOSFET Damage Caused by Charged Operator Touching Lead of

MOSFET Board (Fig. 1) to Objects on Static-Dissipative Bench Top
(8 x 10' Ohms/square) Grounded through 1 Megohm.
MOSFET DAMAGE
PAJmALLY-
TRANSPARENT BAG
(II x 14 INCH)
ANTISTATIC CONDUCTIVE ALUMINUM WITH EXTERNAL
VOLTAGE TOTE BOX TOTE BOX SHEET METALLIZATION
-300 0/3 0/3 1/3

-500 0/3 0/3 2/3 (l shorted)
-1,000 0/3 2/3 (l shorted) 2/3 (l shorted) 2/3 (0 shorted)
-2,000 0/3 3/3 (2 shorted) 3/3 (l shorted)
-5,000 0/3 2/3 (l shorted)
-10,000 0/3 3/3 (2 shorted)
Totals: 0/18 10/18 (6 shorted)
operator was grounded through 1 megohm with a wrist strap. The tests were conducted at
68 percent relative humidity and n°F.
The result was that the antistatic tote box was clearly less damaging than the conductive
box when the operator was charged but only slightly less damaging when the box was
charged. An explanation is that when the box was charged (Table 3) induction became the
dominant factor; that is, MOSFETs were damaged by capacitive coupling when the lead of
the MOSFET board penetrated the field of the charged box. On the other hand, when the
operator was charged (Table 2) the controlling factor was the resistance of the surface being
touched by the lead.
The conclusion is that antistatic boxes are less of a discharge hazard than conductive
boxes if handling techniques are poor. With good operator discipline, conductive boxes are
Table 3. MOSFET Damage Caused by Grounded Operator

Touching Lead of MOSFET Board (Fig. 1) to Charged Objects.
MOSFET DAMAGE
ANTISTATIC CONDUCTIVE ALUMINUM

VOLTAGE TOTE BOX TOTE BOX SHEET
-300 0/3 0/3 2/3

(l shorted)
-1,000 0/3 2/3 1/3
(l shorted) (l shorted)
-2,000 2/3 3/3 3/3
(2 shorted) (3 shorted) (3 shorted)
-3,000 2/3 2/3 3/3
(l shorted) (l shorted) (3 shorted)
Totals: 4/12 7/12 9/12
(3 shorted) (5 shorted) (8 shorted)
PAPER NO.3 187
safe and have the advantage of not relying on a fugitive antistatic additive; however, the
user must assure himself that sloughing of conductive particles is not a problem.
In our assembly operations, modules are placed in Faraday-cage (foil-containing lami-
nate) bags, and then the bags are placed in tote boxes. When a tote box is received at a work
station, the operator removes the bag and sets the tote box aside before removing the mod-
ule. Thus, the tote box is not a threat no matter what its electrical properties. Even ordinary
plastic tote boxes could be used in this manner, but good practice calls for selection of
conductive or antistatic materials in an ESD control program. The blanket rule is that ordi-
nary plastics must be excluded from work stations.
Nonconductive Plastics
When a polyethylene film (0.004 inch) at -13,000 volts (apparent charge, as are all charges
reported in this paper) lay flat on grounded aluminum foil, the apparent charge was reduced
to -300 volts because of voltage suppression, and MOSFET damage by a grounded opera-
tor touching the MOSFET board lead to the plastic surface was 0/3. Similarly, for a vinyl
sheet (0.058 inch) at -8000 volts the apparent charge fell to -300 volts and MOSFET dam-
age was 0/5. The nonconductive plastics were unable to deliver their high charges, and the
suppressed fields were too low to cause damage.
The conclusion is that fields, rather than charges per se, are the hazard in this case. Thus
the MOSFET damage reported in Table 4 when charged nonconductors were touched would
presumably have been caused by a close approach without contact.
Table 4. MOSFET Damage Caused by Fields

from Various Charged Objects.
APPARENT
CHARGE, DISTANCE, MOSFET
OBJECT VOLTS INCHES DAMAGE
Butyrate Tote +7000 2 3/3

Box Lid (3 shorted)
6 1/3
(0 shorted)
Vinyl Tote Box on +500 o(touched) 1.3
Bench (0 shorted)
Vinyl Chair +500 o(touched) 1.5
(1 shorted)
Plastic Identification -2000 3 1/4
Badge (0 shorted)
Butyrate Screwdriver +1700 o(touched) 2/3
Handle (0 shorted)
Polyester Lab Coat +1300 o(touched) 2/5
(0 shorted)
Polyethylene Film -10,000 2/2
(1 shorted)
Aluminum Sheet -3300 4 3/3
(1 shorted)
Painted Metal +100 o(touched) 0/3
188 APPENDIX
Operators
An operator's skin should always be grounded. However, a grounded operator can cause
ESD damage, as seen in Tables 3 and 4, and is only the lesser of two evils versus an un-
grounded operator. Good handling techniques are needed to protect ESDS items from
grounded operators when charges on the items are unavoidable. For example, a module
might be triboelectrically charged by spraying with a conformal coating. Then, if a grounded
operator were to touch a contact a damaging discharge could occur as explained in the
literature. 10.14 A solution to this problem in terms of handling techniques is for the operator
to be conscious of the hazard and avoid touching leads or contacts of ESDS items.
A possibility is that an operator's fingers might be reduced from conductive to static-
dissipative or antistatic surfaces by his wearing appropriate gloves, possibly cotton. This
idea is especially applicable to static-producing assembly or rework operations such as grit
blasting.
Hazards of Fields at the Work Station
Field Strength versus Distance

Table 4 gives typical data obtained when a grounded operator waved the MOSFET board
(Fig. 1) in the vicinity of various charged objects. Rapid motion, however, seemed unneces-
sary; when the lead of the MOSFET board was brought within 1 inch of a plastic surface at
-5000 volts at the slow speed of 5 inches/minute, and then withdrawn at the same speed,
MOSFET damage was 1/2. Presumably a potential difference was slowly built up by ca-
pacitive coupling until the oxide layer of the device was ruptured.
Incidentally, it should be noted that a charge of -3000 volts on the MOSFET board itself,
with both the antennas and the operator grounded, caused 2 of 3 MOSFETs to be damaged
(I shorted) by installation and removal from the board. Damage of components during
insertion into highly charged printed circuit boards has been reported. IS
Figure 3 shows MOSFET damage levels caused by a grounded operator (wearing a wrist
strap with a I-megohm resistor) waving the MOSFET board in static fields. The circles are
for various plastic surfaces, and the triangles are for an isolated 12 x 12 x ~ inch aluminum
sheet charged to various negative voltages by the dc power supply. The black data points
represent MOSFET failures (at least 1/3), and the white points represent no failures (0/3).
Assuming a point source, field strength is directly proportional to voltage on the surface
and inversely proportional to the square of the distance from the surface. Therefore, V plot-
whose slope represents field strength. The same result is given by plotting rv
ted against d 2 , where V is voltage and d is distance from the surface, will give a straight line
versus d,
which was done for convenience. Of course, the source of the field is an area rather than a
point, and the data are not precise enough or numerous enough to establish an exact thresh-
equation for this line (except near the origin) is rv

old field strength to cause damage. However, the line in Fig. 3 illustrates the principle. The
= Cd, or d = MC, where C is a
constant (1.8 in this case) and d is the minimum safe distance in inches from a surface at
apparent voltage V. Fo~example, if the charge were 3000 volts, the calculated minimum
safe distance would be 3000/1.8 = 30 inches. Since no damage occurs below 50 volts, the
line does not pass through the origin, and the threshold for damage by nonconductors was
300-500 volts.
PAPER NO.3 189
0 l
V
7000
1
sooo
--• i-.
JIAZAIDOUS SAFE
~
S 3300 .oil.
... ...
.
Q
/
;0.
r.:I
...../ v
~ 2000 -..., ...
~
i
1700
:;
1300
~ 1000 .A
500
~ .. ~ ...
300
100 / ...
~NONCONDUcroRS
50 "/
o 10 20 30 40 so 60
DISTANCE, INCHES
Fig. 3. MOSFET damage caused by static fields.
This estimate of minimum hazardous field strength is, of course, worst-case because the
long antennas of the MOSFET board served to "gather" static fields. At the other extreme
was a "MOSFET module" in which the antennas (insulated wires) were sandwiched be-
tween aluminum ground planes while the projecting lead shown in Fig. 1 was eliminated.
At 1 foot from a surface at -7000 volts, the "MOSFET module" gave MOSFET damage 0/5
versus 4/5 (two shorted) for the MOSFET board. However, the "MOSFET module" did
allow damage (2/3, 0 shorted) at 2 inches from a surface at -7000 volts. Therefore, the
constant C for the "MOSFET module" is more than 7 but less than 42. Real-life module
assembly situations will lie between the two extremes represented by the MOSFET board
and module.
Air Ionizers
A few ionized-air blowers and nozzles (guns) were evaluated for field hazards by waving
the MOSFET board (Fig. 1) at various distances from their tips while they were running.
The operator was grounded through 1 megohm as usual. Table 5 gives the data.
A safe working distance from the blowers or the pulsed dc ionizer was 10 inches. Inci-
dentally, the pulsed de ionizer had the advantage of not blowing air. An airstream is often
190 APPENDIX
Table 5. MOSFET Damage Caused by Fields

from Air-Ionizing Blowers and Nozzles.
DISTANCE FROM
MOSFET BOARD, MOSFET
EQUIPMENT INCHES DAMAGE
Blower Alternating Current, Corona Discharge 'JYpe:

Model A (5000 volts) 2 2/3
(l shorted)
4 0/5
6 0/5
Model B (5000 volts) 4 4/10
(2 shorted)
10 0/10
Ionizer, Pulsed Direct Current, Corona Discharge 10 0/10
Type, 10,000 volts (Balanced)
Ionized Air Nozzles, Corona Discharge Type:
Model A (4000 volts) 2/5
(0 shorted)
2 0/5
Model B (4000 volts) 3/5
(1 shorted)
2 1/3
(0 shorted)
3 0/5
Model C (7000 volts) 1 0/5
Ionized Air Nozzle, Nuclear Type 0/5
objectionable; for example, it causes undesirable cooling during soldering. However, this
fanless ionizer was as effective as the blowers in rapidity of neutralizing positive or nega-
tive charges on plastics in tests at a range of 3 feet.
The nozzles were operated at 30 Ib/in. 2 gage air pressure. Model C was less damaging
than the others; in this model, the corona-discharge point was mounted "piggyback" out of
the airstream. The nuclear type caused no damage because it has no electrical field. Another
advantage of the nuclear type is its compact size. However, its cost may exceed that of
electrical nozzles over a period of years because of an annual leasing fee, and the poisonous
radioactive material (polonium-210) presents a remote safely hazard. For very close work,
on the basis of field hazard as well as bulk, the nuclear-type nozzle is preferred.
The conclusion is that electrical air ionizers do not present a field hazard if the work is
kept 10 inches or more away from them. However, when nozzles must be held very close to
ESDS items, as in cleaning a module by blowing off particles, the nuclear type is required.
Room Ionization System

Reasons for selecting room ionization as opposed to local ionization-or for using ioniza-
tion at all-are beyond the scope of this paper. We are concerned here only with certain
aspects of a specific dc room ionization system.
PAPER NO.3 191
The equipment which we tested is meant to be operated with the positive and negative
emitters at similar voltages so that there is little net field strength or ion imbalance at the
workbench level 5 feet or more beneath the emitters. However, if the system were acciden-
tally unbalanced would an electrically isolated conductor on the bench become sufficiently
charged to damage ESOS items touched to it?
To answer this question, tests were conducted in the booth described above under "Test
Methods" using only one of the two emitters. The bench surface was static-dissipative as
described under "Bench Tops," above. An aluminum sheet (12 x 12 x ~ inch) was isolated
from the bench surface by two different thicknesses of nonconductive plastic (bubble-wrap)
as listed in Table 6. Thus, a capacitor was produced by the aluminum sheet and the carbon-
paper sublayer of the bench; when the two were separated by ~ inch. for example, the
calculated capacitance of the aluminum sheet was 129 pF. The voltage on the aluminum
sheet was measured with an Electrostatic Field Meter Model 970 from Static Control Ser-
vices; in this case the field meter usually employed was not sensitive enough. Table 6 gives
the data.
The system is normally operated at about 8200 volts; a slight increase, to perhaps 9000
volts, may be required at low humidity. Therefore. Table 6 indicates that even if all the
emitters of one polarity should fail completely there will not be dangerous charging of
isolated conductors lying on the bench. In conclusion, worst-case testing showed no hazard.
Having gained confidence that the room ionization system would not be part of the ESO
problem, we tested its efficacy at neutralizing charges on plastics. Table 7 shows that neu-
tralization of charges at the bench level was reasonably rapid for a variety of materials; the
relative humidity during this test was 56 percent at 74°F. Even with the system somewhat
unbalanced (15,000 positive ions/ml, 22,000 negative ions/ml), results were similar. To
check the effect of humidity. the booth was sealed and the air inside it was partially dried
Table 6. MOSFET Damage Caused by an Unbalanced DC

Room Ionization System.
DISTANCE
FROM OBSERVED
ALUMINUM VOLTAGE ON
SHEET TO ALUMINUM MOSFET
EMITTER VOLTAGE- BENCIITOP. IN. SHEET DAMAGE
+17.000 4 +240 2/3

(1 shorted)
As required to charge aluminum sheet 4 +120 1/3
(0 shorted)
As required to charge aluminum sheet 4 +80 1.5
(0 shorted)
As required to charge aluminum sheet 1/4 +120 2/5
(l shorted)
(0 shorted)
+8200 1/4 <10 0/10
-8200 1/4 <10 0/10
*Only one emitter was operating.

192 APPENDIX
Table 7. Charge Decay Rates with or without Room Ionization.

RESIDUAL
TIME, MINlITES, TO VOLTAGE AT
DECAY TO SOO 5 MINUTES
INITIAL VOLTS wrrn wrrnSYSTEM
MATERIAL CHARGE, VOLTS SYSTEM ON' OFF"'
Polystyrene Foam +5000 1.5 +1200

Polyethylene Film -500 1.8 -4800
Vinyl Sheet -5000 2.5 -800
Carbon-Loaded Polyethylene +5000 2.5 +3200
(Conductive)
Aluminum Foil +5000 1.8 +3300
'15,000 positive ions/ml, 13,000 negative ions/ml.

*"<100 positive ions/ml, 500 negative ions/mI.
with a desiccant. At 10-20 percent relative humidity, the decay rates of the materials in
Table 7 remained rapid.
Another question was the effect of position of charged objects under the emitters; it was
feared, for example, that a positively charged material under a positive emitter might re-
ceive relatively few negative ions and so exhibit a slow decay rate. Therefore a test was run
with a positively charged plastic and a negatively charged plastic at 51 percent relative
humidity and 73°F. Table 8 shows that charge decay rates remained far more rapid at any
position with the system on than when it was off. Tests at commercial installations of this
and a competitive system gave better uniformity, presumably because a grid of emitters
covered the ceiling instead of only two emitters being confined to a booth. Of course, per-
fect uniformity of decay rate across the bench is unnecessary.
Table 8. Effect of Position Under Room-Ionization Emitters

on Charge Decay Rates for Plastics.
PERCENT OF INITIAL CHARGE
INITIAL POSmONON
CHARGE, BENCH IO 20
MATERIAL VOLTS SURFACE MINUTES MINUTES MINUTES
Polyethylene -1700 Under Positive 3

Bubble-Wrap Emitter
Under Negative 2
Emitter
Between Emitters IS 9 <I
System Off 88 82 59
Butyrate +15,000 Under Positive 17 11 3

Tote Box Lid Emitter
Under Negative 3 <I
Emitter
Between Emitters 4 <I
System Off 93 87 80
PAPER NO.3 193
It was desired to compare the effectiveness of the room ionization system with that of a
typical bench-model ionized-air blower (as in Table 5). Polyethylene bubble-wrap
triboelectrically charged to -10,000 volts was found to lose half its charge after a I-minute
exposure (at 60 percent relative humidity and 74°F) either to the emitters at ±8200 volts at
bench level or to the blower at a distance of 5 feet.
In conclusion, the room ionization system tested did not present a field or charge hazard
at the bench level and was indeed effective in removing standing charges on nonconductors
such as plastic packaging materials. Fig. 4 (for 49 percent relative humidity and 73°F), is a
typical example of the utility of room ionization in removing a stubborn charge on plastic;
similar curves appear in a published report. 16 Note that the electrical field of a charged item
attracts ions to cause neutralization. Therefore, a thin sheet of plastic which lies flat on a
workbench and has a largely collapsed field will be neutralized only slowly and incom-
pletely. The bubble-wrap in our test was i inch thick and was folded over on itself so that
the measured surface was elevated about I inch.
Assembly Operations
Many static-producing manufacturing operations are listed in DoD-HDBK-263 and a Navy
training manual,I7 but only two will be discussed here: peeling masking tape from a roll and
grit-blasting a module to remove the coating from components for rework.
A conductive masking tape would have virtually no static field because of charge drain-
age and voltage suppression. Our tests showed that a commercial aluminum foil/fiber tape
had sufficient conformability and tear resistance for masking, but its acrylic adhesive re-
leased incompletely after oven-baking of the masked parts at 150°F. Therefore, the only
deficiency of this tape might be overcome by substituting a silicone adhesive, with good
release, for the unsatisfactory acrylic.
A Micro Blaster (Comco, Inc.) was found to create a charge of +700 volts on FR-4
circuit board material when sodium bicarbonate "grit" was sprayed against the surface using
10 Ib/in.2 gage air pressure. By adding a "piggyback" corona-discharge ionizer (Simco PIN
-18
II)
-----
~160 WITHOUT ROOM IONIZATION
I~~
="'-140
--.......
~-12 0 I~
,
~
a- 100
\
1 ---.
i =:
0
0
~ -4 0
\
-2 0
\. ~WITH 100M IONIZATION
I'
10 20 30 40 so 60
TIME. MINUTES
Fig. 4. Charge decay rates for polyethlene bubble-wrap.

194 APPENDIX
4100034) operating at 7000 volts, with the point 0.6 inch from the spray nozzle, the charge
developed was reduced to only +60 volts. This charge is well below the danger level for
nonconductors (Fig. 2).
Operators
A properly grounded operator has no charge on his skin. However, arm hair has been re-
ported to develop up to 900 volts, even at 50 percent relative humidity, when skirt sleeves
are rolled up; at least one major corporation considers this a serious static problem. 18 Note
that in our worst-case tests as little as 500 volts (on vinyl plastic, Table 4) was sufficient to
damage a MOSFET. Obviously, head hair which dangles and might touch ESDS items is
also a hazard.
A synthetic-fiber smock can carry damaging fields, as seen in Table 4, and DoD-HDBK-
263 makes the point that clothing should never touch ESDS items. An "antistatic" smock is
at least a partial solution, but it should not contain stainless-steel fibers which might fall
onto circuit boards and cause shorts. 19 The operator's chair, if nonconductive, is also a
hazard (Table 4).
The general approach to the problem of unavoidable static fields on an operator's hair or
clothing-or on anything in the work station- is for him or her not to bring ESDS items
near charged surfaces unnecessarily. In other words, do not look for trouble. In addition,
head hair can be tied back, smock sleeves rolled up (with the proviso that ESDS items are
kept away from hair on the forearms), etc.
ESD-Protective Equipment and Materials

Our discussion will be limited to two items: wrist straps and bags.
Wrist Straps
Various designs were tested. Bead chains gave intermittent contact but succeeded in pro-
tecting MOSFETs when the operator charged himself by shuffling his feet on a carpet and
touched the lead of the MOSFET board to ground. The main objection to bead chains is
their tendency to ride up over sleeves and lose contact with the skin. Any design which
presses grounded metal snugly against the skin seems adequate; this includes expanding
stainless-steel watchband types or designs in which a metal "wristwatch" element is held
against the skin by an elastic band. However, wrist straps which depend on the conductivity
of carbon-loaded plastic are sUSpect,20 and a conductive fabric band shed steel fibers up to
0.09 inch long in our tests.
Faraday-Cage Bags
It is well recognized that ESD-protective bags must guard against ( I) internal triboelectric
charging and (2) external static fields/discharges. 9
The bags in Tables 9 and 10 all provide some shielding against external fields. In terms
of the pulse attenuation test (Table 9), 93 percent attenuation must be sufficient because the
foil-containing laminate bag is an excellent Faraday cage. 21
PAPER NO.3 195
Table 9. Shleldlng/Dlscharge Test.

PERCENT
PULSE
BA0 1,21 A'ITENUAll0N
Antistatic Polyethylene 75
Carbon-Loaded Polyolefin (Conductive) 97
With Conductive Ink Grid, A 78
With Conductive Ink Grid, B 50
Partially Transparent with External Metallization 99,8
Same, but Handled (Metallization Cracked on Crease) 59
Partially Transparent with "Buried" Metallization, A 39
Partially Transparent with "Buried" Metallization, B 87
Laminate Containing Aluminum Foil, with Antistatic 93
Polyethylene Lining
Antistatic Polyethylene Bags Inside of and Outside of Carbon-Loaded 92
Polyolefin Bag
Antistatic Polyethylene Bag Outside of Partially Transparent Bag with 93
External Metallization
The high values given by a partially transparent, externally metallized bag (99.7 percent)
and a volume-conductive bag (97 percent) are artifacts, in a way, because conductive mate-
rial "shorts out" the upper and lower plates in the test; wrapping these bags in antistatic
polyethylene lowered their pulse attenuation to the 92-93 percent range. By this test it would
seem that carbon-loaded polyolefin is as good a Faraday cage as aluminum foil, but such is
not the case; discharges from a person's finger can damage ESDS items inside conductive
plastic bags. 3,21
A more significant test is the device-in-bag procedure of Table 10. The conditions were
certainly worst-case because a very low resistance (150 ohms) was used; a resistance of
Table 10. Modified Shleldlng/Dlscharge Test.

MOSFET
BAG (AS IN TABLE 9) VOLTAGE DAMAGE
Carbon-Loaded Polyethylene (Conductive) +5000 3/3 (2 shorted)

With Conductive Ink Grid, A +5000 1/3 (l shorted)
Partially Transparent with External Metallization +5000 3/3 (3 shorted)
Partially Transparent with "Buried" Metallization, A +5000 1/3 (1 shorted)
Laminated Containing Aluminum Foil, with Antistatic +5000 0.5
Polyethylene Lining +10,000 0.5
Partially Transparent with External Metallization ** -8000 (est)'" 2.3 (0 shorted)
Same, but Handled**· -8000 (est)· 2/3 (2 shorted)
Partially Transparent with "Buried" Metallization, A -8000 (est)· 3/3 (0 shorted)
*The EDS-200 probe was replaced by the fmger of it person triboelectrically charged by walking on a
carpet at 24% relative humidity and 73°F.
**Surface resistivity between sides = 256 ohms/square at 0.2 volt.
***Surface resistivity between sides = 2 x 10" ohms/square at I 00 volts.
196 APPENDIX
1500 ohms is specified in the human ESD model of DoD-HDBK-263, and a realistic esti-
mate for human resistance is said to be 350 kilohms.6 However, the foil-containing laminate
bag survived even this test.
MOSFETs in a partially transparent, externally metallized bag were damaged by a dis-
charge from the finger of a statically charged person (Table 10), and the spark burned off a
small area of metal. Published data6 seem to disagree, but conditions differed somewhat
from ours; for one thing, the capacitive probe had no leads as ours did. Also, the bag may
not have been pulled tightly over the upper plate of the probe as it was on our test. As
expected, MOSFET damage was worse (Table 10) when the externally metallized bag had
a low-conductivity crease caused by cracking of the metallization by handling. Also note
that externally metallized bags, being good conductors, can participate in discharges at the
work station (Table 2).
The conclusion, which we had reached earlier,21 is that only foil-containing laminate
bags provide absolute protection against worst-case fields and discharges. A suitable com-
mercial product is a three layer laminate with aluminum foil (0.00035 inch thick) sand-
wiched between antistat-treated spun-bonded polyethylene on the outside and antistatic
polyethylene on the inside. Note that Mll..-B-81705, Type I, is a foil-containing laminate
but has an unsatisfactory liner which does not prevent internal triboelectric charging. 4•9
Equipment/Material Selection
and Handling Techniques
One school of thought recommends exclusion of conductive materials 13 while another, the
"conductive approach," visualizes most objects in the work station as conductive, grounded,
and therefore at zero potential. 22 Either approach, or a mixture of the two approaches, will
be effective with proper maintenance of materials (e.g., periodic treatment of antistatic
surfaces with topical antistat) and handling techniques which accommodate the limitations
of the equipment/materials. Table 11 illustrates the principle of complementary materials
and techniques.
The fact is that ESD-protective materials must be used correctly to be effective. 23 In
other words, operators are more important than equipment. For example, a grounded con-
ductive chair would be more dangerous to ESDS items then an ordinary chair if a careless
operator without a wrist strap caused discharges by brushing device leads against the uphol-
stery. Conversely, a very skilled operator can protect ESDS items with only the bare essen-
tials of equipment, including a wrist strap, because he knows where the dangers lie. Therefore,
operator training is essential for ESD control programs as prescribed in DoD-STD-1686. A
wide range of equipment/materials is described in DoD-HDBK-263, and it is up to the user
to integrate these with operator disciplines in a unified approach.
In our program, handling techniques accommodate a mixture of conductive,
static-dissipative, and antistatic materials. Some basic rules are: (1) Never touch ESDS
device leads with the fingers unnecessarily, and keep ESDS items away from all surfaces
except the bench top and the assembly to which the item is being attached. (2) Keep ESDS
items in Faraday cages, such as foil-containing laminate bags, when away from the static-
safe work station. (3) Clear the immediate work area of all but essential objects. A tote box
might fail to drain completely and bear a charge-but if kept to one side it will be harmless.
(4) Be sure to ground conductors, such as soldering iron tips, which touch leads or contacts
and can discharge directly into ESDS devices. (5) Recognize the limitations of supplementary
PAPER NO.3 197
Table 11. Complementary EqulpmentlMaterlals and Techniques.

ESDHAZARD EQUIPMENT/MATERIAL TECHNIQUES
Bench top Static-dissipative laminate Avoid unnecessarily touching

any surface, even this, with
leads of ESOS items.
Ordinary plastics Ionization equipment, Identify and exclude if possible;
topical antistat treat with antistat; use
local or room ionization.
Chair If possible, use conductive Don't closely approach with
chair on grounded ESOS items.
conductive surface.
Tote box Conductive Be aware of possible sloughing.
Tote box Antistatic Periodically treat with antistat
solution.
Tote box Either type Ensure grounding when on
work surface; avoid touching
with ESOS items; place
ESOS items in Faraday
cage, such as foil-containing
laminate bag, before adding
to box; keep box to one side
when working on ESO items.
Operator's skin Wrist strap Always use strap and check its
resistance often.
Operator's arm hair and None Tie head hair back; don't
head hair closely approach hair with
ESOS items.
Operator's clothing Antistatic smocks Don't closely approach with
ESOS items.
Unavoidable static fields Field meter Avoid in accordance with
(eg, from electrical Figure 3.
equipment)
Static fields and Foil-containing laminate Never remove ESOS items
discharges away bags or other Faraday from Faraday cage except
from work station cages at static-safe work station.
protective measures such as humidification or ionization and do not be overly dependent

on them. Humidification merely lessens ESD hazards, and ionized air requires time to
neutralize charges or cannot neutralize them at all insofar as the static field is collapsed by
proximity to a ground plane. (6) Check wrist straps frequently. Various resistance testers
are on the market. (7) Check with a field meter for static fields in the work station. For
example, a nearby window might generate a high enough charge from dry, dusty wind
blowing over it to create a dangerous field in the work area. If fields are unavoidable, apply
the equation d = #11.8 as explained in the text above. (8) Be conscious that any assembly
operation, such as peeling tape from a roll, creates static charges. Deal with these as required;
for example, pull masking tape off the roll slowly and "wash" it in ionized air before applying
it to contacts on a circuit board. (9) Remember that you, the operator, are a primary hazard
even though grounded because you can participate in discharges. (10) Be scrupulous in
excluding common plastics, which may be personal items such as sandwich bags, from the
198 APPENDIX
work station. As a last resort, ordinary plastics which definitely cannot be excluded must be
treated with topical antis tat.
In general, equipment and materials must be selected carefully from the gamut of prod-
ucts available, and then handling techniques must be adapted to the products chosen.
Conclusions
1. Tests confirmed that static-dissipative bench tops are preferred to conductive sur-
faces not only for personnel safety reasons but because a slower discharge rate can reduce
damage to ESDS items.
2. A charged conductive tote box is only slightly more hazardous than a charged anti-
static box when an ESDS item enters the static field of the box. But when boxes rest on a
grounded static-dissipative work surface, a conductive box participates much more readily
in harmful discharges than does an antistatic box. Both types of boxes have their pros and
cons, and either is suitable if used properly.
3. Statically charged nonconductive plastics can present a field hazard when the appar-
ent charge is as little as 500 volts. However, a charge of as much as 13,000 volts is not
deliverable to an ESDS item by contact (discharge) when the field is suppressed to 300
volts apparent charge by proximity to a ground plane.
4. Under worst-case conditions, using a MOSFET with long "antennas," a rough equa-
tion for the minimum safe distance d (in inches) from a surface charged to voltage V is: d =
.[V/I.8. When the antennas are shielded, as they often are in practice on circuit-board
modules, the constant in this equation rises from 1.8 to at least 7.
5. Based on the above data, basic rules are: (1) Ground all conductors so that none can
have a charge of over 50 volts. (2) Check with a field meter and allow no more than 300
volts apparent charge on nonconductors such as ordinary plastics. (3) If static fields are
unavoidable, keep ESDS items d inches away from charged surfaces according to the worst-
case equation d = .[V/I.8.
6. Electrical air-ionizing blowers or nozzles were found not to present a field hazard if
ESDS items are kept at least 10 inches away from the corona-discharge points. For close
work, nuclear-type air nozzles are indicated.
7. A pulsed dc benchtop ionizer, with no fan, was found to be as effective as conven-
tional ac blowers. The lack of an airstream is advantageous. (Note: Slow-pulsed ionizers
later proved hazardous; see "Ionization" in Chapter 2.)
8. A room ionization system with dc emitters proved to be effective-about equal to a
typical bench-model blower at 5 feet- and relatively uniform regardless of the position of
charged objects on the bench surface under the emitters. At worst-case conditions (emitter
of only one polarity operating), an isolated conductor at bench level (capacitance 129 pF)
was not charged sufficiently to damage MOSFETs.
9. The possibility of a conductive masking tape of foillfiber construction is suggested.
The aluminum foil prevents a static field by charge drainage and voltage suppression.
10. A corona-discharge point mounted above a small-scale grit blaster reduced the charge
developed on circuit-board material from 700 volts to the harmless level of 60 volts.
II. A bead-chain wrist strap is effective but tends to ride up over sleeves. Snug-fitting
straps in which metal contacts the skin are recommended.
12. In out tests, MOSFETs were damaged in partially transparent metallized bags, under
worst-case conditions, when the bags were touched by an operator charged to an estimated
PAPER NO.3 199
8000 volts by walking on a carpet. In contrast, MOSFETs in a foil-containing laminate bag

were unaffected by a lO,OOO-volt discharge from a simulator with human-level capacitance
but three orders of magnitude less resistance than a person. Only foil-containing laminate
bags are recommended for Faraday-cage protection against worst-case fields and discharges.
13. A synthesis of complementary equipment/materials and handling techniques is
necessary for an effective ESD control program in accordance with Dod-HDBK-1686.
Examples are given.
14. Operators are more important than materials. Some basic operator disciplines are, in
brief: (I) keep ESDS items away from all surfaces except the bench top and the assembly to
which the item is being attached, (2) use Faraday-cage bags/containers, (3) clear the imme-
diate work area of all but essential objects, (4) beware of charged conductors, (5) do not be
overly dependent on supplementary protection such as humidification or ionization, (6)
check wrist straps frequently, (7) check for static fields with a meter, (8) beware of static
charges caused by assembly operations, (9) remember that a grounded operator can partici-
pate in harmful discharges, and (10) rigorously exclude common plastics.
(Note: Later research modified some of the above conclusions, as seen in the text of this
book. For example, we became more cautious about ionization.)
References
I. I. M. Kolyer and W. E. Anderson, "Pennanence of the Antistatic Property of Commercial Anti-
static Bags and Tote Boxes," Reliability Analysis Center EOSIESD Symposium Proceedings,
pp. 87-94, Las Vegas, NY, 1983.
2. B. A. Unger, "Electrostatic Discharge Failures of Semiconductor Devices," Reliability Physics
Symposium, April 1981.
3. I. R. Huntsman, D. M. Yenni, Jr., and O. E. Mueller, "Fundamental Requirements for Static
Protective Containers," presented at 1980 NEPCON/West Conference, Anaheim, CA.
4. D. M. Yenni, Ir., and I. R. Huntsman, "The Deficiencies in Military Specification MIL-B-81705:
Center EOS/ESD Symposium, Denver, CO, 1979.
5. "Latent ESD Failures: A Reality," Evaluation Engineering Magazine, p. 80, April 1982.
6. I. R. Huntsman and D. M. Yenni, Ir., "Test Methods for Static Control Products," Reliability
7. Electronics Industry Association Interim Standard No.5, January 1983.
8. "Electromagnetic Shielding Effectiveness," technical infonnation sheet from the Bemis Co.,
Inc., ESD Protective Materials Dept., undated, received 1983.
9. N. B. Fuqua and R. C. Walker, "ESD Controls Study, Final Report," Reliability Analysis Center
No. 01115-30-2, September 1981.
10. B. Unger, R. Chemelli, P. Bossard, and M. Hudock, "Evaluation of Integrated Circuit Shipping
Tubes," Reliability Analysis Center EOSIESD Symposium Proceedings, pp. 57-64, Las Vegas,
NV, 1981.
II. J. R. Huntsman and D. M. Yenni, Jr., "Charge Drainage vs. Voltage Suppression by Static Con-
trol Table Tops," Evaluation Engineering Magazine, March 1982.
12. I. R. Mileham and N. I. Safeer, "Selection of Static Eliminating Decorative Table Top Mats or
Laminates," presented at 1984 NEPCON/West Conference, Anaheim, CA. The material we
tested is No.6 in the tables of this reference.
13. D. C. Anderson, "ESD Control: To Prevent the Spark that Kills," Evaluation Engineering Maga-
zine, pp. 120-131, July 1984.
200 APPENDIX
14. R. G. Chemelli, B. A. Unger, and P. R. Bossard, "ESD by Static Induction," Reliability Analysis
Center EOSIESD Symposium Proceedings, pp. 29-35, Las Vegas, NY, 1983.
15. J. E. Berry, "Static Control in Bare Board Testing," Electronic Packaging and Production, pp.
161-162, August 1981.
16. C. F. Mykkanen and D. R. Blinde, "The Room Ionization System: An Alternative to 40 Percent
RH," Evaluation Engineering Magazine, pp. 76-88, September 1983.
17. "Electrostatic Discharge Training Manual," NAVSEA SE 003-AA-TRN-0IO, Published by Di-
rection of Commander, Naval Sea Systems Command.
18. "Arm Hair Pin-Pointed as New ESD Hazard," Evaluation Engineering Magazine, p. 70, May
1984.
19. GIDEP (Government-Industry Exchange Program) Alert No. D5-A-84-OI, May 21,1984.
20. GIDEP (Government-Industry Exchange Program) Alert No. MX-A-82-02, March 21, 1983.
21. J. M. Kolyer and W. E. Anderson, " Selection of Packaging Materials for Electrostatic Dis-
charge-Sensitive Items," Reliability Analysis Center EOSIESD Symposium Proceedings, pp.
75-84, Las Vegas, NY, 1981.
22. R. Euker, "ESD in Integrated Circuit Assembly," Static Digest, published by Static Control
Systems Division, 3M Co., July 1983.
23. G. E. Hansel, "The Production Operator: Weak Link or Warrior in the ESD Battle?" Reliability
Analysis Center EOSIESD Symposium Proceedings, pp. 12-16, Las Vegas, NY, 1983.
PaperNo. 4
Presented at the 7th Annual Electrical Overstress/Electrostatic Discharge

Symposium, Minneapolis, Minnesota, September 10-12,1985, Sponsored by
EOS/ESD Association and ITI Research Institute, EOSIESD Symposium
Proceedings, EOS-7, 1985, page 111.
PERFORATED FOIL BAGS: PARTIAL

TRANSPARENCY AND EXCELLENT ESD PROTECflON
John M. Kolyer and William E. Anderson

Anaheim, CA 92803-4192
Note: The following is a condensed version of this paper. See the Symnposium
Abstract
Commercial foil-containing laminate bags provide "Faraday-cage" protection for electro-
static discharge (ESD)-sensitive items but are opaque. However, foil may be perforated to
give see-through capability if the metal is thick enough and the stamped holes have out-
wardly directed burrs which may act as "lightning rods" to intercept discharges. A wire
screen serves the same purpose, with its high points presumably receiving the discharges.
This principle is applicable to various other containers, e.g., tote boxes. Thus, partial trans-
parency need not mean a sacrifice in ESD protection.
Introduction
Commercial "Faraday-cage" bags are intended to protect against electrostatic fields and
discharges. A secondary goal is protection against EMI (electromagnetic interference), spe-
cifically at 1-10 GHz per MIL-B-81705, Type I. These bags fall into four classes: (1)
foil-containing laminates which give good ESD and EMI protection but are completely
opaque, (2) metallized bags in which the metallized layer is thin enough to allow partial
transparency with a considerable sacrifice of ESD and EMI protection (also, external met-
allization can crack on handling, thus lowering shielding effectiveness), (3) bags with a grid
of conductive ink that are fairly transparent but give poor ESD protection, and (4) metal-
fiber-containing bags that are also fairly transparent but give poor ESD protection.
Therefore, our objective was a new bag laminate which combines the "best of both worlds"
to a considerable extent: see-through capability along with good ESD protection and EMI
protection in the 1-10 GHz (radar) range.
201
202 APPENDIX
Materials Tested
Nine commercial ESD-protective bag materials were tested (Table 1): antistatic poly,
carbon-loaded, ink grid, buried metallization A, external nickel, copper fibers, buried met-
allization B, buried metallization C, and foil laminate. Several of these products have been
described and characterized by us. 1.2,3
The other six materials in Table 1 were fabricated in the laboratory from antistatic
polyethylene (6 mils) and either aluminum foil, perforated aluminum foil, or metallic
screens.
The test conditions are listed in Table 2 and diagrammed in Figs. 1,2, and 3.
Table 1. Bag Materials Tested.

THICKNESS TRANSPARENCY •
DESIGNATION MILS % DESCRIPTION
Antistatic Poly 6 87 Antistatic Polyethylene (MIL-B-81705, Type ll)

Carbon-Loaded 4 0 Conductive, Carbon-Loaded Polyethylene
Ink Grid 4 68 Laminate with Buried Grid of Conductive Ink
and Antistatic Liner
Buried Metal- 3.5 74 Polyester + Vapor-Deposited Metal + Antistatic
lization A Polyethylene Liner
External Nickel 2.5 35 Vapor Deposited Nickel + Polyester + Antistatic
Polyethylene Liner
Copper Fibers 3.7 78 Fine Copper Wires Dispersed in Polyethylene
Buried Metal- 3.3 60 Resinous Protective Coating + Vapor-Deposited
lization B Aluminum + Carrier Film + Antistatic
Polyethylene Liner
2-mil Perforated 12 38 Perforated Foil Encapsulated in Antistatic
Foil Polyethylene
Buried Metal- 3 60 Polyester + Vapor-Deposited Aluminum +
lization C Antistatic Polyethylene Liner
5-mil Perforated 15 38 Perforated Foil (See Text) Encapsulated in
Foil Antistatic Polyethylene
Foil Laminate 10 0 Antistic Spun-Bonded Polyethylene + 0.35-mil
Aluminum Foil + Antistatic Polyethylene Liner
1.4-mil Foil 7.4 0 1.4-mil Aluminum Foil + 6-mil Antistatic
Polyethylene Liner
5.6-mil Foil 11.6 0 Four layers of 1.4-mil Aluminum Foil + 6-mil
Antistatic Polyethylene Liner
Aluminum Screen 23 59 Aluminum Screen (See Text) Sandwiched Be-
tween Sheets of 6-mil Antistatic Polyethylene
Copper Screen 35 41 Copper Screen (See Text) Sandwiched Between
Sheets of 6-mil Antistatic Polyethylene
Table 2. ShieldingIDischarge Test Conditions.
TEST NUMBER
2 3 4 5 6 7 8
Capacitive No No Yes Yes Yes Yes Yes Yes

Sensor
Volts +25,000 About 35,000 +5000 -6000 +10,000 About 35,000 +24,000 About 35,000
to-8000
Resistance, 0 Tesla Coil 150 Human 150 Tesla Coil 150 Tesla Coil
ohms
Capacitance, 167 Tesla Coil 150 Human 185 TeslaCoil 185 Tesla Coil
picofarads
Time, seconds (Discharge) 30 (Discharge) (Discharge) (Discharge) 30 (Discharge) 30
Surface Nonconductive Nonconductive Grounded Grounded Grounded Nonconductive Grounded Grounded
Under Bag
~
a
~
....
fA
204 APPENDIX
ROHCOlIDUCTlVE SURFACE
Fig. I. Tests number 1 and 2 (Table 2).
us ISTOI.
PR>BI
(TISTS 3.5.7)
Fig. 2. Tests number 3, 4, 5, 7, and 8 (Table 2).
HONCONDUCTlVE SURFACE
Fig. 3. Test number 6 (Table 2).

PAPER NO.4 205
Table 3. Shielding/Discharge Data Summary.
Bag Material
Antlatatlc Poly F
Carbon·Loaded F
Ink Grid (F) F

Burled Metallization A (F) f F
External Nickel F F
Copper Fiber. P
Burled Metallization B P F f
2·mll Perforated P (F) f F
foil
,
Burled Metallization C P P (F) F F F F
5-mll Perforated p p p p F
foil
Foil laminate p p p p (F) f

1....mll foil p (F)
5.8·mll Foil p P
Aluminum Screen p p
Copper Screen p p
Rating
Lagend: P = pealed (no MOSFETI damaged)

(F) = barely failed /115(081')
F .. failed (1115(18) or fa lure ratio> 115)
Results
The data are summarized in Table 3. Note the excellent results given by screens. After
publication of this paper a screen bag was commercialized; the aluminum screen was sand-
wiched between layers of antistatic polyethylene. Perforated foil was not as effective as
screen and just as expensive. so it was never commercialized.
Conclusions
1. Metallized. carbon-loaded. or ink grid bags have been reported not to withstand a
direct static charge above 25 kV. Our own tests indicate a much lower threshold for these
materials. Note that personnel walking on a carpet at 1~20 percent relative humidity gen-
erate typically 35 kV.
2. Two independent groups of investigators have recommended (EOS/ESD Sympo-
sium. 1984) that highly ESD-sensitive items should always be protected by metal-foil bags.
206 APPENDIX
3. We have now found that metal foil can be perforated and still shield against ESD and
radar. Wire screen is even better. Outwardly directed burrs on the foil, or prominences on
the screen, may act as "lightning rods" to intercept discharges.
4. The effectiveness of perforated foil or screen laminates as a moisture vapor barrier
would depend on the permeability of the plastic film used.
5. Preliminary designs with performed foil or wire screen would cost twice as much as
a commercial foil laminate. However, added expense might be justified in protecting valu-
able, high-reliability electronics.
References
1. J. M. Kolyer and W. E. Anderson, "Selection of Packaging Materials for Electrostatic Discharge-
Sensitive Items," Reliability Analysis Center EOSIESD Symposium Proceedings, pp. 75-84,
Las Vegas, NV, 1981.
2. J. M. Kolyer and W. E. Anderson, "Permanence of the Antistatic Property of Commercial Anti-
static Bags and Toite Boxes," Reliability Analysis Center EOSIESD Symposium Proceedings,
pp. 87-94, Las Vegas, NV, 1983.
3. J. M. Kolyer. W. E. Anderson. and D. E. Watson. "Hazards of Static Charges and Fields at the
Work Station." Reliability Analysis Center EOSIESD Symposium Proceedings. pp. 7-19. Phila-
delphia. PA, 1984.
PaperNo. 5
Presented at the Electrical Overstress Exhibition, Anaheim, California,

January 21,1986. Proceedings of the Technical Program,
1986 Electrical Overstress Exhibition, page 34.
COST-EFFECTIVE METIIODS OF
TESTING/MONITORING WRIST STRAPS
John M. Kolyer and Donald E. Watson

Anaheim, CA 92803
Abstract
An ungrounded operator is extremely dangerous to ESD (electrostatic discharge)-sensitive
devices as illustrated by a test at high relative humidity (54 percent) in which the operator
damaged transistors by rolling his chair. An ESD-protective smock did not help to protect
devices. A maximum resistance from operator to ground of 10 megohms is recommended,
and this value was exceeded in 3 of 36 tests under manufacturing conditions. Because of
flexing of the cords and other abuse, wrist straps have a limited lifetime, but they are too
expensive to scrap in a replacement cycle and instead must be monitored. Periodic monitor-
ing is unsatisfactory because ESD-sensitive items will be exposed to an ungrounded opera-
tor for some period of time, even if only an hour, when a strap fails. Then material review
action might be required on all hardware handled since the last successful test of that strap.
Therefore, continuous monitors may be "the wave of the future." Commercial continuous
monitors work on the capacitance principle, but another method is a continuity check across
the operator's skin, up and down a two-conductor cord, and in and out two separate ground
connections. Prototype monitors of this two-conductor design are being tested in a produc-
tion environment. High-cost items, e.g. $1000 modules, especially deserve continuous
monitoring.
Introduction
This paper is presented in the form of slides shown at the Electrical Overstress Exposition
(EOE) in Anaheim, CA, January 21,1986. Accompanying each slide is additional discus-
sion and clarification as required.
Wrist straps have a limited lifetime I because of frequent flexing and other abuse. Since a
time-based replacement cycle would scrap many good straps and be wasteful, monitoring is
necessary. This monitoring is preferably continuous, because the failure of a wrist strap in a
207
208 APPENDIX
periodic check raises the questions: "How long ago did this strap fail?" and "Have products
been damaged in the meantime?"
As shown in the following outline, the danger of a failed strap was demonstrated, simple
methods for testing resistance were devised, straps in an assembly plant were checked, and
means of monitoring were investigated.
OUTLINE
• DAMAGE CAUSED BY LACK OF WRIST STRAP
• RESISTANCE VS CHARGE ON OPERATOR
• ALLOWABLE RESISTANCE, OPERATOR TO GROUND
• RESISTANCE TEST METHODS
• THROUGH STRAP
• OPERATOR TO GROUND
• EQUIPMENT
• MEGOHMMETER
• ZAPFLASH
• RESISTANCE TESTING DATA
• FACTORS IN RESISTANCE, OPERATOR TO GROUND
• MONITORING
• PERIODIC VS CONTINUOUS
• CONTINUOUS DESIGNS
• CAPACITANCE TYPE
• TWO-CONDUCTOR TYPE
• ADVANTAGES OF TWQ..CONDUCTOR TYPE
• CONCLUSIONS
The charge on the operator was read by means of a Trek Model 512 field meter aimed at
an electrically isolated I-ft2 aluminum plate to which the operator was attached by a wrist
strap. The temperature was nOE
The transistor was a Motorola 2N4351 MOSFET. The testing procedure and damage
criteria have been described. 2
PAPER NO.5 209
DAMAGE CAUSED BY
LACK OF WRIST STRAP
• GROUNDED OPERATOR IS LESSER OF TWO EVILS VS
UNGROUNDED OPERATOR
(KOLYER ET AL, EOSIESD SYMPOSIUM 1984)
• UNGROUNDED OPERATOR IN ESC-PROTECTIVE SMOCK,

54% RELATIVE HUMIDITY
• ROLLED CHAIR: 500 VOLTS, TRANSISTORS DAMAGED
• SHUFFLED FEET: 1000 VOLTS, TRANSISTORS DAMAGED
A charged operator is a very great danger because he can damage ESD-sensitive devices
by direct discharge or "injection" rather than indirectly by induction from a static field.
The published data3 were obtained with an operator holding a probe attached to an elec-
trometer and an oscilloscope. The Trek meter and l_ft2 plate were used as described under
the preceding slide. Since 10 volts was our limit of detection. the result at I megohm agrees
with that obtained by the more sophisticated method: a I-megohm resistor in the wrist strap
allows negligible voltage on the operator.
RESISTANCE VS CHARGE ON OPERATOR

RESISTANCE, OPERATOR PEAK VOLTAGE
TO GROUND, MEGOHMS ON OPERATOR REFERENCE
<2
10 11
50 30
100 80
INFINITE (NO STRAP) 1550
1 <10 2
INFINITE (NO STRAP) 1100 2
1. HUNTSMAN AND YENNI, EOSIESD SYMPOSIUM 1982

2. OUR DATA
210 APPENDIX
ALLOWABLE RESISTANCE,
OPERATOR TO GROUND
• FOR DEVIDCES SENSITIVE TO 50 VOLTS MIN: 50 MEGOHMS
• FOR DEVICES SENSITIVE TO 20 VOLTS MIN: 10 MEGOHMS
• CONSERVATIVE LIMIT FOR ALL DEVICES: 10 MEGOHMS
These thresholds are based on the data of the preceding slide. A charge of 11 volts on
the operator is hannless to most devices at the present time; in the future this allowable
resistance might have to be lowered. A I-megohm resistor is presently standard for wrist
straps.
As an added precaution, a static-limiting floor finish (UL288, Hanson-Loran Chemical
Company) was used. The chair-rolling or foot-shuffling operator in the slide before last
would have generated little charge with UL288. In the test, a conventional acrylic finish
was on the tile floor.
A Beckman L-lO megohmeter at 10 volts was used. For quick tests, a Zapflash (see two
slides below) is suitable.
RESISTANCE TEST METHOD,

THROUGH STRAP
ALLIGATOR
CUPS
PAPER NO.5 211
RESISTANCE TEST METHOD,

OPERATOR TO GROUND
The same megohmeter at 10 volts was used. The metal can (tin-coated steel) is a conve-
nient, hand-filling probe.
The Zapflash was tested with various resistance levels. At 1.0 megohm the light (indicat-
ing continuity) was very bright. At 3.0 megohms the light was relatively bright, but at 5.2
megohms it was dim and at 6.2 megohms it was barely visible.
RESISTANCE TEST METHODS, EQUIPMENT
• MEGOHMMETER AT 10 VOLTS, EG. BECKMAN MODEL L-10
• ZAPFLASH (ANDERSON EFFECTS). HIGH·RESISTANCE

CONTINUITY TESTER. LIGHTS AT ABOUT 6 MEGOHMS
MAX. GOOD FOR QUICK CHECKS
Various combinations of wrist straps and operators gave a total of 36 tests. These mea-
surements were made in a manufacturing situation. One operator had spliced the cord of his
bead chain, thus increasing its resistance from 1.0 to 1.8 megohms; this is not a serious
increase, but it might have been. Another operator was pleased with her loose-fitting expan-
sion band; resistance from her skin to ground was 30 megohms for this reason.
212 APPENDIX
RESISTANCE TEST DATA

RES~TANCE,MEGOHMS
TYPE NUMBER NUMBER THROUGH OPERATOR TO

STRAP SAMPLES OPERATORS STRAP GROUND
EXPANSION 5 7 1.00-1.05 1.0-9.0
BAND (1)
EXPANSION 5 8 1.00-1.05 1.05-30
BAND (2)
EXPANSION 2 1.0 2.0,4.3
BAND (3)
CLOTH BAND 3 4 2.7-110 1.5-65
(CONDUCTIVE)
BEAD CHAIN 4 8 1.0-1.8· 1.5-40
·SPLICED CORD GAVE 1.8
These examples illustrate the need for operator discipline, especially in the absence of con-
tinuous monitoring of the straps.
The three out-of-spec cases occurred in a program of periodic (weekly) checks and illus-
trate the desirability of continuous monitoring. In a periodic check, the operator may slide
the band up his arm to tighten it or make some other adjustment to pass the test. An analogy
is annual inspection of automobile exhaust emissions; once the driver has passed inspec-
tion, he might readjust his engine to exceed the allowable level of smog-producing gases.
FACTORS IN RESISTANCE,
OPERATOR TO GROUND
• OVERSIZED EXPANSION BAND
• INCREASED RESISTANCE OF CLOTH BAND (SOILING)
• STRETCHING OF CLOTH BAND, CAUSING LOOSENESS
• SOILING OF BEAD CHAIN
• DRYSKIN
• IN 36 TESTS, RESISTANCE EXCEEDED 10 MEGOHMS IN 3 CASES:

EXPANSION BAND (LOOSE FIT), CONDUCnVE CLOTH BAND, BEAD
CHAIN
PAPER NO.5 213
PERIODIC VS CONTINUOUS MONITORING
• PERIODIC
• ADVANTAGE: LOW EQUIPMENT COST
• DISADVANTAGE: ESD-SENSmVE DEVICES ENDANGERED

FOR SOME nME INTERVAL, EVEN IF ONLY AN HOUR
• CONnNUOUS
• ADVANTAGES: DEVICES ALWAYS PROTECTED, WRIST

STRAP QUALITY SECONDARY, LESS OPERATOR DISCIPLINE
NEEDED
• DISADVANTAGE: RELAnVE HIGH EQUIPMENT COST (BUT COST OF

LABOR FOR PERIODIC MONITORING MUST BE
SUBTRACTED)
CONTINUOUS MONITOR,
CAPACITANCE TYPE
• CHECKS CAPACITANCE, EG, 150 pF
• ALARM FLASHES AT ALLOWABLE MAX

OPERATOR-TO-GROUND RESISTANCE, EG, 3.5 MEGOHMS TOTAL
• USES CONVENnONAL WRIST STRAP
• COMMERCIALLY AVAILABLE FROM MORE THAN ONE SUPPLIER
Commercial products include the Simco Wrist Strap Monitor Model M50A, the Ground
Gard from Static Prevention, Incorporated, and the GM-ICT from Westek. Charleswater
Products has a developmental product to be introduced early in 1986.
214 APPENDIX
Note that in the absence of continuous monitoring, all hardware handled since the last
successful test of a failed strap might be subject to a material review action.
CONTINUOUS MONITOR,
CAPACITANCE TYPE: DIAGRAM
1·MEGOHM
RESISTOR
This is a very simplified representation of the Simco Wrist Sgrap Monitor Model M50A.
CONTINUOUS MONITOR,
TWO-CONDUCTOR TYPE
• CHECKS LOOP ACROSS SKIN, INTO FIRST GROUND
CONNECTION, AND OUT SECOND GROUND CONNECTION
• ALARM FLASHES AT 20 MEGOHMS, OPERATOR·TO-GROUND
• SPECIAL SPLIT·CONDUCTOR WRIST STRAP NEEDED
• PROTOTYPES TO BE TESTED IN PLANT

PAPER NO.5 215
The product under evaluation is from Semtronics Corporation. Westek is developing

their own model.
CONTINUOUS MONITOR,
TWO-CONDUCTOR TYPE: DIAGRAM
",
.......... ------------_/T
This is a very simplified representation of the Semtronics prototype. The commercial

system, Part No. EN435S, will comprise a Sentinel monitor, a dual coil cord, and a wrist
strap.
ADVANTAGES OF TWO-CONDUCTOR
TYPE CONTINUOUS MONITOR
• FAILURE PROBABLY IN ONE CONDUCTOR, NOT BOTH, SO

ESD-SENSITIVE ITEMS NEVER JEOPARDIZED
• CHECKS GROUND CONNECTIONS AS WELL AS OPERATOR
• NOT FOOLED BY HIGH-CAPACITANCE OBJECT TOUCHING

STRAP
• WHEN BOTH CONDUCTORS ARE UNPLUGGED FROM BOX,

MECHANICAL SWITCH SHUTS OFF ALARM, EG, WHEN
OPERATOR TAKES BREAK
216 APPENDIX
At least one maker of the capacitance-type monitor expects to add a feature to check
ground connections. Also, the problem of the strap being deceived by a high-capacitance
object may possibly be overcome. With these improvements, the capacitance-type monitor
would more nearly equal the two-conductor type, so that a choice between the two might
largely depend on cost.
Cost effectiveness is not easy to calculate at this point, since the costs of periodic inspec-
tion and of hardware damaged due to lack of continuous monitoring must enter the equa-
tion. In general, periodic checks may be adequate for low-cost products, but continuous
monitoring is indicated for expensive assemblies. After all, most of us consider the wrist
strap our first line of defense, and if it fails we have dropped our guard, however briefly, in
the absence of continuous monitoring.
CONCLUSIONS
• WRIST STRAP FAILURE CAN EASILY CAUSE DAMAGE OF

ESD-SENSmVE DEVICES
• RESISTANCE TEST METHODS ARE DESCRIBED
• 3 FAILURES (RESISTANCE OVER 10 MEGOHMS) IN 36 CHECKS

OF STRAPS IN PLANT
• LOOSE BANDS AND DRY SKIN RAISE RESISTANCE
• CONTINUOUS MONITORING PREFERRED
• TWO-CONDUCTOR MONITOR PREFERRED OVER CAPACITANCE TYPE
• TWO-CONDUCTOR MONITOR PROTOTYPES TO BE TESTED IN PLANT
• HIGH-COST ITEMS, EG, $1000 MODULES, DESERVE

CONTINUOUS MONITORING
PAPER NO.5 217
References
1. A. P. Hahl, "A Wrist Strap Life Test Program," Reliability Analysis Center EOSIESD Sympo-
sium Proceedings, pp. 94-96, Philadelphia, PA, 1984.
2. J. M. Kolyer, W. E. Anderson, and D. E. Watson, "Hazards of Static Charges and Fields at the
Work Station," Reliability Analysis Center EOSIESD Symposium Proceedings, pp. 7-19, Phila-
delphia, PA, 1984.
3. J. R. Huntsman and D. M. Yenni, Jr., ''Test Methods for Static Control Products," Reliability
PaperNo. 6
Presented at the 8th Annual Electrical Overstress/Electrostatic Discharge

Symposium. Las Vegas. Nevada. September 23-25. 1986. Sponsored by
EOS/ESD Association and aT Research Institute.
EOSt ESD Symposium Proceedings. EOS-8. 1986. page Ill.
METHODOLOGY FOR EVALUATION OF

STATIC-LIMITING FLOOR FINISHES
John M. Kolyer and Dale M. Cullop

Anaheim. CA 92803
Hanson Loran Chemical Co .• Inc.

6700 Caballero Blvd.
Buena Park. CA 90620
Note: The following is a condensed version of this paper. See the Symnposium
Proceedings for the complete test.
Summary
Methodology is described for evaluating floor finishes which limit static buildup on per-
sonnel wearing ordinary shoes and drain charges from personnel wearing conductive foot-
wear or heel grounders. Procedures include triboelectric charging of various synthetic shoe
sole materials and an accelerated scrubbing test to predict durability. The objective is a low-
charging. easily maintained. and cost-effective product.
Introduction
Several commercial ESD-control floor finishes were evaluated in mid-1985. Some were
unsatisfactory in limiting static generation. while others had questionable floor finish
proterties such as scrub resistance; in general. non-ESD properties seemed compromised
by antistat addition. All the products were judged too expensive to be cost-effective. These
deficiencies prompted our study.
Our objective was a low-charging. easily maintained. and cost-effective acrylic coating.
and the search involved development of methodology for evaluating commercial finishes
and perfecting new and better ones. The emphasis was on simple. relatively inexpensive
equipment and realistic performance tests.
218
PAPER NO.6 219
We prefer the tenn "static-limiting" because this property, not conductivity per se, is
what is desired if ordinary shoes are worn. "Static dissipative" and "antistatic" refer to
surface resistivity ranges, and surface resistivity correlates imperfectly with triboelectric
charging of shoe soles, as will be seen. "Static-preventing" would be a misnomer or false
claim because all antistatic materials penn it some charging, especially of certain "problem"
synthetics.
When conductive footwear or heel grounders are prescribed, low surface resistivity is
the key ESD property, and tests for triboelectric charging are superfluous. However, our
methodology remains pertinent because it predicts durability in tenns of surface resistivity
as well as triboelectric charging propensity.
Floor Finishes Tested

The nine commercial products tested represented most of these offered in the U.S. in mid-
1985. These were acrylic, modified with wax in some cases, with a minimum solids content
of 15-24.5 percent and pH values in the 8-10 range. Note that deficiencies in these prod-
ucts may have been more or less corrected after the time of our testing, because floor finish
fonnulations can be changed at a moment's notice. Also, several new products have ap-
peared. Hence this paper is concerned with general methodology rather than evaluation of
specific products.
The four experimental products tested prepresent stages in the evaluation of a new com-
mercial product which was a wax-modified acrylic with a nominal solids content of 20
percent and a pH of 9.7.
Grounded aluminum (alloy 6061, temper T6) was included in the tests as a reference
material to demonstrate that a conductive floor surface can charge synthetic shoe soles.
Thus a reduced surface resistivity is necessary but not sufficient.
The various experimental methods are shown in Fig. I through 7. Fig. 8 shows that the
actual voltage in the walk test can deviate considerably from the meter reading because of
inertia of the needle. To be conservative, we multiply meter readings by a factor of 2.
Cost Effectiveness
Cost effectiveness is a vital "property" and must be considered in selecting a floor
finish.
In conjunction with ordinary ungrounded shoes, the use of a static-limiting floor finish,
like humidification and room ionization, is a supplementary procedure, as opposed to pri-
mary ESD control methods such as the use of wrist straps. Supplementary procedures are
intended to create a relatively benign background or "safety net" for imperfect primary
methods. A perfect acrobat doesn't need a safety net, and a perfect operator who always
wears his wrist strap doesn't need a static-limiting floor finish, but in reality the reduction
of background hazards is worth some expense. The question is: how much expense?
220 APPENDIX
lxI-INCH TILE SAMPLE

(CAPSULE HOLDS SIX SAMPLES,
ONE INSIDE EACH FACE)
Fig. 1. Heel mark test.
ADJUSTABLE
SCALE
RUIIER
RING
STRING
(TO WINDLASS)
Fig. 2. Topaka slip test.
CROSS SECTION
Fig. 3. Gloss test.

PAPER NO.6 221
CABLE
(TO RECIPROCATING
MECHANISM)
Fig. 4. Detergent or water scrub test.
Fig. 5. Roller test.
Table I compares estimated costs of using a static-limiting floor finish derived from
experimental product No.4 versus a standard nonantistatic fmish of similar chemical nature
(acrylic and polyethylenic polymers). It is interesting that the chemicals account for only a
small fraction of the annual cost of floor maintenance: 4.2 percent for the conventional
finish and 6.5 percent for the static-limiting finish. For 10,000 ft2 of floor space, the pre-
mium for using the static-limiting finish is $235/year, which seems a low cost for any sig-
nificant additional EDS protection.
Note that the special finish could help justify saving mishandled items in material review
actions; for example, if an operator touched an ESD-sensitive item without wearing his
wrist strap, the argument might be made that the probable charge on his skin was below the
sensitivity level of the item based on walk-test monitoring data.
222 APPENDIX
f ....
, ...-c..': \
':.:.~
.... ' !'.' FARADAY
.........
.."':' . . "
.~..
":
. .
CUP
- -. .. ,.'. :.':> GROUNDED STEEL PAl L

STRING (TO
WINDLASS)
Fig. 6. Drag test.
AUJillrul
SHEET (1 "2)
STAND-OFF INSULATOR
(POLYETHYLENE BEAKER WITH
SLIT FOR ALlJUNUM SHEET)
Fig. 7. Walk test.

PAPER NO.6 223
0
1000 0
.... 0
~ 0
i
-i= 0 00
f
t
500 cP
0
...
vi
-'
i
0
VOLTS, ACTUAL
Fig. 8. Apparent versus actual voltage in walk test.
Table 1. Estimated Annual Costs of Using Conventional versus

Static-Limiting Acrylic Floor Finish.
COST, S/IO,OOO Fr2
CONVENTIONAL FINISH, STATIC-LIMITING FINISH,

$ IO/GALLON SI5/GAlLON
Finish (8 Coats @ 5 Gallons/Coat) 400 600

Neutral Cleaner (30 Gallons) 165 N/A
Antistatic Neutral Cleaner N/A 300
(30 Gallons)
Maintenance* (5 h/day@ $IO/h, 13,000 13,000
5-DayWeek)
Totals 13,565 13,900
Premium o 235
*Scrubbing/recoating, restoring, damp-mopping, dust-mopping. and possibly spray-buffmg. These

operations are done piecemeal and intennittently, as required.
224 APPENDIX
An advantage of the use of static-limiting floor finish with ordinary shoes as a supple-
mentary procedure is that no disturbing or even visible change is made in the work environ-
ment. In contrast, humidification can cause discomfort due to "mugginess," and room
ionization has been known to create anxiety about its imagined electrical effects on people.
However, thorough testing must assure that voltages are in fact sufficiently controlled with
the great variety of shoe soles which personnel may wear. For example, certain vinyl com-
pounds may be "problem" materials, as indicated by high charges in both the walk test and
laboratory tests.
Static-limiting floor finishes may be applied to conductive tile without destroying its
grounding capability. For example, two coats of the product derived from experimental
product No.4 increased the surface resistivity of a popular brand of conductive tile from
1 x 107 to 3 X 107 ohms/square and increased the resistance to ground (through a 5-lb,
2.5-inch-diameter NFPA 56A electrode) from 7 x 106 to 3 X 107 ohms at 10 volts. These
threefold or fourfold increases do not prevent rapid electrical drainage from conductive
footwear, and the finish provides wear-resistance and an attractive gloss.
Conclusions
1. Nine products commercially available in mid-1985 were evaluated to illustrate our
methodology and found to be generally unsatisfactory in performance and too high in cost.
Resistance to scrubbing and foot traffic seemed compromised by antistat addition. How-
ever, formulations are easily changed, and today's versions may be improved.
2. The products tested were not literally "zero-charge." They limited triboelectric charg-
ing of shoe soles but did not stop it. As of May 1986, some suppliers felt that ordinary,
ungrounded shoes could be worn without charging enough to damage ESD-sensitive de-
vices, but others felt that conductive footwear or heel grounders were needed to drain charges
including these created by movement of clothing.
3. The "acid test" of a static-limiting floor finish is the measurement of charges on
people who walk on it. When ordinary shoes are worn without conductive straps, surface
resistivity correlates imperfectly with charging and can be misleading.
4. The ever-changing shoe sole market presently includes a variety of synthetics such
as polyurethane, vinyl, ethylene-vinyl acetate, and styrene-butadiene rubber (SBR). These
materials vary in susceptibility to triboelectric charging; vinyls are high and polyurethanes
tend to be low (hence they are often used, with additives, in ESD-control footwear). Also,
different synthetics may respond in varied and unexpected ways to different floor finishes.
Therefore, an averaging effect is desired, and screening of candidate finishes should in-
clude several synthetics, not just Neolite (an SBR compound) as in AATCC Test Method
134-1979.
5. Compared to synthetics, leather soles are a minor problem. They hold only 15 per-
cent of the market, are usually used only on expensive shoes, and in general triboelectrically
charge much less than synthetics.
6. In laboratory tests, cylinders of various synthetics were rolled down a ramp of coated
tile, or pieces cut from actual shoe soles were dragged across the surface; in both cases the
specimens fell into a Faraday cup to measure the charge. The results roughly agreed with
voltages observed on people with random shoe soles when a variety of synthetics and sev-
eral people were measured to get average effects.
PAPER NO.6 225
7. Based on these ESD tests, methodology is proposed for developing/evaluating new

products in the laboratory. Basic floor finish properties such as slip, gloss, and scrub resis-
tance are included. The test equipment is simple and easy to use.
8. Correlation of laboratory tests with a floor test in a busy supermarket was good for
floor finish properties: durable finishes were distinguished from those subject to hazing and
scuffing. However, ESD properties deteriorated in all cases due to spray-buffing under the
wrong conditions (coatings not thick enough). It is recommended that spray-buffing of
static-limiting floor finishes be done with caution. A thin restorer coat keeps most of the
luster while reviving ESD properties.
9. With proper maintenance, including mopping with a solution of antistatic cleaner, a
good static-limiting floor finish should control the charging of walking personnel to 200
volts maximum (peak voltage of spikes) for most, but not necessarily all, synthetic soles at
50 percent relative humidity and 72°F. An experimental finish was successful at 33 percent
relative humidity and 72°F in limited testing.
10. Our tests were done in the vicinity of 50 percent relative humidity, which proved
acceptable for comparative purposes (screening of candidate finishes), but final on-the-
floor testing must involve the lowest relative humidity to be met in practice. Literature data
suggest a threefold rise in charging of rubber soles when the relative humidity drops from
50 to 15 percent at 72°F, but the increase could be much more.
11. Static-limiting floor finishes lower triboelectric charges on rolling equipment as well
as walking people. For example, annoying charges on supermarket shopping carts with
polyurethane wheels were reduced to an innocuous level.
12. The use of static-limiting floor finish with ordinary, ungrounded shoes is only a supple-
mentary procedure which provides a "safety net" for primary procedures such as the use of
wrist or ankle straps. Therefore, the added expense of using the finish must be moderate for
this backup procedure to be cost-effective. An example is given in which switching from a
conventional to a durable, easily maintained static-limiting finish costs a premium of only
$235/year/lO,OOO ft2 of floor. On conductive tile, the finish permits grounding while pro-
viding wear-resistance and gloss. Since some sort of finish must be used, why not a static-
limiting or dissipative one at little added expense?
Note: This paper appeared as an article in EOSIESD Technology Magazine, April 1987,
page 20:
Getting ESD Floor-Finish Tests

Rolling Toward Standarization
~
. ... By John M. Kolyer, Rockwell International and
.-....... Dale M. Cullop, Hanson loran Chemical Co.
,1~::=; These tests are good candidates for a standard method of
~'!:...... ___ evaluating the static·limiting properties of floor finishes.
PaperNo. 7
This paper is reprinted. by permission. from EOSIESD Technology.

October/November 1987.
TOTE BOX MATERIAL: HOW GOOD IS IT?
1. M. Kolyer. W. E. Anderson and D. E. Watson
Rockwell International Corp.•

Electronic Operations.
Anaheim.CA
Until recently. all tote boxes for ESD control were made of polyolefin and either topically
applied antistat. extruded-in antistat. or extruded-in graphitic (conductive) carbon. Each of
these materials has its limitations.
Topically applied antistat is only an expedient because the antistat wears off after some
undetermined period of use. Also. a wall of plain polyolefin. even with an antistatic surface.
provides little Faraday-cage shielding protection from external static fields or discharges
(Table I).
Extruded-in antistat provides a longer life than topically applied antistat because it
continues to bleed to the surface for some time. creating a weakly conductive sweat layer
from atmospheric moisture. However. handling. heat. contact with paper products. or
exposure to solvents will eventually deplete all the antistat (Ref 1). Again. shielding is
poor.
Extruded-in conductive carbon offers the advantage of permanence. but it has several
problems. A carbon-loaded polyolefin tote-box wall is conductive enough to endanger people;
a current of over 100 rnA, which is usually fatal (DOD-HDBK-263). can be carried at 110
V (Table 2). If the conductive box itself is charged it is more dangerous to devices than
antistatic or nonconductive boxes (Table 3 and Ref 2).
A high-conductivity surface is also dangerous to devices (Ref 3). especially if the opera-
tor should be charged and touches a sensitive lead to the box (Table 4).
However. the wall of a carbon-loaded polyolefin box is not conductive enough to be a
good Faraday cage; note that the volume resistivity of carbon-loaded polyolefin is on the
order of 102 Q-cm versus 1~ Q-cm for aluminum foil. Refs 3 and 4 agree that highly ESD-
sensitive components should always be protected by metal-foil bags. not carbon-loaded
bags or bags with thin. see-through metallization. This conclusion is applicable to other
containers such as boxes. Furthermore. carbon-loaded polyolefin imparts high triboelectric
charges to nonconductors stroked on it (Table 5 and Ref 1) and sloughs conductive particles
that could fall into open microelectronic devices and cause shorts.
Looking at New Materials

The basic defect of conventional boxes is that they are insufficient Faraday cages. So. two
multilayer designs with permanent ESD properties have been investigated: fiberboard-foil
226
PAPER NO.7 227
Table 1. Shlelding/Discharge Test.

CAPACITANCE RESISTANCE MOSFET
TOTE BOX TYPE (pF) (0) v DAMAGE·
Antistatic or carbon- Human Human -8,000 3/3 (OS)

loaded polyolefin
(0.140 in. wall)
Same as above 97 1,500 +5,000 2/2 (2S)
Same, but with air gap"'''' 97 1,500 +15,000 2/2 (2S)
Carbon-loaded poJyo- 97 1,500 +5,000 0/3
lefin as above, but
lined with 0.0003-in.
aluminum foil
Corshield containing one Human Human -8,000 0/10
layer of 0.00025-in. Tesla coil Tesla coil 35,000 1/5 (IS)
aluminum foil
Corshield folded to give Tesla coil Tesla coil 35,000 0/5
two layers of the foil
Vinyl (0.011 in.) on Tesla coil Teslacoil 35,000 0/5
20-gauge aluminum
sheet (0.0375 in.)
Sandwich of aluminum Tesla coil Tesla coil 35,000 0/5
screen sandwiched
between O.06O-in. sheets
of "Forbon" hard
vulcanized fiber (NVF
Co., Container Div.)
*For example, 2/5 (IS) would mean that five MOSFETs were tested, two were damaged and one of those damaged
was shorted.
**I-in. gap between each electrode and inner surface of box.
and vinyl-metal sheet. Multiple layers are necessary because a homogeneous wall is not
capable of providing both a safe antistatic or nonconductive surface and Faraday-cage
protection.
Commercially available fiberboard-foil construction ("Corshield" by Conductive Con-
tainers Inc.) consists of aluminum foil sandwiched between layers of fiberboard. The fiber-
board has a naturally antistatic surface but is covered with an antistatic coating to seal in
Table 2. Current Carried by Tote Boxes.

RESISTANCE (0) CURRENT (mA)
TOTE BOX TYPE (OOV 200V IIOV 220V
Antistatic 2x H)9 2xl()9 5 x 10- 5 1 x 10"4

Carbon-loaded polyolefin (950 n at 1.5 V) 120 230
Corshield 5 x 10' 4x 10' 2 x 10"4 5 x 10"4
Vinyl-aluminum sheet >10 12 2x 10" <10-7 1 x 1()-6
228 APPENDIX
Table 3. Damage to MOSFETs by Grounded Operator

Touching Charged Tote Box.
VOLTAGE OF CHARGED roTE BOX
roTA!. MOSFETs
roTE BOX TYPE +SOO +1000 -8000 DAMAGED
Antistatic 0/1 0/5 1/5 1/15

Carbon-loaded polyolefin 1/5 2/5 3/5 6/15
Corshield 1/5 0/5 1/5 2/15
Vinyl-aluminum sheet 0/5 0/5 1/5 1/15
Table 4. Damage to MOSFETs

by Charged Operator
roTE BOX TYPE MOSFET DAMAGE
Antistatic 0/5
Carbon-loaded polyolefin 3/5 (15)
Corshield 0/5
Vinyl-aluminum sheet 0/5
Bare aluminum sheet 4/5 (35)
Table 5. Trlboelectrlc Charging Data.

V ON COUPON nCONCOUPON
roTE BOX TYPE FR-4 ACRYlIC AI. FR-4 ACRYlIC AI.
Antistatic 70 700 0 0.6 6.4 0.0

Carbon-loaded polyolefin 400 800 0 3.8 8.5 0.0
Corshield 200 200 0 1.0 2.9 0.1
Vinyl-aluminum sheet 200 200 800 3.3 2.6 2.5
sulfur-containing impurities that might tarnish silver-plated leads. The foil provides su-
perior Faraday-cage protection. For example, a discharge from a key held by a person
charged to 8000 V caused no damage to field effect transistors (MOSFETs), whereas an
impractically heavy wall (0.140 in.) of carbon-loaded polyolefin allowed transistor dam-
age (Table 1).
Even the Tesla coil test was passed when the Corshield box was folded to give two layers
of foil (Table 1). In previous work (Ref 5), only constructions with heavy foil or metal
screen passed this test. Note that the Tesla-coil test is the worst case in electrical stress and
positioning of the device in the box. However, this test is not the worst case in statistical
significance because only five parts are tested for a "pass" rating. Furthermore, our Tesla-
coil test does not use worst-case acceptance criteria because subtle damage may not be seen
by the curve tracer, and the MOSFETs used are sensitive to 100 to 200 V whereas some new
devices may be affected by only 20 volts. If a cost-effective material can pass this test, we
would use that material.
PAPER NO.7 229
Fiberboard-foil costs less than other materials. Also, it can be stored easily as flat sheets
and then folded into boxes when needed. Its only major defect is its limited durability, but
heavier fiberboard will probably prove sturdy enough for most applications.
The vinyl-metal sheet design should satisfy the market niche requiring extreme durabil-
ity. The metal, either steel or aluminum, provides high structural strength and is coated on
both sides with tough vinyl, e.g., 0.010 in. thick, by either lamination or powder-coating.
The resulting nonconductive surface is safe for people. Also, in a contrived charge-device
model test (Table 6), the nonconductive surface was even safer for devices than an antistatic
surface.
The relatively heavy metal wall, 0.0375-in. aluminum (20 guage), is a virtually impreg-
nable Faraday cage and suppresses the voltage of a static charge, no matter how much the
surface may be stroked, so that the box never has an appreciable E field when the metal is
grounded via bare metal feet on the bottom. In our test, the effectiveness of draining off
charges onto an antistatic bench-top was better for a vinyl-metal sheet box than for a
carbon-loaded polyolefin box (Table 7).
However, good contact with the bench surface, aided by flatness of the bottom of the
box, can be critical. The slightly flexible Corshield box benefited from being conformable
and lying flat, whereas the rigid antistatic or carbon-loaded polyolefin boxes were slightly
"dished" (concave) so that only edges or corners made contact.
The nonconductive vinyl surface's only defect is that it can triboelectrically charge con-
ductors, whereas an antistatic or conductive surface cannot (Table 5). However, charging of
nonconductive surfaces, e.g., conform ally coated circuit-board modules, seems a more im-
portant issue, and nonconductive vinyl was less of an offender than carbon-loaded polyolefin
(Table 5).
Test Methods
Below are the various test methods that were used to derive the results shown in each of the
tables accompanying this article.
Shielding/Discharge Test
(Results in Table 1)
An electrode 1.5 in. square was taped against the inside surface of one wall of the tote box and
connected to the substrate-case lead of a Motorola 2N4351 MOSFET, and a similar elec-
trode was taped against the inner surface of the bottom of the box and connected to the gate
Table 6. Charged-Device Model Test

TOTE BOX TYPE MOSFET DAMAGE
Antistatic 2/5 (IS)

Carbon-loaded polyolefin 2/5 (OS)
Corshield 1/5 (IS)
Vinyl-aluminum sheet 0/5
Bare aluminum sheet 2/5 (IS)
230 APPENDIX
Table 7. Drain Time Test.

v (NEGATIVE)
'J'O're BOX TYPE AFI'ER I SEC AFI'ER 5 SEC
Antistatic 8000 8000

Carbon-loaded polyolefin 5000 900
Corshield 2500 600
Vinyl-aluminum sheet with bolt heads for feet 900 600
lead. Then a discharge was made to the outer wall of the box over the electrode. This dis-
charge was from a charged person holding a key, from a capacitor connected to a resistor
and a steel probe, or from a Tesla coil operated for 30 sec. The box sat on a grounded plate
during the test.
Current Carried by Tote Boxes Test

NFPA 56A electrodes were placed 1 in. apart on the box surface, and the resistance was
read with a Beckman Model L-IO megohm meter.
Damage to MOSFETs by Grounded

Operator Touching Charged Tote Box
Test (Results in Table 3)
A grounded operator held the substrate-case lead of a MOSFET and touched the gate lead
to the charged tote box resting on a nonconductive plastic stand-off.
Damage to MOSFETs by Charged

Operator Test (Results in Table 4)
An operator charged to +1000 volts held a MOSFET (as in Table l) by the substrate-case
lead and touched the gate lead to the grounded tote box.
Triboelectric Charging Test

1.5-in.-square coupons of uncoated aluminum or FR-4 epoxy circuit-board laminate, un-
coated or coated with acrylic conformal coating, were shaken in the tote box for 30 sec and
then dropped into a Faraday cup or measured with a static field meter. All charges were
positive.
PAPER NO.7 231
Charged-Device Model Test

The capacitor (1300 picofarads) representing a charged circuit board was FR-4 epoxy lami-
nate, 0.096 x II x 15 in., copper-plated on both sides with I-in. unplated borders. The
substrate-case lead of a MOSFET (as in Table I) was connected to the lower side of the
capacitor, which was suspended by nonconductive twine. Then the upper plate was charged
to +1000 volts, and the gate lead of the MOSFET was touched to the grounded tote box
being tested.
Drain Time Test (Results in Table 7)

The tote box, suspended by nonconductive twine, was charged to -8000 volts, placed on a
melamine-fonnaldehyde laminate table top (1010 il/sq.) for either 1 or 5 sec, and lifted
again; then the field on the box was measured with a static meter.
Summary
Table 8 summarizes our evaluations. Together, the two new multilayer boxes should satisfy
all in-plant handling needs for an ESD-control program. These boxes are able to afford
secure Faraday-cage protection for even the most sensitive items when electrically continu-
ous lids are being used.
TableS.
TOlE BOX TYPE
CARBON- CARDBOARD- VINYL-

LOADED ALUMINUM METAL
CHARACTERISTIC ANTISTATIC POLYOLEFIN FOIL SHEET
ESD shielding Poor Poor Excellent Excellent

Drain time Fair Good Very good Excellent
Triboelectric charging
Nonconductors Low High Low Low
Conductors None None Almost Some
none
Danger to devices
Box charged Low Moderate Low Low
Operator charged Low High Low Low
Danger to people Low High Low Very low
Permanence (ESD) Poor-fair Excellent Excellent Ex.cellent
Durability (physical) Good Good Fair Good
Cost Moderate High Low High
232 APPENDIX
Cost advantage depends on many factors, most notably the number of units produced
and the fabrication method. In general, a vinyl-aluminum sheet box would be competitive
with an injection-molded carbon-loaded polyolefin box. A vinyl-steel sheet box, though
cheaper than aluminum, would be almost three times heavier in the same gauge. In contrast
with the above choices, the fiberboard-foil box is inherently inexpensive (Table 8).
Last of all, fiberboard boxes (Corshield) are commercially available at this time: how-
ever, vinyl-metal sheet designs are still in the prototype stage.
Another interesting multilayer design is one which utilizes an aluminum screen or foil
sandwiched between layers of hard vulcanized fiber (Table 1). The vulcanized fiber mate-
rial is naturally antistatic, even at low humidity (5 x 1011 O/sq after seven weeks' storage at
nop and 12% RH). As in the case of the vinyl-metal sheet box, the commercial success of
this design would depend upon the development of a practicable fabrication method, but
both appear to be excellent alternatives for future ESD control.
References
1. J. M. Kolyer and W. E. Anderson, "Pennanence of the Antistatic Property of Commercial Anti-
static Bags and Tote Boxes," Reliability Analysis Center EOSI.ESD Symposium Proceedings,
EOS-5, Las Vegas, NY (1983): 87-94.
2. J. M. Kolyer, W. E. Anderson and O. E. Watson, "Hazards of Static Charges and Fields at the
Work Station," Reliability Analysis Center EOSIESD Symposium Proceedings, EOS-6,
Philadlephia, PA (1984): 7-19.
3. R. O. Enoch and R. N. Shaw, "An Experimental Validation of the Field-Induced ESO Model,"
Reliability Analysis Center EOSIESD Symposium Proceedings, EOS-8, Las Vegas, NV (1986):
224-231.
4. G. C. Holmes, P. J. Huff and R. L. Johnson, "An Experimental Study of the ESO Screening
Effectiveness of Antistatic Bags," Reliability Analysis Center EOSIESD Symposium Proceed-
ings, EOS-6, Philadelphia, PA (1984): 78-84.
5. J. M. Kolyer and W. E. Anderson, "Perforated Foil Bags: Partial Transparency and Excellent
ESO Protection," Reliability Analysis Center EOSIESD Symposium Proceedings, EOS-7, Min-
neapolis, MN (1985): 111-117.
PaperNo. 8

Symposium, Orlando, Florida, September 29-October I, 1987, Sponsored by
EOSIESD Association and ITI Research Institute, EOSIESD Symposium Proceedings,
EOS-9, 1987, page 41.
ELECTROSTATIC DISCHARGE (ESD) CONTROL

IN AN AUTOMATED PROCESS
John M. Kolyer, Ronald Rushworth, and William E. Anderson

Autonetics Electronic Systems
Sensors and Aircraft Systems Division
Anaheim, CA 92803
Summary
In the robot-based, computer-controlled installation of devices such as OP-AMPs and CMOS
on printed circuit boards, an exaggeratedly sensitive model assembly served to locate ESD
hazards and later monitor the process for their absence. This "coupon" approach applies
uniquely to automation with its freedom from unpredictable operator error.
Introduction
The objectives were (l) to reveal ESD hazards in an automated, robot-based manufacturing
process for mounting components on printed circuit boards (PCBs), (2) to remove or con-
trol these hazards, and (3) to monitor the process for continued ESD safety.
The process and the ESD-sensitive components of concern will be described. Then ESD
damage mechanisms, hazards, and remedies will be reviewed, and possible hazards will be
related to process steps. Next, the "coupon principle" of using an exaggeratedly sensitive
model assembly to locate and evaluate ESD hazards will be explained and illustrated with
specific coupon designs. Finally, test data obtained with coupons will be discussed. Haz-
ards were eliminated to give an ESD-safe process which could be checked periodically by
means of fresh coupons.
233
234 APPENDIX
Description of Automated Process

In 1986, Rockwell International at Anaheim, CA, opened a new Automated Manufacturing
Cell (AMC) for placing and reflow-soldering surface-mounted components on PCBs to
build modules for guidance systems. The AMC (Fig. 1) is a highly automated, computer
integrated, robot-based manufacturing system comprising several work stations arranged in
a loop. An Automated Guided Vehicle moves around the cell to shuttle parts between sta-
tions. Materials are delivered to the AMC from an Automated Material System (AMS).
With reference to Fig. I, the assembly steps are: (1) removal of parts in tote boxes from
an automated carousel, (2) fluxing, tinning, cleaning, lead-forming, and kitting, (3) dis-
pensing of dots of solder paste at the rate of two per second, (4) dispensing of adhesive and
placement of flatpack, radial, and axial components at the rate of one per 18 seconds, (5)
curing of adhesive and baking out of volatiles from the solder paste, (6) vacuum oven bak-
ing and vapor-phase soldering, (7 and 8) cleaning to remove flux and other contamination
as well as solder balls, (9) manual installation of over-sized or heat-sensitive components
and also rework, and (10) inspection using sonic digitizers. Units 7 and 8 are connected by
a conveyor belt.
Work instructions and parts information are stored in a central, fault-tolerant computer
system. At each station, a laser bar code scanner is used to enter data. Necessary manufac-
turing data appear on cathode ray tubes (CRTs). This paperless system allows easy product
changeovers, and several products can be built simultaneously. The precision afforded by
automation reduces component scrap so that subassembly costs are minimized and quality
is maximized.
The relative humidity in the AMC room is controlled above a minimum of 30 percent (at
nOF). This reduces triboelectric charging but of course does not prevent it.
Coupon Principle
An automated process, being free of unpredictable operator error, uniquely lends itself to
a "coupon" approach in which an exaggeratedly ESO-sensitive model assembly is passed
through the process as a coupon to reveal ESO hazards. Various coupon designs used in
our tests all contained a 2N4351 MOSFET. The assumption was that if this MOSFET
showed no measurable damage, in terms of gate-source threshold voltage and shape of
the curve on a curve-tracer, there would be no subtle damage (not detected in quality
control tests) to devices on a real module. Undetected damage could lead to latent failure
and hence shorter lifetime and reduced reliability. Latent failures are known to occur with
CMOS devices.
A coupon is shown in Fig. 2, and a close-up of the MOSFETs on this coupon is shown in
Fig. 3.
The unedited paper describes how the coupons were tested before and after passing through
various process steps. ESO hazards revealed by the coupons were removed, and fmally
coupons were passed without damage through the entire process and the AMC was declared
ESO-safe.
@ ® CD
VAPOR
o o
O6 PHASE
SOLDER ING
AMC
\
7"
.:0-
--
ROBonc
O4 ASSEMB LY
WORK 7'
STAnON
SOLDER PASTE
APPLlC AnON
WORK STATION
~
;II
:u
~
011
Fig. 1. Automated manufacturing cell.

S!
238 APPENDIX
Fig. 2. Coupon.
Summary and Conclusions

1. Model assemblies (coupons) which were more ESD-sensitive than real PCB assem-
blies proved effective in locating ESD hazards in an automated assembly process. This
coupon approach was feasible because automation is free from unpredictable operator error.
2. Relevant ESD damage mechanisms were direct injection (01), the field induced model
(FIM), and damage by the static (8) field from the triboelectrically charged board itself
(FFB). The most sensitive coupons were damaged by 500 V by DI, 30,000 V at 12 inches by
FIM, and 700 V by FFB.
3. The FFB mechanism threatened coupons consisting of bare-board laminates with
mounted MOSFETs, but fields on the actual PCBs were minimized by voltage suppression
by metallic heat rails and ground planes.
Fig. 3. Close-up of encapsulated MOSFET (simulating a resistor) and

plugged-in MOFSET.
PAPER NO.8 237
4. Some proposed damage mechanisms never materialized. Triboelectric charged by

hot, blowing air was experimentally discredited, and flowing electronic heat-transfer liquid
in the vapor-phase soldering process did not damage the most ESD-sensitive coupons.
5. One coupon design was a simulated resistor for handling by robots in the component
preparation and adhesive-bonding steps, while other designs were simulated PCB assem-
blies for passing through the rest of the process. Designs differed in ESD-sensitivity, and
when a more sensitive design was damaged, a less sensitive, more realistic design was tried
because all the designs were far more ESD-sensitive than real PCB assemblies.
6. ESD damage to coupons led to corrective actions. These included eliminating syn-
thetic brushes, removing supplier-installed grounding wires mixed with brush bristles, and
making sure all machinery was grounded.
7. The test program culminated in successful passage of five replicate coupons through
the entire process. Monitoring would be done in the same way.
8. In conclusion, the empirical coupon approach is recommended for revealing ESD
hazards in automated processes and assuring safety after hazards are removed.
9. However, the coupon must be demonstrably more sensitive than actual components
to all pertinent ESD damage mechanisms. Introduction of a new component, or a change in
the process, could necessitate coupon redesign.
10. Finally, periodic checks with coupons are necessary because continued ESD safety
is not guaranteed. For example, a ground connection might fail and machinery parts might
be charged by electrical leakage.
Note: This paper was presented as an article in EOSIESD Technology Magazine,

August/September 1988, page 10. A good summary was included:
Controlling ESD with Coupons

by John M. Kolyer, a_lei llllhworth and WIlliam E. ~,
Rockwell International Corp., Autonetics Electronic Systems, ~_~...,.
Sensors and Aircraft Systems Div.
A comprehensive and detailed overview of
Rockwell International's use of coupons to detect and
manage EOSjESD hazards in an automated assembly facility.
Human unpredictability can cause many static problems, but
humans also monitor and control ESD. Here's how Rockwell
International uses coupons to detect and manage EOSjESD
hazards in an automated assembly facility where few humans
need go.
Etll,,,,'s Note: AI,b," IIUIO"""- bly 11M to IISSU" 111111 qUllllty de-
etlllSsembly ,limllUl'" "","y of ,II, ,p/It 0 mucII "'uc,tI IIu""," _k
p,06/,ms CIIUS,tI by ,II, mowmelll fDlU, Rockwtll 1"'HlUltlllllGl IIltd
11M mlS/obs of IIumllll bellllS. til, _tllods d"cribetl below. TII,y
,II,n nmllllll ,II, probltm of tI"ecl- "" btu,d 011 ,II, coupOll prlllcipl"
illl 11M cOII,ro/lill, ESD I" ,,., lib- w"",11I 0 ,ubtu_bly _ Itllll-
se/lCt of IIumllll supI/'V/II011. III II tlw '0 ESD tllllll ,1101' '0 be _
limltttl WilY, ESD COIItroi becomlS fllCtll"d Itllnds III f", til, IlIltH till,..
l1li t1u,cilt III nmolt Stllli/ll. III, qUIIlijiClltlOllIlM "9UIIlijiclltlOll
III ",tI" '0 spo, IIn11S of ESD vul- of til, f«lIity.
",,,,bill,y tlurilll lIu'omlltttlllssem-
PaperNo. 9
Presented at the 10th Annual Electrical Ovestress/ Electrostatic Discharge

Symposium, Anahein, California, September 27-29, 1988 Sponsored by
EOS/ESD Association and rrr Research Institute. EOSIESD
Symposium Proceedings, EOS-IO, 1988, page 99.
CORROSION AND CONTAMINATION BY ANTISTATIC

ADDITIVES IN PLASTIC FILMS
John M. Kolyer and Jack D. Guttenplan

Autonetics Electronic Systems
Sensors and Aircraft Systems Division
Anaheim, CA 92803
Abstract
One commercial brand of MIL-B-81705, Type n film contained organic acid and caused
corrosion of solder-coated device leads on circuitry. However, solderability was unaffected
in accelerated tests. Even acid-free antistats can stresscrack polycarbonate, fog instrument
mirrors, weaken adhesive bonds, and discolor epoxy paint. These problems are reduced by
a new generation of Type n films.
Introduction
Contamination of surfaces by antistats from antistatic films can cause (l) corrosion of sol-
der (when the antistat contains organic acid), (2) stresscracking of polycarbonate plastic,
(3) fogging of instrument mirrors, (4) loss of strength of adhesive bonds, and (5) discolora-
tion of epoxy-polyamide paint Other adverse effects, e.g., contamination of gyroscope balls,
have been discussed in the literature. I
Corrosion of Solder
In early 1987 a Government-Industry Data Exchange Program (GIDEP) Alert2 described
corrosion of solder-coated leads by traces of n-octanoic acid in one commercial brand of
"pink poly" antistatic film (MIL-B-81705, Type II). Therefore, we ran tests to determine
the amount of acid present and the nature of corrosion caused by it
In our tests, extraction of a suspect (acid-contaminated) bag with refluxing ethanol gave
acidity corresponding to 514 ppm by weight n-octanoic (caprylic) acid, in agreement with
400-650 ppm found by methylene chloride extraction by other investigators.3 The acidity
238
PAPER NO.9 239
of a sample of neat (undiluted) antistat from the manufacturer of the suspect film corre-
sponded to 8.9 wt. % n-octanoic acid, and other investigators3 found up to 15 wt. %. From
= =
our acidity data, the wt. % of antistat in the film (514) (100)/89,000 0.6, which is
roughly the level the manufacturer said had been added to the polyethylene by blending and
extrusion.
Our corrosion testing is summarized in Table I. Degradation of solderability has been
suggested2.3 but was not found in our tests. When painted with neat acid-contaminated anti-
stat, without aging, leads remained wettable by solder, and strength of solder joints was
normal. In long-term tests, circuit-board specimens and resistors with freshly tinned copper
leads were stored in suspect bags at elevated temperature and humidity or immersed in
moist antistat or moist n-octanoic acid at 120°F. The conclusion was that corrosion of pack-
aged items does occur but is only superficial (dull, gray tarnish) because so little acid reaches
the solder surface. There was no effect on solderability except in the drastic test with neat n-
octanoic acid where enough corrosion products built up to cause dewetting of solder. When
the acid content was very low, e.g., 0.05 wt. % as n-octanoic acid in the neat antistat, antistats
of the diethanolamide type ("proposed antistat" in Table I) were noncorrosive, and the
cosmetic problem of tarnishing disappeared. Most items already stored in acid-contaminated
packaging were left as is, but new items were packed in acid-free containers. Panicky
repackaging that could deform leads or cause electrostatic discharge (ESD) damage was
avoided.
Stresscracking of Polycarbonate
Polycarbonate, which is notably subject to stresscracking, was attacked by all antistats tested,
though to different degrees. Stresscrack agents for polycarbonate include commonplace
liquids such as ethylene glycol, heptane, and com oil in the literature4 and, in our tests, pink
liquid hand soap, No. 30 motor oil, and RMA (rosin, mildly activated) solder flux. There-
fore, it was not surprising that liquid antistats, which have considerable solvent power, were
also stresscrack agents. At 2.4% tensile strain on the surface of ~-inch-thick bent strips of
polycarbonate, ethoxylated tertiary amine or diethanolamide-type antistats caused cracks to
propagate completely through the plastic within 24 hours. At 0.3% strain, the antistats still
caused catastrophic cracking, but the time was much longer: 2 to 7 weeks.
When conventional MIL-B-81705, Type II films or new-generation "dry" films (dis-
cussed below) were rubbed on the surface of bent polycarbonate strips at 2.4% strain, the
new films caused significantly less cracking than did the "old." The explanation is that the
new films have minimal liquid antistat on their surface. Thus they may be safe when the
stress level on the polycarbonate surface is very low, but the cautious recommendation is to
avoid contacting polycarbonate items with any plastic that contains a liquid antistat.
Fogging of Instrument Mirrors

Fogging of instrument mirrors by outgassed antis tat during storage was observed and was
simulated by suspending an aluminum plate inside an antistatic bag in an oven at 160°F.
After one week, the plate was rinsed with acetone, and the solution was analyzed by infra-
red spectroscopy. It was found that about 20% of the original antis tat on the inner surface of
240 APPENDIX
Table 1. Corrosion Study.

GIDEP ALERT E9·A·86·02
ISSUED JANUARY 23,1987
• RESULTS
• ONE COMMERCIAL BRAND OF PINK POLY AN11STATIC FU.M - ACCELERATED TESTS
(MIL-B-81105, TYPE 0) CONTAINS TRACES OF N-ocTANOIC • HEAT ANTISTAT (SUSPECT BRAND) CAUSED DULLJNO/
ACID GRAYING, NO EFFECT ON SOLDERABILITY
• PURE n-oCTANOIC ACID FORMED HEAVY CORROSION
• n-ocTANOIC ACID REACTS WITH SOLDER, FORMING
PRODUCTS, SOME EFFECT ON SOLDERABILITY
TIN/LEAD SALTS
• FfIR ANALYSIS - CARBOXYLIC DERIVATIVES (ORGANIC
• SOLDERABILITY CAN BE REDUCED ACID SALTS) OF TIN AND LEAD
• PROPOSED NEAT ANTISTAT (ESSENTIALLY FREE Of
ORGANIC ACID) HAD NO EFFECT ON SOLDER OR
TESTING PROGRAM SOLDERABILITY
- LONG-TERM EXPOSURE TESTS

• 8O'l'o R.H. AT 100"F
• DESIGNED TO ANSWER FOLWWlNG QUESTIONS:
- SEALED AND UNSEALED BAGS
- IF OPERIJIONAL PARrS OR MODULES ARE STORED IN - 3-MONTHS EXPOSURE
nus MATI!RIAL, WR.L THERE BE PHYSICAL OR • AMBIENT HUMIDITY AND TEMPERIJURE
I!LI!CJ1t1CAL DAMACB, I.B., CORROSION, IMPAIRED -SEALED AND UNSEALED BAGS
SOLDBRABn.JTY, CONDUCTIVE PATHS? -I-YEAR EXPOSURE
- IF THERE IS DAMAOE, WHAT DISPOSmON SHOULD BE - LONG-TERM STORAGE TESTS

MADE OF PARrS AND MODULES PRESENTLY STORED OR • SUSPECT BRAND OF BAGS CONTAINS LOW LEVEL,
PREVIOUSLY EXPOSED? TYPICALLY SOD PPM, n-oCTANOIC ACID
• OTHER BRANDS ESSENTIALLY FREE Of ORGANIC ACID
• CORROSION EfFECTS IN SUSPECT BAOS LIMITED TO
• TEST SPECIMENS
DULLING/GRAYINO; NO EFFECT ON SOLDERABILITY
- PCB SPECIMENS, SOLDER-COATED CIRCUITRY BOTH SIDES • NO CORROSION/NG EFFECT ON SOLDERABILITY IN
OTHER COMMERCIAL BRANDS
- PCB MATERIAL COUPONS, COPPER-CLAD BOTH SIDES,
SOLDER-DIPPED OR SOLDER-PLATED AND Rl!A.OWl!D
- RESISTORS WITH COPPER LEADS, FRESHLY TINNED APPEARANCE OF

PCB TEST SPECIMENS AFTER
ACCELERATED IMMERSION TESTS
• ACCELERATED EXPOSURE TESTS
- PARTIAL IMMERSION AT 120"1' FOR 7 DAYS
• MOIST N-oct'ANOIC ACID • IMMERSION IN PURE n-ocTANOlC ACID
• MOIST NEAT ANTISTAT (FROM SUSPECT BRAND) SATURATED WITH WATER-HEAVY CORROSION
• MOIST PROPOSED NEAT ANTISTAT PROOUCT ACCUMULATION IN LIQUID
- HUMIDITY TEST (MIL-STD-202, METHOD 106) • IMMERSION IN SUSPECT BRANO HEAT

• PAINT ON THIN FILM ANTJSTAT - SLIGHT GRAYING IN LIQUID
- D-OCTANOIC ACID
• IMMERSION IN PRGPOSED NEAT
-NEAT ANTISTAT (FROM SUSPECT BRAND)
ANTJSTAT - NO EFFECT
- PROPOSED NEAT ANTISTAT
• CYCLE RM. TEMP. TO UO·F AT 90-98% R. H., TWICE
DAILY FOR \0 DAYS
ACTION TAKEN
• EVALUATION METHODS • PURCHASE OF SUSPECT BRAND SUSPENDED

- PCB SPECIMENS
• ADD SOLDER PASTE AND VAPOR PHASE SOLDER • PARTS AND MODULES PRESENTLY STORED: LEFT
• VISUALLY DETERMINE DEGREE OF SOLDER WETTING "AS IS" (SOLDERABILITY NOT AFFECTED;
REPACKAGING COULD CAUSE ESD DAMAGE, LEAD
- SOLDER-COATED COUPONS
DAMAGE, CONTAMINATION, ETC.)
• SGLDER DIP
• VISUALLY DETERMINE DEGREE OF SOLDER WETTING
• PARTS AND MODULES PREVIOUSLY EXPOSED; LEFT
- RESISTOR LEADS "AS IS" (NO PROBLEM OR FAILURE MODE HAS
• MEASURE SOLDERABILITY ON MENISCOORAPH BEEN ATTRIBUTED TO THIS EFFECT)
PAPER NO.9 241
the bag (as detennined by rinsing a fresh bag) had migrated as vapor to contaminate the
plate. The same level of transfer is estimated to occur in 6 months at room temperature,
based on a Cox chart which relates vapor pressure of organic compounds to temperature.
The solution to the fogging problem obviously is to use ESD-protective packaging mate-
rial free of fugitive, surface-seeking additives. Another example of the undesirability of
transferable antistats is the packaging of items such as resistors whose surface must main-
tain high electrical resistivity. In this instance the weakly conductive "sweat layer" that is
desired on the package surface becomes a liability on the surface of the stored item.
Loss of Strength of Adhesive Bonds

Intuition suggests that contamination of surfaces by "greasy" antistat would greatly reduce
the strength of adhesive bonds. An experiment was conducted using aluminum (alloy 2024
T3) treated with Forest Products Laboratories etch solution (30 parts by weight deionized
water, 10 parts concentrated sulfuric acid, and 1 part sodium dichromate) at 150°F for 20
minutes followed by rinsing with deionized water and drying at 150°F for 20 minutes. Lap-
shear specimens were prepared using a 2-mil (0.OO2-inch) bond line of Scotchweld EC2216
(3M Co.), which is a filled epoxy-polyamide adhesive. These bonded aluminum specimens
were cured at 155°F for 15 hours and tensile-tested by a standard method. 5 Before bonding,
the surfaces had been (1) left clean, (2) contaminated with a slight amount of antistat by
rubbing with an antistatic bag, or (3) liberally smeared with neat antistat. Bond strength
data (averages of four detenninations) are given in Table 2.
Failures were cohesive (within the adhesive) except as noted in the table. As shown by
the relative bond strengths, slight contamination by rubbing with the bags had a negligible
effect, and gross contamination by ethoxylated tertiary amine or diethanolamide antistat
has surprisingly little effect on bonding with one frequently used structural adhesive. How-
ever, other adhesives might behave differently and would have to be tested. Our result
merely shows that fugitive antistats do not necessarily cause severe bonding problems.
Table 2. Adhesive Bond Strength Data

AV.BOND STANDARD RELATIVE
CONTAMINANT STRENGTH, PSI DEV.,PSI BOND STRENGTH
None 2588 199 1.00

Antistat from rubbing with
suspect bag 2468 357 0.95
Antistat from rubbing with new-
generation "dry" bag 2583 100 1.00
Ethoxylated tertiary amine
antistat 1954 438 0.76*
Suspect amide antistat containing
8.9% n-octanoic acid 2196 132 0.85**
"About 10% adhesive failure (bare aluminum exposed).

""Trace, perhaps 0.1 %, adhesive failure.
242 APPENDIX
Discoloration of Epoxy-Polyamide Paint

Instrument panels painted with gray epoxy-polyamide paint (MIL-C-22750, Type I) were
discolored by prolonged contact with the acid-contaminated brand of MIL-B-S1705, Type
II film. The film had been pressed tightly against the surface, and wrinkles in the film
corresponded to a pattern of whitish streaks or marbling. Abrasive rubbing compounds
failed to remove these blemishes, and waxes did not fully restore gloss. Repainting was
necessary.
When drops of liquid were left on a painted surface for 24 hours at 120°F, n-octanoic
acid caused the most attack (swelling and softening of the paint), followed by the acid-
contaminated antistat, a low-acid diethanolamide antistat, and an ethoxylated tertiary amine
antistat. The explanation for this order of attack is that the acid was the best paint solvent,
the amide was intermediate, and the amine was the poorest solvent.
At room temperature, the acid-contaminated film caused staining in I day when a pres-
sure of 5 psi was maintained but no staining in the absence of pressure. A sample of new-
generation pink antistatic bubble-wrap, which had minimal antistat on the surface, caused
no staining even after 6 months at 5 psi.
Attack on paint probably is rarely noticed because the film is not usually tightly pressed
against the bagged item as chanced to happen in our case. Our temporary solution was to
use MIL-B-SI705, Type I, which of course is antistat-free.

1. One commercial brand of MIL-B-SI705, Type II film was contaminated with n-
octanoic acid and caused corrosion of solder, as revealed in a GIDEP Alert. However, our
corrosion tests indicated that the problem was only cosmetic because solderability was
unaffected; in fact, the acid resembles flux, which is also an organic acid. Therefore, pan-
icky repackaging was avoided in our operations. The issue is now past, because the offend-
ing product has been removed from the market.
2. Antistat-containing films are suitable for most purposes and have a long history of
success in ESD control, but they must be used with discretion because the fugitive, surface-
seeking additive can stresscrack polycarbonate, fog instrument mirrors, and discolor epoxy
paint. Adhesive bond strength also may be reduced, but the effect was less than expected in
tests with a commonly used epoxy adhesive.
3. Liquid antistats have considerable solvent power and can even attack rubber and
wire insulation, but in practice the damaging effects are limited by the small amount of
antistat transferred by rubbing or volatilization. For example, n-octanoic acid in sufficient
quantity will destroy solder completely, but a minimal amount of the acid contained in
rubbed-off antistat merely dUlled solder surfaces.
4. Antistat contamination problems are reduced, but not eliminated, by a new genera-
tion of "dry" MIL-B-SI705, Type II films with a minimal amount of "acid-free, amine-
free" amide antistat on the surface. By "acid-free" is meant negligible acidity, and
"amine-free" refers to a low amine content, e.g., 4-6 wt. % diethanolamine. Amines are not
corrosive to solder under conditions of use, and ethoxylated tertiary amine has long been
used as an antis tat.
PAPER NO.9 243
5. Assuming that their permanence6 proves adequate, the new-generation bags and
bubble-wrap will be a great impovement over the old "wet" or "greasy" products but are not
a panacea because the difference is quantitative, not qualitative. Packaging materials free of
liquid antistat still must be used in critical applications such as shipment or storage of poly-
carbonate items. Potential users should run tests to find the most cost-effective packaging
system.
Acknowledgment
The assistance of R. Rushworth, D. R. Violette, A. E. Carmellini, and T. J. Hester in per-
forming chemical analyses and corrosion tests is greatly appreciated.
References
I. M. K. Bernett, H. Ravner, and D. C. Weber, "Electroactive Polymers as Alternate ESD Protec-
tive Materials," EODIESD Symposium Proceedings, pages 115-119, orlando, Fl., 1982.
2. GIDEP Alert No. E9-A-86-02, on "Materials, Plastic, Antistatic," issued January 23, 1987.
3. J. Anderson, R. Denton, and M. Smith, "Antistatic Polyethylene Package Corrosion," EOSIESD
Symposium Proceedings, pages 36-40, Orlando, Fl., 1987.
4. Modern Plastics Encyclopedia, 1978-79, pages 528-529.
5. "Standard Test Method for Strength Properties of Adhesives in Shear by Tension Loading (Metal-
to-Metal)," ASTM D 1002-72 (reapproved 1983).
6. J. M. Kolyer and W. E. Anderson, "Permanence of the Antistatic Property of Commercial Anti-
static Bags and Tote Boxes," EOSIESD Symposium Proceedings, pages 87-94, Las Vegas, NY,
1983.
PaperNo. 10
Presented at the 11 th Annual Electrical Overstress/Electrostatic Discharge

Symposium, New Orleans, Louisiana, September 26-28,1989, Sponsored by
EOS/ESD Association and m Research Institute.
EOSIESD Symposium Proceedings, EOS-II, 1989, page 23.
CONTROLLING VOLTAGE ON PERSONNEL
John M. Kolyer, Donald E. Watson, and William E. Anderson

Autonetics ICBM Systems Division
Anaheim, California 92803
and Dale M. Cullop
Hanson Loran Chemical Co., Inc.

6700 Caballero Blvd.
Buena Park, California 90260
Abstract
Controlling voltage on personnel is a vital and demanding requirement in ESD protection.
New data show that the allowable resistance to ground must be lower than was thought.
Furthermore, wrist-strap systems often fail, so monitoring must be continuous; resistive-
type monitors are best. A cost-effective program is described.
Introduction
Controlling voltage on personnel has long been recognized as of primary importance in
preventing electrostatic discharge (ESD) damage to electronics. The reason for concern is
that operators' fingers often touch leads or contacts of ESD-sensitive (ESDS) devices, and
a flow of current from person to device can destroy or "wound" it by the direct injection
(01) mechanism. A "wound," passing undetected in routine tests, may result in latent fail-
ure. Thus, improperly grounded people are the principal ESD hazard or at least prime sources
of ESD for destroying parts.
A resistance of about I megohm is generally agreed upon for current-limiting resistors to
protect operators wearing wrist straps. However, questions remain: (I) What should be the
244
PAPER NO. 10 245
maximum allowable resistance to ground (ARTG) for protection of ESDS items as opposed
to protection of operators? (2) How can the triboelectric charging of operators be mini-
mized to make the resistance to ground less critical? (3) Should this resistance be checked
at intervals or continuously? (4) If a continuous wrist-strap monitor is used, which design is
best? (5) How can skin-voltage control and monitoring be made part of a cost-effective
program? We have attempted to answer these questions based on experience as well as
experimental data.
An operator sat at a workbench and charged himself either by stroking a sheet of Aclar film
on the bench surface or by shuffling his feet. The resulting data are shown in Fig. I. Fig. I
represents worst-case charging conditions, but surveys of operator grounding in manufac-
turing areas showed many bad cords and instances of high resistance to ground because of
operators' dry skin. Therefore, we recommend continuous monitoring with units like the
one illustrated in Fig. 2 and 3.
A Cost-Effective Program
Including Skin-Voltage Control
We emphasize basic, proven methods of ESD control and deemphasize methods which are
of questionable value and may even be part of the problem. The major elements in our
suggested program for ESDS items in Class I of MIL-STD-1686A are:
1. Operator Disciplines
A sufficiently skilled operator could work with almost no special equipment, and, con-
versely, the most expensive appliances won't prevent damage by an ignorant or careless
operator. Remember that a grounded operator is only the lesser of two evils versus an un-
grounded operator because a charged device could be damaged by rapid discharge to the
grounded operator's finger (the Charged Device Model or CDM). To maintain their skills,
operators must be retrained periodically.
2. Use of Test Equipment to Aid Operator Judgment

An ESD-educated operator needs access to basic test equipment including a field meter, a
continuity checker such as a Zapflash, and a surface resistivity meter. Several operators
might share these instruments.
3. Skin Voltage Control

As described in this paper, the ARTG of 10 megohms is met using continous wrist-strap
monitors and antistatic lotion.
l~ri-------------------------------------------------------------------------------------'
I
o • VOlTME IY PYT-SOO
• • VOLTME IY TlEK 512
( ) • MOSFET IWWiE
I2
I
'1
/
/
III
(2/10) 1 ( 5 1 5 ) "/
i 'OO (115) ~ ,
••••••••••••••••
o • ••••••••••••• , ••••••••••••••• " •••••••• V
(0/5) I (0/5) I'
I /
I /
1 • / /
/
/ SHOES, STATJC- ,r/"

LlMITJIIi FLOOR FINISH
~\I// ---r4. /.
/
/
100 1000
MEGOIfG
Fig. 1. Plot of triboelectric charging data.

PAPER NO. 10 247
CONTINUOUS MONITOR,
TWO-CONDUCTOR TYPE: DIAGRAM
.ALARM
\' "',,
\
,
' " ........... ~~T

--------- -,-"
Fig. 2.
4. Exclusion of Nonconductors
from Workstations
This precaution must be rigidly enforced, with frequent checks. A field meter and surface
resistivity meter are essential. Nonconductive surfaces which can't be eliminated may be
treated with topical antistat, necessarily chloride-free in some applications, and so labeled.
Bench tops should have static-dissipative surfaces and a buried ground plane to suppress
voltage.
5. Grounding of All Conductors

This is easily checked with a Zapflash. If a conductor, such as a microscope eyepiece hous-
ing, cannot be conveniently grounded, the operator must satisfy himself that it never be-
comes dangerously charged. The best policy is to ground even small-area conductors.
6. Control of Electrical Fields

Paper No.3 gives safe distances for loo-volt-sensitive devices from surfaces at various
apparent voltages as determined by a field meter. The operator should run frequent checks
on his static-safe workstation to locate charges. Then, if these can't be eliminated, he will
work a safe distance from them.
248 APPENDIX
Fig. 3. Two-conductor continuous monitor with strap.
7. Minimization of Triboelectric
Charging of Operators
Static-limiting floor finish is very cost-effective as seen in Paper No.6. Choose a brand
based on its ability to limit tribroelectric charging of operators wearing shoes with various
types of soles. We recommend a "walk test" with a dozen of your personnel: the voltage on
them is measured with a field meter or with a Voyager PVT-300 Personnel Voltage Tester.
Surface resistivity of the finish should be at most 10" ohms/square, but note that surface
resistivity does not correlate with triboelectric charging for materials in general and corre-
lates imperfectly for floor fmishes. No matter what finish is used, operators should not
fidget or shuffle their feet unnecessarily; there is no sense in "looking for trouble."
PAPER NO. 10 249
8. Use of Ionizers
Be cautious with room ionization as mentioned above. Cleanrooms, by the way, are a
special case; an ionization system, correctly engineered, may reduce the number of airborne
particles. Use local ionizers only for a clearly defined purpose, such as to control charge
buildup in processes like grit-blasting. Never use ionization as a vaguely conceived "safety
net."
9. Humidification
In very dry areas, humidification is desirable because it makes antistatic materials with
"sweat layers" function better and it reduces triboelectric charging. But don't let it build
false conficdence, and beware of corrosion or other problems as mentioned above. As with
ionization, the use of humidity control must be judicious.
10. Static-Safe Packaging

We recommend "Faraday-cage" packaging with aluminum foil as shielding, because inde-
pendent studies agree that foil is the only adequate protection from external fields and dis-
charges for highly ESD-sensitive items. When away from a static-safe workstation, an ESDS
item must be in a Faraday-cage container such as a tote box for in-plant handling or a
sturdy, suitably cushioned Faraday-cage package for shipment.
11. Elimination of Unnecessary Equipment

Our recommended program is cost-effective because it emphasizes operator skills coupled
with basic, clearly useful equipment. Money is saved by eliminating marginally valuable
appurtenances such as special smocks, footwear, chairs, or floors. "Buy the basics, forget
the frills." Besides being a waste of capital, some frills may be counterproductive by instill-
ing false confidence so that operator disciplines are relaxed.
12. Program Administration

Of course, good record-keeping and day-to-day management are essential. Operators
should be encouraged to keep records, e.g., of hazards such as E fields that they're en-
countered, and to make suggestions. Incidents of hardware mishandling should be dis-
cussed with the group. Operators will monitor their individual workstations by using test
equipment as recommended above. With alert, knowledgeable operators and an emphasis
on the basics, ESD control becomes a simple matter except for silicon wafer processing,
paint-spraying, or other processes which involve triboelectric charging beyond the
operator's control. Static-limitation must be designed into such processes, e.g., by using
conductive solvents or blowing ionized air over surfaces. Automated processing is, of
course, out of the operator's hands by definition and requires special approaches such as
the use of ESDS coupons to locate hazards. In large companies, a Program Coordinator
should be appointed to lead a working group, conduct monthly meetings, brief upper
management, etc.
250 APPENDIX

1. An operator, grounded through various resistances, triboelectrically charged himself
by shuffling his feet or by stroking a plastic film (representing a garment sleeve) on a work-
bench surface. Devices sensitive to 80-100 V were damaged by being touched by the op-
erator at a little over 70 megohms resistance to ground in the foot-shuffling tests and a
little over 7 megohms in the plastic film tests. An explanation for the difference involves
rapidity of voltage rise, which is greater for the thinner dielectric (film versus shoe soles).
2. Considering such factors as the methodology of the tests, the effect of relative hu-
midity, and the fact that women's shoe soles tend to be thinner than men's (as used in the
test), we recommend an allowable resistance to ground (ARTG) of 10 megohms. This is the
maximum, and a suitable minimum for most work is 1 megohm, so as to limit current to
well under 1.0 mAo When high voltages are being handled, ground fault circuit interruptors
will guard against the remote possibility of an electrically overstressed resistor being car-
bonized so that it falls below a safe resistance.
3. To minimize triboelectric charging, operators shouldn't fidget or shuffle their feet
unnecessarily. Also, a static-limiting floor finish is effective; for example, the voltage on an
ungrounded, foot-shuffling operator was reduced from 3300 V with an ordinary finish to
180 V with the special finish. With the special finish, the ARTG could have been increased
by a factor of about 8, because the charging of the operator was so low even when he was
ungrounded. However, the ARTG must be set assuming normal floor surfaces.
4. The static-limiting floor finish remained effective, though less so, after severe scuff-
ing and scraping to simulate wear. We advise selecting a durable brand of finish which is
easily maintained by standard procedures.
5. One-conductor wrist straps show poor reliability. In fact, checks on wrist straps in
electronic assembly areas disclosed not only a disturbing rate of cord failure but many
instances of excessive resistance to ground caused by loose wrist bands or operators' dry
skin. Clearly, monitoring is essential.
6. This monitoring should be continuous because ESD damage can occur in nanosec-
onds and an ESDS item can be "zapped" during the interval, however brief, between peri-
odic checks. All hardware handled since the last successful check of a failed strap could be
subject to a material review action and might even have to be scrapped! Furthermore, cords
sometimes fail in an intermittent way which could go undetected in periodic checks but be
revealed in continuous monitoring. We highly recommend these monitors, but some users
have been troubled by features such as the audible alarm. Potential buyers should run trials
in their own assembly areas.
7. Our preferred monitor design is the two-conductor resistive type which checks ground
connections as well as the operator. Also, the two conductors provide redundancy (if one fails,
the other keeps working), and the monitor is not fooled by a high-capacitance object touching
the strap. The property of interest-resistance-is measured directly. Antistatic lotion is often
required for dry skin, and the monitor gives notice when the lotion must be renewed.
8. This paper sketches a cost-effective program, including skin-voltage control, for pro-
tecting items in sensitivity class 1 of MIL-STD-1686A. Elements of the program include
operator disciplines, test equipment for operators' use, continuous monitoring of skin volt-
age, exclusion of nonconductors, grounding of conductors, control of E fields, minimiza-
tion of triboelectric charging of operators, judicious use of ionization and humidification,
static-safe packaging, cost-saving by eliminating unnecessary equipment ("Buy the basics,
forget the frills"), and astute program administration.
Paper No. 11
This paper is reprinted, by permission, from EOSIESD Technology,

October/November 1990, pages 18-23.
ESD-CONTROL MYTHS, OLD AND NEW
Rockwell Intemational Corporation

Defense Electronics
San Bernardino, California
By "myth" we mean a more or less widely held belief that sounds credible but proves false
when examined critically. Many such beliefs have clouded ESD-control standards. First,
we'll discuss some old myths, then some new ones.
Among old ESD-control myths, now put to rest in most people's minds by long repeti-
tion of the truth, are the following:
Wrist straps render human skin harmless. The truth is that a grounded operator
versus an ungrounded operator is only the lesser of two evils because of the Charged
Device Model (CDM). A charged device may discharge to the grounded operator's fin-
gers and be damaged. If the operator is ungrounded and chances to be at the same poten-
tial as the device, no damage will occur. So grounding, in this unlikely case, would be an
"evil."
Components are safe once they're mounted on circuit boards. Circuit lines may act
as antennas to increase the likelihood of ESD damage and discharges caused by the Field
Induced Model (FIM). Also, more than one device may be damaged by Direct Injection (01)
and electrical overstress (EOS) when a single discharge occurs to a branched circuit line.
Carbon-loaded plastic provides a Faraday cage. Carbon-loaded plastic, which was
the earliest ESD-protective material (in the sense of not holding triboelectric charges) is not
conductive enough to shunt a direct discharge safely away from an ESD-sensitive item
inside a bag. Instead, the discharge acts through the wall of the bag and can cause damage if
the device leads are in a vulnerable position.
High humidity solves all ESD handling problems. Triboelectric charging is merely
reduced by high humidity, not eliminated. Operator disciplines cannot be relaxed when
humidification is provided. Humidification is expensive and may introduce problems such
as corrosion or operator discomfort.
Ionization allows the use of plain, nonconductive plastics. This statement is untrue
because nonconductive plastics can become triboelectrically charged and cause ESD dam-
age by the FIM before the ions have time to neutralize the charge.
Ionization is a substitute for grounding of personnel (by wrist straps, for
example). This egregious idea was actually promoted by an equipment supplier several
years ago. The fact is that operators can become triboelectrically charged (by shuffling their
feet, for example) and damage ESD-sensitive items with 01 by touching them despite a
shower of ions that neutralize their skin relatively slowly.
251
252 APPENDIX
Devices with protective networks are not ESD-sensitive. Protective networks raise
the threshold for ESO damage to a limited extent, such as from 100 V to 2,000 V by the
Human Body Model (HBM), not infinitely. In other words, the sensitivity class in accor-
dance with MIL-STD-1686A may be raised (from Class 1 to Class 2 in the previous ex-
ample), but the device will remain ESO-sensitive. In our system of ESO control, the
precautions wouldn't change because we treat all ESO-sensitive classes the same as a mat-
ter of convenience; maintaining "islands" of different disciplines within a plant is impractical.
So much for old myths. Unfortunately, a second generation of myths has arisen, created
partly by vendors' biases. We'll debunk several of these new, less obvious, and more insidi-
ous untruths. The following discussion includes our own biases, of course, but these are
based on data and not influenced by sales concerns. Readers are welcome to contact us c/o
ESDlEOS Technology on points of disagreement.
Background Ionization Is needed when highly ESD-sensitive Items are being
handled. This statement suggests a cure-all or panacean view of ionization, but ioniza-
tion is not a cure-all. In fact, ionization for no valid, well-defined reason can be useless and
actually become part of an ESO problem by creating unbalanced charges on ungrounded
conductors. The possible occurrence of such charging has been noted for an electrically
unbalanced AC ionizer (Ref. 1). Hence, ionizer manufacturers have shown much concern
about balancing which in many models is now automatic.
In contrast to humidification, which reduces triboelectric charges by inhibiting them in
the first place, ionization can neutralize a charge only after it has formed. Thus, ionization is
a process of correction rather than prevention. However, it does have its place in ESO
control when used judiciously. Local ionization helps control triboelectric-charging pro-
cesses, such as peeling tape from a roll, and room ionization does control standing charges
on common plastics and garments. Also, room ionization may have an incidental air-clean-
ing effect in cleanrooms, but potential users should run tests or at least review published
data to prove that vendors' claims of reduced particles are true.
While malfunctioning electrical ionizers could charge isolated conductors as we men-
tioned earlier, nuclear ionizers have a possible problem of polonium 210 contamination.
One brand of nuclear ionizers was recalled by the Nuclear Regulatory Commission in Feb-
ruary 1988. Actually, this danger is remote. In terms of cost, nuclear and electrical ionizers
may be similar; you must make your own calculations.
In sum, ionization is not a harmless vitamin pill but strong medicine that has possible
dangerous side effects and should be used only when risks are carefully controlled and
outweighed by benefits.
The resistance-to-ground for personnel may be 100 mn or more because charges
drain otT even at very high resistance as shown by steady-state tests. The fallacy of
this statement is that a steady-state test is unrealistic. The voltage spikes on seated operators
being triboelectrically charged under real conditions, such as shuffling feet or brushing a
garment sleeve over the work surface, must be measured. We did these measurements and
found that MOSFETs were damaged at only about 8 mn resistance-to-ground when an
operator stroked a plastic film, representing a sleeve, on a workbench under conditions
where the individual generated -3,800 V when ungrounded (Ref. 2). The implications of
such data are serious, and we recommend an allowable resistance-to-ground of 10 mn
maximum (with the usual 1 mn minimum for safety). The maximum resistance should be
controlled with a continuous wrist-strap monitor (see the next myth).
Occasional checks of wrist straps are sufficient. Logically, this statement is inde-
fensible because almost everyone agrees that 01 from operators' fingers is a major, or
PAPER NO. 11 253
even the major, ESD-damage mechanism, and it is well known that this damage occurs in
a matter of nanoseconds. Therefore, we can't drop our guard for even a fraction of a
second-monitoring of skin resistance-to-ground must be continuous. Even if wrist straps
were checked every hour (the real interval is usually daily or even weekly), when a failed
strap is discovered, all the ESD-sensitive items handled by the wearer during the last hour
are suspect. The only reasons continuous monitoring is illogically avoided are its per-
ceived cost and trouble. But, in reality, this monitoring is practical. More and more con-
tinuous monitors are being used in industry, especially on government programs, and we
have found certain models to be quite cost-effective. We prefer the two-conductor resis-
tive rather than the capacitance type.
Note that dryness of operators' skin may raise the resistance-ta-ground far above 10 mO,
even with a perfect wrist strap and good skin contact. The monitor also reminds operators to
apply antistatic lotion, usually once or twice during a shift.
All ESD damage is caused by a spark. As a bare, unqualified statement, this is mean-
ingless. The first problem is semantic: What is a "spark"? Is it a visible, audible discharge?
If so, an ungrounded operator's finger at 1000 V will damage a MOSFET, for example, by
touching a lead even if there is no spark visible or audible.
Even if a spark is defined as "any discharge causing damage by 01," electrical fields can
act through space to destroy MOSFET's by the FIM with no spark at all. In the latter case,
the spark occurs inside the device, but this definition of spark as "an internal discharge"
results in a useless truism that all ESD damage is caused by discharges inside devices,
which gives us no guidance in preventing these internal discharges. We want to know what
happens outside a device to cause the "spark" inside. Therefore, a more productive state-
ment is: "All ESD damage is caused by fields (FIM) or discharges (01, CDM)."
The only useful idea in the "spark theory" is that a rapid discharge may cause damage
under conditions where a slower discharge won't (see the next myth).
Conductive surfaces are dangerous; antistatic surfaces are safe. We object to this
generalization, though it contains more than a grain of truth. MIL-HDK-773 strongly
favors antistatic surfaces as being safer for ESD-sensitive items, and our own data sup-
port this. For human safety, conductive surfaces are obviously undesirable because they
can carry a lethal current (over I milliampere). However, even static-dissipative surfaces
can cause ESD damage to sensitive devices such as MOSFET's by the CDM, and avoid-
ing conductive surfaces completely is impracticable because tweezers, soldering iron tips,
and operators' fingers are ubiquitous in workstations. Our approach, which we consider a
prudent compromise, is to use a mix of antistatic and conductive surfaces, with the latter
being minimized.
If handling procedures are correctly developed and followed (admittedly this is a big
"if'), conductors will be grounded and ESD-sensitive items will be uncharged so the CDM
can't happen with either a conductive or an antistatic surface. In other words, we're avoid-
ing the trouble of covering fingertips with antistatic cots and so on by relying on operator
skills to prevent 01 and CDM, which can still happen with antistatic surfaces, though to a
much lesser extent. A relatively (not completely) "forgiving" work environment in terms of
discharges, with all antistatic or static-dissipative surfaces, seems unfeasible to us, and damage
by fields (by the FIM) could still happen.
Antistatic surfaces (of DIP tubes, for example) do not triboelectrically charge pack-
aged ESD-sensitive items by rubbing against them. In 1983, we were surprised to find
that antistatic polyethylene (MIL-B-81705, Type II) charged epoxy-glass circuit-board
material, and researchers at another major corporation were skeptical at first but verified
254 APPENDIX
our results (Ref. 3). The truth is that the contamination effect (antistat transferring to the
rubbed surface so that liquid separates from liquid with minimum charge generation) greatly
reduces triboelectrification but does not prevent it. An antistatic material whose surface
layer of antistat has been removed by volatilization or wear will not benefit from the con-
tamination effect nor will permanently antistatic materials with dry surfaces. Furthermore,
packaging materials that depend on a fugitive antistat will hold a charge when the antistat is
sufficiently depleted.
Triboelectric charging is quite material-specific, so tests should be run with the particu-
lar packaged item and packaging material (both fresh and aged) in question.
Highly conductive surfaces (of metals, for example) do not participate in triboelec-
tric charging and are safe for ESD·sensltive items. Nothing could be further from the
truth! Conductors only seem not to be triboelectrically charged because when they are
grounded, the charge runs off almost instantaneously. An isolated sheet of aluminum or
steel takes a charge of equal magnitude and opposite polarity to the charge taken by a
plastic sheet stroked against it. In fact, grounded metals and carbon-loaded, conductive
plastics are effective in charging nonconductors and causing damage by the COM, which
can occur during the charging process with no discharge when, for example, an intense,
uneven field is produced on a circuit board with mounted devices (we have called this the
field-from-board or FFB damage mechanism). For example, epoxy-glass circuit-board lami-
nate is readily charged by sliding on stainless-steel shelves in an oven, and people are charged
by walking with insulated, rubber-soled shoes on a conductive floor. Ironically, such a floor,
which was designed for use with conductive footwear, actually can make matters worse
with ordinary shoes.
Conductive flooring is always beneflcial for ESD control. We've just mentioned
that a conductive floor may be the opposite of beneficial when ordinary, nonconductive
shoes are worn. We say "may be" because shoe sole materials vary greatly in charging
propensity. Rubber and certain vinyl compositions tend to give especially high charges.
Another problem with conductive floors is safety. In one of our tests, operators with wet,
leather-soled shoes (as they might have been ifthey'd come in out of a rainstorm) showed a
resistance-to-ground low enough for a dangerous current of over I milliampere to pass
through their bodies if they touched a 110 V hot lead. Ground-fault interrupters may be
needed when such situations could occur. The point is that special footwear must be used
with conductive floors, and the safety issue must be considered. Only then are conductive
floors beneficial.
Conductive chairs and special smocks are essential In a rigorous ESD·control
program and allow some operator disciplines to be relaxed. Conductive chairs are
possibly dangerous to both people and devices, insofar as a lethal current might be carried
or the COM might be encouraged by the conductive surface, but we do not reject them on
that basis alone. Our main reason not to use conductive chairs is that they are unnecessary;
ordinary vinyl-upholstered chairs are not a hazard according to our measurements of
electrical-field strength in the work area on the bench.
Though vinyl does exhibit static charges, the fields from them are sufficiently attenuated
by distance and the shielding effect of the operator's body to be harmless at the bench
surface where ESO-sensitive items are being handled. This setup assumes that operators
will not bring the items near the upholstery, as one of our operator disciplines specifically
forbids. Thus, we are depending on operator skill to avoid the cost of expensive, foolproof
equipment, but operators are always the critical factor no matter how much money is thrown
at the problem of ESO control.
PAPER NO. 11 255
As for smocks, the shielding effect of fabric with widely spaced conductive threads is
dubious and, in any event, isn't needed with tight-fitting shirts or blouses with short or
rolled-up sleeves. Proximity to the skin suppresses fields on the cloth, and distance to the
operator's hands, in the normal working position, further weakens fields to the point of
being harmless. You can make measurements with a hand-held field meter to support or
refute this conclusion in your own assembly area. Another consideration is cost. Smocks
are good uniforms for raising ESD awareness and impressing visitors, but they are rather
expensive to collect, clean, and reissue.
No equipment justifies relaxing operator disciplines. Operators are the "captains of the
ship," and ineptitude on their part can sink the most lavishly capitalized ESD-control pro-
gram. Forget expensive frills and invest in operator training.
The hair hazard is eliminated by area ionization. Here's another case of expecting
too much from ionization. As stated earlier, area (room) or any other kind of ionization
should be used only for clearly defined purposes, and controlling static charges on hair is
not one of them. Long hair will constantly recharge as it swings around, and the resulting
field could cause damage by the FIM before air ions neutralize the charge. The best course
is to require hair to be tied back so static voltages on it are suppressed by nearness to the
skin.
Use a hand-held meter to verify that the field in the vicinity of ESD-sensitive items is
harmlessly weak, even with the operator's head bowed for close examination. We use the
"Charge-Distance Rule," an empirical relation between the apparent voltage on a surface
and the safe distance from that surface (Ref. 4). The equation is d = ~~:, where d is the
distance in inches and V is the apparent charge in volts measured vith a hand-held meter.
For example when V is 1,000 V, d is 18 in.
Ordinary floors in ESD-controlled areas must always be resurfaced or specially
treated. This assertion seems reasonable at first, but concrete floors, especially new ones,
often contain enough moisture to be weakly conductive. The presence of salts in the con-
crete also may help. Sealers may be soapy and afford low triboelectric charging of ordinary
shoe soles, as does the thin layer of antistat on floors coated with a static-limiting floor
finish. Therefore, concrete floors should be checked for surface resistivity if conductive
footwear is used, or triboelectric charging of people (which correlates imperfectly with
surface resistivity) should be checked if ordinary shoes are worn.
Soiled. uncoated vinyl-asbestos tile, though unattractive in appearance, may also impart
relatively low charges to people walking on it. In general, evaluate the floor before you
consider altering it. If nothing else, baseline data will show the improvement obtained and
help, in retrospect, to justify the cost of resurfacing (with conductive tile, for example) or
treating (such as with static-limiting floor finish).
All cellulosic materials change from antistatic to nonconductive at moderately low
relative humidity (15 to 20%). We believed this myth until our testing disclosed a coated,
plasticized cellophane (regenerated cellulose) that remains antistatic down to about 4%
relative humidity at room temperature. (For further information, see EOS/ESD Technology,
OctoberlNovember 1989, p. 9.)
Antistatic masking tapes are ESD-safe. As with many myths, the fallacv of this one
lies in oversimplification. Yes, antistatic masking tapes are safe in the sense that they will
not show a field because incipient charges drain to the worker's hand, while any field from
the nonconductive adhesive layer is suppressed by the weakly conductive tape layer. How-
ever, when the tape is pulled from a nonconductive surface (for example, epoxy-glass
circuit-board laminate), that surface may be charged enough to damage mounted devices by
256 APPENDIX
the FFB mechanism. The point is that charging of the substrate, as well as the tape itself,
must be considered before generalizing that the tape is ESD-safe.
Short drain times of charges from objects, such as tote boxes, deposited on work
surfaces is of major importance and justifies expensive laminates and grounding
schemes. Our experience and judgment indicate that situations where short drain time is
critical are rare. A commonly cited scenario in support of short drain time is a grounded
operator reaching into a charged tote box at rest on a bench and touching a sensitive device
lead. Damage occurs by the CDM because the body of the device is charged by induction
along with the tote box. However, such an incident is unlikely because the grounded opera-
tor usually will touch the box and discharge it before reaching in. In fact, reaching in and
touching a device lead without touching the box wall might be a difficult trick. We're satis-
fied with a reasonably short drain time without taking extreme measures, such as using a
conductive surface, which also has dangers, to minimize the time. Drain time has always
been a secondary consideration, not a controlling factor, in our selection of work surfaces
and grounding techniques.
Overkill protection Is necessarily expensive. This myth may stem from the idea that
you get what you pay for. However, overkill can come cheap. An example is a stone house
built near a quarry; the stone will last for thousands of years but is cost-competitive with
wood, which decays in a lifetime. Similarly, aluminum foil is a far more effective ESD
shield than see-through, lOO-angstrom-thick metallization but is similar in cost.
A laminate bag consisting of wire screen sandwiched between layers of antistatic poly-
ethylene is indeed overkill, resisting even 35,000 V from a Tesla coil, but this packaging at
a few dollars for a 10 x 12 in. bag is not really expensive when a $100.000 module for a
vital aerospace or defense application must be protected. Continuous wrist-strap monitor-
ing. discussed earlier, also may seem like overkill-though logically it is a necessity-but
costly failures or material review actions are avoided. In fact, you can't afford not to have it.
Latent ESD failures are commonplace. Perhaps we should call this one a "sus-
pected myth" because no one is sure. A consensus among leading device experts, whom
we questioned two years ago, was that latent failures are real but rare. In 1988, British
Telecom researchers stressed components to 90% of their HBM threshold and then sub-
jected them to accelerated testing to look for premature (latent) failures; however, they
found none (Ref. 5). Of course, this negative result doesn't rule out latent failures under all
conditions; any negative generalization is difficult if not impossible to prove.
Intuition supports the latent failure concept when we picture stressed devices as "walking
wounded" because wounded soldiers indeed may die early, but devices are not human
nor wounded in the sense that flesh can be. Intuition often uses false premises and is
wrong-though it could be correct by accident in this instance. The point is that skepticism
is in order until strong evidence for latent failures accumulates. But even the possibility of
a few latent failures in critical equipment, such as missile guidance systems or life-supporting
medical appliances, is reason enough for a rigorous ESD-control program.
Myths, old and new, are oversimplifications or bad generalizations that tend to lull be-
lievers into a false sense of security while they rely on a product or approach that has fatal
shortcomings or is needlessly expensive. For example, someone concerned only with "sparks"
would ignore the FIM and FFB, and someone relying on ionization as a cure-all could have
ESD-sensitive items damaged by poorly trained operators. In stating the truth as we see it,
our object has been to dispel irrational fears, roused by some vendors' scare tactics, as well
as foolish complacency. We have suggested when to spend and when to save because throwing
money at the problem in the wrong places (for expensive, unnecessary equipment) and
PAPER NO. 11 257
misguided thrift (refusal to buy continuous wrist-strap monitors) are both counterproduc-
tive. Buy the basics, forget the frills.
Our opinions are often controversial but supported by realistic tests as well as experience
and common sense. They form the basis of a flexible, cost-effective system of ESD control
that has proven itself over years of practice (Ref. 6).
References
1. GIDEP Alert H7-A-85-02, 1985.
2. Kolyer,1. M., D. E. Watson, and W. E. Anderson. "Controlling Voltage on Personnel," EOSIESD
Symposium Proceedings, EOS-ll, 1989, p. 23.
3. Private communication with I. R. Huntsman and D. M. Yenni, 3M, 1983.
4. Kolyer, I. M., W. E. Anderson, and D. E. Watson. "Hazards of Static Charges and Fields at the
Workstation," EOSIESD Symposium Proceedings, EOS-6, 1984.
5. Woodhouse, I., and K. D. Lowe. "ESD Latency: A Failure Analysis Investigation," EOSIESD
Symposium Proceedings, EOS-IO, 1988. p. 47.
6. Kolyer, I. M., and D. E. Watson. ESD From A to Z: Electrostatic Discharge Controlfor Elec-
tronics. Van Nostrand Reinhold, New York, NY, 1990.
Paper No. 12
This paper is reprinted, by permission, from

Evaluation Engineering, September 1991, pages 115-119.
ESD TESTING OF SILICON WAFERS
John Kolyer and Donald Watson

Defense Electronics
As a typical ESD-sensitive device, the Motorola 2N435 I Metal Oxide Semiconductor Field-
Effect Transistor (MOSFET) has made an excellent "white rat" for our research. 1 However.
this destructive testing has drawbacks: the cost of the MOSFETs as well as annoying delays
when more pieces must be ordered. Now we have found that hundreds or even thousands of
tests can be run on a single silicon wafer.
The wafer is P-type silicon coated with 1,000 A of oxide. With either MOSFETs or
wafers, a voltage differential, which may be momentary, punctures the oxide in a micro-
scopic dielectric strength test.
The wafer is 4" in diameter and 0.020" thick and has a resistivity of 5 to 60 n'cm,
tvpicallv 15 to 25 n·cm. The oxide thickness is 800 to 1,200 A, nominally 1,000 A. This
type of wafer serves to monitor sodium contamination in furnaces where the oxide is grown,
so the upper surface holds a hundred sputtered aluminum dots for making electrical mea-
surements to detect sodium.
To make electrical contact with the underlying semiconductive silicon, one side of the
wafer was broken off and solder drops (labeled "s" in Fig. I) were applied 2" apart on the
exposed edge and pressed by spring-loaded contacts with terminals A and B. Resistance
between A and B was I x lOs n at 10 V.
The probe, electrically isolated with Teflon™, was held by a pivoted wooden arm which
could be adjusted for reach or radius as well as rotated. The procedure was to start at maxi-
mum radius and sweep the wafer at successively smaller radii. At first, an attempt was made
to utilize the aluminum dots, but the best procedure, which provides an almost unlimited
number of test points, was to avoid the dots and probe directly on the oxide. The steel probe,
0.112" in diameter with the tip polished to a radius of O.OOS", was springloaded to press the
oxide surface with a force of 0.6 lb (Fig. 2).
Resistance across the oxide layer was determined with a Beckman L-12 Megohmmeter
with shielded leads (RG-58IU coaxial cables) at 10 V. The low or negative megohmmeter
lead was connected to terminal T (which was connected to probe terminal P) while the high
or positive megohmmeter lead was connected to A. Undamaged oxide read approximately
5 x 1012 n, with the reading influenced by capacitance of the wafer apparatus. Damaged
oxide almost always read less than 5 x 10 10 n, usually in the lOS to 109 range but sometimes
lower. Slight degradation, e.g., 5 x 1011 n, rarely occurred. Intermittents occasionally oc-
curred, as with damaged MOSFETs.
258
PAPER NO. 12 259
Fig. 1. Wafer test setup.
CURRENT FLOW ~
PROBE
ESD I
EVENT
r
.. : ....\.,:.........................:.... .
.....-
~-- P - TYPE SILICON
s s
(solder) (solder)
Fig. 2. Diagram of silicon wafer test.

260 APPENDIX
For comparison tests, MOSFETs were similarly measured with the low megohmmeter
lead to the substrate-case lead of the MOSFET and the high megohmmeter lead to the gate
lead of the MOSFET. Fresh MOSFETs read over 1013 0, and degraded MOSFETs usually
read less than 1010 0, but subtle damage, e.g., 2 x 1012 0, occasionally occurred. The arbi-
trary criterion for failure was a resistance of less than 1 x 1012 O. This is more rigorous than
our former definition of failure as a shift of more than 0.1 V in gate-source threshold volt-
age, which allowed considerable damage to the gate oxide. 2
The difference between the wafer and a MOSFET gate was leakage current. For ex-
ample, when the voltage across the gate of a MOSFET was slowly raised, the resistance fell
to 2 x 1011 0 at 102 V (0.5 nA current) and 1 x 1010 0 at 110 V (11 nA) before failure
occurred and the resistance fell several orders of magnitude. In a similar test, the wafer
oxide began to leak (2 x 1012 0,0.05 nA) just before failure at 107 V.
In MOSFETs, perhaps the major leakage path is over the surface of the chip to the nearby
source and drain metallization. This leakage is significant in Charged Device Model (COM)
tests involving charge drainage from a pin to a surface. However, in testing where the rise
time is fast, gate leakage does not protect the MOSFET and it is successfully modeled by
the wafer.
The voltage sensitivity of the wafer apparatus was checked with a modification of the
Human Body Model (HBM). 3 Terminal A was grounded, and a 1,500-0 resistor connected
to a charged l00-pF capacitor was touched to terminal T. MOSFETs were tested by simi-
larly touching the gate lead with the substrate-case lead grounded. Both wafer and MOSFETs
began to fail at about 80 or 100 V (Table 1) as expected for 800 to 1,000 A of silicon oxide.
The following four examples illustrate that the wafer method correlates with the MOSFET
device tests. The examples deal with:
• Safe distance of ESO-sensitive devices from charged surfaces.

• Maximum allowable resistance to ground for operators.
• Hazard of small, charged, conductive objects on the workbench.
• Superiority of foil-containing ESD-protective packaging material.
Example I. To demonstrate the Field-Induced Model (AM) for ESO damage, a I ftl
aluminum plate on a nonconductive standoff was connected to terminal P, and terminal A
was grounded. AcIar™ film (9" x 12") triboelectrically charged to -15,000 V was slowly
moved toward the plate to induce a charge. In repeated tests, when the film was approxi-
mately 40" from the plate, the oxide on the wafer failed as noted by a burst of RF heard from
an AM radio tuned between stations. This result is consistent with our Charge-Distance
Table 1. Voltage Sensitivity Test.

FAILURE RATE (FAlLEDrrnsTED)
VOLTAGE PROBE ON WAFER 2N4351 MOSFET
--60 0/5 0/5

-80 115 0/5
-100 1/5 115
-120 2/5 2/5
-140 SIS 4/5
-160 SIS SIS
PAPER NO. 12 261
(CD) Rule, which specifies a conservative safe working distance of 68" from surfaces charged
to 15,000 V.2
Example 2. In tests for allowable resistance to ground (ARTG) for personnel, an un-
grounded person with a capacitance of approximately 200 pF stroked the Aclar™ film on a
carpet. The film was quickly lifted, which charged the person to -4,000 V as read by a field
meter on an electrically isolated aluminum plate connected to the individual's skin. The
charging process was repeated with the person grounded by a wrist strap with a I-MQ
resistor and a finger pressed against T (A was grounded). The failure rate (number
failed/number tested) was 0/10. However, when the resistance in the strap was increased to
15 MO, the failure rate was 6/10. This result is consistent with previous data using MOSFETs.4
The explanation is that the voltage surge or spike exceeded 100 V and damaged the oxide
when the resistance to ground was too high; the charge on the person could not be drained
fast enough. In retrospect, the test could have used a wafer instead of MOSFETs to demon-
strate this principle and set the maximum ARTG at 10 MO.
Example 3. The setup in Fig. 1 modeled a circuit-board module with the lead from T
to P representing circuitry leading to the oxide and the wafer representing circuitry beyond
the oxide. The capacitance on the probe side was determined by charging T to -1,000 V
with the probe resting on Teflon and reading the charge with a Keithley Model 617 Pro-
grammable Electrometer. The charge was 12 nCo Therefore, C = QIV = 12 pF. By this
method, a ceramic capacitor labeled 100 pF held 103 nCo On the wafer side, C was 34 pF. A
discharge to T represented the so-called floating model in which an ungrounded ESO-sen-
sitive item is touched by a charged surface.5 The question was: Are low-capacitance con-
ductive objects, such as small screwdrivers, at comparatively low voltages really a hazard?
To obtain an answer, the objects in Table 2 were touched to a power supply lead, and then
to T. Alternatively, the probe was rested on Teflon, and the gate lead of a MOSFET was
connected to P by means of a jumper while the substrate-case lead was connected via a
jumper to B. The jumpers raised the capacitance from 12 to 14 pF on the probe side and
from 34 to 36 pF on the wafer side. The voltages in Table 2 are at the moment of contact.
For example, the ceramic capacitor was touched to the power supply lead at -1,000 V but
lost 50 V, by leakage or by attracting air ions, during the one-second transfer to T.
The data in Table 2 show the usual statistical scatter, but the wafer and MOSFET tests
agree that small conductive objects are indeed hazardous at about 500 V/5 pF or 3,000 V/l
pF, even with considerable capacitance (12 pF) in circuitry leading to the oxide layer. The
inverse situation is the COM in which charged DIPs with a capacitance of only 1 to 3 pF are
damaged by discharge from a lead to a workbench or other surface. s
Table 2. Floating Model Test.

CAPACITANCE, pF,
BYELECI'ROMETER FAILURE RATE (FAILEOImSTED)
CAPACnuR METIiOD VOLTAGE PROBE ON WAFER 2N43S1 MOSFET
Ceramic, rated 3.3 -480 2/5 0/5

at 3.3 pF -950 3/5 4/5
Small screwdriver with 3" 4.5 -450 3/5 1/5
blade and plastic handle, -900 5/5 2/5
held in grounded hand
I" nail embedded in rubber, 0.9 -1,000 0/5 0/5
held in grounded hand -3,000 3/5 4/5
262 APPENDIX
Example 4. A capacitive probe was constructed to two O.75"-diameter brass disks,

0.009" thick, separated by 0.5" of nonconductive plastic. The inner conductors of l' lengths
of RG-58/U coaxial cable were soldered to the disks and the cable shields were grounded.
The probe was placed inside an ESD-protective bag which was taped to a grounded alumi-
num plate so that the bag material was stretched tightly over the upper disk. The lead from
the upper disk was connected to T while the lead from the lower disk was connected to A. A
200 pF capacitor charged to -3,800 V was touched to the bag above the upper disk, and then
the probe was disconnected and the wafer oxide was tested as usual.
Table 3 shows the results, which agree with the MOSFET data in similar testS. 2•6 The
antistatic polyethylene and black conductive plastic bags were not sufficiently conductive
to prevent a damaging electrical pulse between the disks. The nickel-coated bag was super-
ficially conductive, but the spark discharge removed the metallization, resulting in a bare
spot 0.01" wide. Only the foil-containing bag succeeded in preventing a momentary poten-
tial of 100 V or more across the oxide.
Table 3. Packaging Test.

FAILURE RATE (FAu.EDtrESTEDl
BAG BY PROBE ON WAFER
Antistatic polyethylene (MIL-B-81705, Type ll), 0.006" wall 5/5

Black conductive plastic, 0.004" wall 5/5
Metallized polyethylene with 100 Aexternal nickel 5/5
Laminate containing 0.00025" aluminum foil 0/5
Summary
In conclusion, the wafer test duplicates results with actual devices when the rise time is fast.
Data are somewhat scattered, just as with MOSFETs. The practical object is to obtain a
failure rate of 0/5 in a worst-case test to establish safe working conditions and to select
ESD-protective packaging.6 A MOSFET test has been recommended for packaging specifi-
cations, but the method of Example 4 would do as well.·
References
1. J. M. Kolyer and O. E. Watson, ESD from A to Z: Electrostatic Discharge Control for Electron-
ics, Van Nostrand Reinhold, 1990, pp. 21 and 96.
2. J. M. Kolyer, W. E. Anderson, and O. E. Watson, "Hazards of Static Charges and Fields at the
Work Station," EOSIESD Symposium Proceedings, 1984, p. 7.
3. OoO-HDBK-263, May 2,1980, p. 24.
4. J. M. Kolyer, O. E. Watson, and W. E. Anderson, "Controlling Voltage on Personnel," EOSIESD
Symposium Proceedings, 1989, p. 23.
5. O. J. McAteer, Electrostatic Discharge Control, McGraw-Hili, 1989, pp. 176 and 189.
6. J. M. Kolyer and W. E. Anderson, "Perforated Foil Bags: Partial Transparency and Excellent
ESO Protection," EOSIESD Symposium Proceedings, 1985, p. 111.
PaperNo. 13
This paper is reprinted, by permission, from Evaluation Engineering,

October 1991, pages 11 0-117.
CDM AND WORK SURFACE SELECTION

Defense Electronics
Work surfaces for ESD control must be conductive enough to allow charges to drain from
conductive objects, yet resistive enough to restrict the current flow to a safe level in terms
of the Charged Device Model (CDM).I An unsafe level of current flow has been equated
with sparking. 2•3 A study was undertaken to determine the best compromise between drain
time and CDM protection based on device damage data, not mere intuition.
Using a Beckman L-12 Megohmmeter set at 10 V with the high or positive lead con-
nected to the grounding lug of l-ft2 panels or mats, or clamped to the edge of bag samples,
the resistance to ground was measured in two ways:
• With the low or negative lead of the megohmmeter connected to a dual in-line package
(DIP) square-tipped lead 0.015" wide and 0.009" thick (terminal D in Fig. I) pressing
the surface (S) with a force of 0.008 lb.
• With the low lead of the megohmmeter connected to a National Fire Protection Asso-
ciation (NFPA) 99 electrode (5 lb., 2.5" diameter, aluminum foil over rubber) resting
on the surface.
Eleven surfaces were tested with the following results (resistance in ohms through D is
given in parenthesis):
• Aluminum alloy 6061-T4 sheet (approximately 10 0, but varied from 3 to 1,100 0).
• Carbon-loaded conductive polyolefin bag (l x l()!l 0).
• Hard laminate with conductive fibers (5 x 1()!l 0).
• Bag with external nickel (2 x 101 0).
• Rubber mat (I x 109 0).
• Special vinyl mat (6 x 109 0).
• Three hard laminates (2 x 1010 0, 5 X 10 10 0, 7 X 1010 0).
• Antistatic polyethylene (MlL-B-81705, Type II) (1 x 1011 0).
• Painted steel as used for office furniture (5 x 1012 0).
Except for the aluminum and the nickel-coated bag, the NFPA 99 values are plotted
against the D values as shown in Fig. 2.
263
264 APPENDIX
""--.11',
If.
\
\
\
\
IJ2
I
CODE: D • DIP LEAD SOLDERED TO WIRE w

B • TEFLON BLOCK ON WHICH D RESTS
S • SURFACE UNDER TEST
T • REMOTE TERMINAL FOR PROBE
P • PROBE TERMINAL
W • WAFER
s • SOLDER
A,B • WAFER TERMINALS
J1, J2 • JUMPERS
G • GATE LEAD OF MOSFET
SC • SUBSTRATE-CASE LEAD OF MOSFET
Fig, 1. CDM apparatus.
The CDM was modeled using the setup of Fig. 1 as explained in Ref. 4. To represent a
charged device, the switch was closed, and the powered (negative) lead from a calibrated
power supply (Power Designs. Inc., Model 3KlOB) was connected through an 8-MO resis-
tor to terminal A before turning on the power. With the power supply turned on before
connecting it to terminal A, the 1,000 A oxide layer on the wafer and a MOSFET with its
leads shunted were damaged. This illustrates the fact that shunting does not necessarily
provide ESD protection. The MOSFET is tested by moving the probe to a Teflon™ surface
and using the jumpers as shown in Fig. 1.
With the assembly charged, the power supply lead was disconnected from terminal A,
the switch opened, and block B (0.25" thick) removed to let D fall to the grounded surface
S under test. As shown schematically in Fig. 3, capacitor C2 (representing the capacitance of
the wiring from terminal T to P and the probe itself when contacting the wafer) then drained
into the test surface through resistance R 2, while C. (representing the capacitance from the
oxide layer through terminal A, the switch, the wire w and to terminal D) was insulated by
the wafer oxide or by the gate oxide of the MOSFET. The l4-pFcapacitanceofC. is worst-
case for a large DIP; ordinary DIPs are 1-3 pF at the work surface. s The RC time constant
for C2 was as expected, given measurement uncertainties, e.g., approximately 3 s for anti-
static polyethylene vs 1.7 s calculated, with the fall in voltage at terminal A being moni-
tored with a Trek Model 512 Field Meter.
After 5 s, the switch was closed and terminal T was touched with a grounded I50-MO
resistor to discharge the setup before measuring the resistance of the wafer oxide or MOSFET
PAPER NO. 13 265
gate with the megohmmeter. Failure was judged by a significant resistance drop as ex-
plained in Ref. 4.
The wafer oxide failed when the voltage at C2 decreased by more than 100 V, indicating
that the dielectric strength of the oxide was exceeded. Even with the resistance through D
(Rz in Fig. 3) at 4 X 1012 n, the wafer oxide finally failed at 20 s when drainage of CI was
allowed to continue. At the arbitrary drainage time of 5 s, only painted steel was safe among
the surfaces tested.
In contrast to the wafer oxide, MOSFETs were partially protected by leakage current, as
explained in Ref. 4. The failure rate (number failed/number tested) decreased with increas-
ing resistance R2• The rate did not approach zero until approximately 1010 n at 500 V/14 pF,
1011 nat 500 V/214 pF, or 1012 nat 1,000 V/14 pF as shown in Fig. 4. The failure rate
reflected no sudden "sparking point" as the resistance decreased; rather, there was a smooth
rise in the probability of failure. Without leakage current, MOSFETs would be damaged
even by antistatic polyethylene at 500 V/14 pF, as is the wafer oxide. But, in fact, as the
resistance through D increases, the leakage rate exceeds the drainage rate; therefore, the
MOSFETs are not damaged.
By shunting the wafer with a lOS n resistor between terminals P and A, wafer failure
began abruptly below lOS n as expected. Thus, the shunted wafer can serve for testing
purposes as an idealized 50%-failure-rate MOSFET as shown in Fig. 4.
The conclusion drawn from Fig. 4 is that the optimum resistance through D is in the 1010
to 1011 n range, corresponding to an NFPA 99 measurement of 107 to lOS n for the data
points in Fig. 2. The rubber mat with a resistance through D of 1 x 109 n (l x 10«' n by
NFPA 99) was judged too conductive (approximately 40% MOSFET failure), while the
~ 10
f
M
81
~
u.
z
1)
i
~
C!) 6
g
§ 5
III
Jj
I
a: 4
'0
J
3
4 5 6 7 8 9 10 11 12 13
Log of Resistance to Ground through 0, Ohms
Fig. 2. Correlation of resistance to ground data.

266 APPENDIX
12pF lor Probe on Warer. 14 pF lor MOSFET

34pF lor Probe on Warer, 36 pf lor MOSFET
105 otmls (rrom A to B. Figure 1)
Resistance to Ground through 0 (Figure 1)
Total Resistance rrom Olide 10 Ground
Fig. 3. Schematic for CDM apparatus (Fig. 1) with switch open and D resting on S.
vinyl mat at6 x I()'I n was borderline, The hard laminate at 5 x 0 1 ' was clearly unsuitable,
but fortunately most hard laminates for ESD control give 1010 to 1011 n through D as desired.
Incidentally, the CDM danger of conductive bags lying on the work surface is apparent
from Fig. 4.
Though the preferred hard laminates are safe for MOSFETs at 500 V, damage does occur
at 1,000 V, especially at higher capacitance levels, as shown in Fig. 4 and reported previ.
ously.6.7 Therefore, eve n these surfaces must be considered possibly hazardous todevices in
Class I ofMIL·STD·1686A. If drain time is not an issue, a nonconductive surface with an
underlying conductive layer for voltage suppression is best.'
Skin resistance varies greatly but was on the order of 107 n through 0, corresponding to
a MOSFET failure rate of 60%. 'Therefore, a grounded operator is the lesser of two evils vs
an ungrounded operator.' This COM hazard can be alleviated by wearing antistatic gloves
or finger cots or avoided altogether by wearing nonconductive latex gloves or cots.' How·
ever, the wearer of nonconductive gloves or cots must remember that charged objects can·
not be discharged by touching as they nonnally are, and even small charged objects at
relatively low voltage, e.g., a screwdriver at 5 pF and 500 V, can cause ESO damage by
direct injection or the floating model.'
PAPER NO. 13 267
Tests also were run with 1% tolerance, 1/20-W thin-film resistors (Class 1 of MIL-STD-
1686A) shunted between terminals P and A in Fig. l. At 3,000 V and C1 = 214 pF, a hard
laminate (5 X 1010 0 through D) caused less than 0.01 % resistance shifts for 10-0, 100-0,
lO-kO, or 3OO-ill resistors. However, the carbon-loaded conductive bag shifted the resis-
tance of the 3OO-k 0 resistor by -0.05%, and the aluminum surface caused shifts in all the
resistors to a maximum of -0.38%. The conclusion is that the CDM is not a major hazard
for thin-film resistors with I % or even 0.1 % tolerance when the work surface is a hard
laminate with 107 to lOS 0 resistance to ground by NFPA 99.
ICs or discrete devices usually will not exceed 500 V when near the work surface be-
cause of voltage suppression. In fact, the inputs of most ICs now are protected by zener
diodes to at least 1,000 V. Therefore. the major concern may be for Class I devices mounted
on circuit boards that have relatively high capacitance and also high voltage because a large
board tilted with respect to the work surface escapes effective voltage suppression. Accord-
ingly modules should be checked with a field meter and, if necessary, neutralized with
5/5 ////////////////////////////////////~"""""""'~
Shunted Wafer. ~ '~

500orl000V. ~ \.;
12pF / ~
•
~ \.;
/ ~
4/5
~ Unshunted Wafer. ~
~ 500 or l000V. ~
~ 12pF ~
500V ~ \.;
/ ~
~ \.;
/ ~
3/5 () ~ () ~
~\ \.;
____ ~%..:ai~re_ _ _ _ _ _ _ /'~ _ _(OOOV ~
,~ ~
•
~ ~
215
•
MOSFETpata
500V.14 pF
.\\~ () ~
~
,
\
() 1000V. 14 pF \ ~
D 500V. 214 pF
'. :-.;
1/5 ~~ . \r500V.
\~ 214pF
··
2 3 4 5 6 7 8 9 10 11 12 13
Log of Resistance to Ground through D, Ohms
Fig. 4. MOSFET and wafer failure rates.

268 APPENDIX
ionized air before a sensitive tenninal or edge contact touches the bench surface or the
operator's fingers.
In conclusion, while supporting the popularity of work-surface laminates with an NFPA
99 resistance to ground of 101 to lOS 0, the data make the important point that these surfaces
are only relatively CDM-safe. They are nonsparking, but discharges can cause ESD dam-
age without an evident spark. For example, in many cases, D fell to surface S and damaged
MOSFETs with no RF heard from an AM radio tuned between stations. Therefore, operator
disciplines must assure that the ability of the surface to drain charged objects does not
become a liability by damaging voltage-sensitive devices.
References
1. Carlton, D., "Selecting an ESD Workstation," Evaluation Engineering, January 1991, p. 94.
2. "New Test Proposed for Checking ESD Safety of Materials," Compliance Engineering, Fall
1990, Vol. VII, Issue 5, p. 77.
3. "Electrostatic Discharge Protective Packaging," MIL-HDBK-773, April I, 1988.
4. Kolyer, J. M. and Watson, D. E., "ESD Testing With Silicon Wafers," Evaluation Engineering,
September 1991, p. liS.
5. McAteer, O. J., Electrostatic Discharge Control, McGraw-Hill, 1989, p.176.
6. Kolyer, J. M., Anderson, W. E., and Watson, D. E., "Hazards of Static Charges and Fields at the
Work Station," EOSIESD Symposium Proceedings, 1984, p. 7.
7. Kolyer, J. M., Anderson, W. E., and Watson, D. E., "Tote Box Material: How Good Is It?"
EOSIESD Technology, OctoberlNovember 1987, p. 13.
8. Kolyer, J. M. and Watson, D. E., ESD from A to Z: Electrostatic Discharge Control for Electronics,
Van Nostrand Reinhold, 1990, p. 121.
9. Kolyer, J. M. and Watson, D. E., ESD from A to Z: Electrostatic Discharge Control for Electronics,
Van Nostrand Reinhold, 1990, p. 46.
PaperNo. 14
This paper is reprinted, by permission, from EOSIESD Technology,

February/March 1992, pages 27-28.
IS YOUR WORK SURFACE CDM SAFE?
John M. Kolyer

Defense Electronics
The charged device model (CDM), in which charge on the circuitry of an ESD-sensitive
device flows from a contact, lead, or terminal to a benchtop or other surface, has become
the focus of increasing concern. The flow has been found to create sufficient voltage differ-
ential or current within the device to puncture gate oxide, degrade thin resistor films, or
cause other overstress damage. I Because of this, MIL-HDBK-773 warns against conduc-
tive surfaces.
Obviously, a highly conductive surface, such as metal, is unsafe by the defmition of the
CDM as a discharge-caused failure model. Conversely, a nonconductive surface is assumed
to be safe because discharge is avoided. However, some surface conductivity is desired to
drain static charges from objects such as tote boxes. A tradeoff is needed; there must be a
surface-to-ground resistance high enough to suppress the CDM while not overextending
the drain time.
One study found that the resistance to ground for work surfaces was correlated with the
CDM failure rates of MOSFETs sensitive to lOOV. (These MOSFETs are more easily dam-
aged than most, but not all, devices).2 The conclusion was that the resistance through a lead
that contacts the surface should be above 1010 ohms, because the MOSFETs were protected
by a 1010 ohm leakage path. When the resistance through the lead was 109 ohms, for ex-
ample, leakage around the gate oxide of the MOSFET was not fast enough to prevent a
damaging voltage differential of lOOV.
As seen in Fig. I, a conductive rubber mat and a laminate with conductive fibers were
unsafe, while painted steel was safe but nonconductive. Between these extremes are the
best surfaces: hard laminates resistive enough to prevent CDM damage yet sufficiently
conductive to drain charged objects.
How can you evaluate your own work surface? Here are three methods:
1. The conventional method measures resistance to ground with a 5-lb., 2.5-inch-diameter

NFPA 99 electrode connected to the negative (low) lead of a megohmmeter. The positive
(high) lead is connected to the grounding lug of the work surface. At 10 V the resistance
should be 10'_108 ohms as seen in Fig. l. Note that results may be misleading for
unhomogeneous surfaces because the large electrode cannot detect small conductive spots
or fibers that a device lead might find. For example, a laminate with conductive fibers on
269
270 APPENDIX
4 5 6 7 8
Log of Resistance to Ground through

NFPA 99 Electrode, Ohms
80
Brlghl
(I mAl
Z"
'# c:
oj ~
iii :>
60 CC 2-
I!!
~
~
I-
S
.r::
.~ Dim
W CO
LL (I l1li)
lJ) ,Q
0 :>
::E CO
40 c:
g
Z Very Dim
(0.3l1li)
Dark
(O.lI'A)
20
Log of Resistance to Ground through

DIP Lead, Ohms
•• •• •
2 3 4 5 6 7 8 9 10 11 12
LAMINATE SKIN CONDUCTIVE HARD PAINTED

WITH RUBBER LAMINATES STEEL
CONDUCTIVE MAT
FIBERS
Fig. 1. MOSFET failure rate by CDM versus resistance to ground and neon bulb brightness.
the surface gave a resistance through a device lead of 5 x 10' ohms with the lead resting on
a fiber vs. 5 x 1011 ohms between fibers.
2. The method discussed in Ref. 2 measures resistance at 10 V through a dual in-line
package (DIP) lead (0.015 x 0.009 in. in cross section) pressing the surface with a force of
0.008 lb. The lead is shifted from place to place to identify the lowest resistance, which
should be above 1010 ohms. (See Fig. 1.)
3. A neon bulb test was evaluated using the DIP lead with the specified force, because
pressure on soft surfaces (e.g, a rubber mat) greaty affects resistanceY An N-2 neon bulb
with maximum striking voltage of 90 V and design current of 0.6 rnA was in series between
the DIP lead and the negative lead of a megohmmeter at 500 V, while the positive
megohmmeter lead was connected to the grounding snap or conductive layer of the material
under test. The bulb became very dim at a little above 109 ohms through the DIP lead,
corresponding to 0.3 microampere.
PAPER NO. 14 211
The test did not detect COM-hazardous surfaces at the low end of the MOSFET failure
rate curve. However, the bulb did provide a visual demonstration of the worst COM-offenders
by glowing brightly and flashing on contact. With unhomogeneous, ungrounded surfaces
not amenable to a resistance measurement, a flash of the bulb would suggest a COM hazard.
In conclusion, COM-safe laminates or other COM-safe materials, such as soft mats, can
be identified by a resistance measurement. Measurements can be taken using the standard
NFPA 99 electrode or a small-area probe that simulates a device lead and detects conduc-
tive inhomogeneities.
References
1. McAteer, O. I., Electrostatic Discharge Control, McGraw-Hill, 1989, p. 173.
2. Kolyer, 1. M. and Watson, D. E., "The Charged Device Model and Work Surface Selection,"
Evaluation Engineering, October 1991, pp. 110-\17.
1990, p. 77.
4. Anderson, D. C., "A Simple Approach to ESD Damage Prevention," EMC Technology,
March/April 1991, p. 38.
PaperNo. 15
This paper is reprinted. by permission. from EMC Test and Design.

September 1993. pages 28-31.
REALISTIC TESTING OF ESD MATERIALS
John M. Kolyer

Defense Electronics
San Bernardino. California
Data-sheet material properties such as surface resistivity. shielding effectiveness. and tri-
boelectric charging propensity are measured under arbitrary. fixed conditions and cannot
predict performance under the varied. dynamic conditions of use. Thsts that realistically
simulate electrostatic discharge (ESD) damage mechanisms are needed for selecting ESD-
protective materials with confidence. This article briefly describes each of the common
ESD damage mechanisms~2,3 and gives examples of relevant tests.
Human-Body Model (HBM)

In the HBM. a charged person touches a lead or terminal of an ESD-sensitive device and
causes a damaging flow of electrons to or from the device. The current may cause damage
by its energy (current-sensitive devices. e.g.• thin-film resistors). or it may produce a differ-
ential voltage that exceeds the dielectric strength of a thin oxide layer [voltage-sensitive
devices. e.g.• MOSFETs (metal-oxide semiconductor field-effect transistors)).
The HBM mechanism is a major. if not tlu! major. cause of ESD damage. so the skin
voltage of personnel handling ESD-sensitive items must be strictly controlled by ground-
ing. Wrist straps are the usual means. and the skin-to-ground resistance should be limited to
a maximum of 10 Mel to prevent harmful surges of voltage when operators triboelectrica1ly
charge themselves by motions such as foot shuffling.l
Recently. a cordless wrist strap of novel design appeared on the market. The principle of
operation appears to be corona discharge from fibrous material in the strap rather than the
draining of charges to ground. A sample strap was tested in two ways.
First. a person wearing the cordless strap aimed a static (field) meter at a grounded steel
surface while stroking his rubber shoe soles on a nylon carpet. When he lifted one foot to
reduce his capacitance the meter read +2 kV and his skin voltage was -2 kV. The same
result occurred without the wrist strap. demonstrating that the strap was ineffective under
the conditions of the test.
In the second test. the strap-wearer sat in a chair on the same carpet and shuffled his feet
without lifting them. Then he touched the cap of a 3M Static Event Detector™ (SED) with
the case grounded. The SED is diagrammed in Fig. 1. (The shunt with the neon bulb pro-
tects the SED from voltages above 5 kV that could damage it.) A general discussion of the
272
PAPER NO. 15 273
NE-2
SED Bulb
Fig. 1. Protected SED.
use of the SED in ESD process control has been published. 4 An advantage of the SED is
ease of use, because a liquid crystal display (LCD) changes color when the monitor is
tripped, and no measurements or readings are needed.
The SED used in this test had a capacitance of 150 pF and was tripped by voltage surges
of 92 V or more, so it represented a MOSFET with a damage threshold of approximately
100 V. In repeated trials, the seated strap-wearer tripped the SED, demonstrating that
MOSFETs would have been damaged by being handled under analogous conditions (gate
lead touched, substrate-case lead grounded). The conclusion was that the HBM hazard was
not significantly reduced by the cordless wrist strap. A cord to ground, including a current-
limiting resistor to protect the wearer from shock, remains a necessity.
To reduce the chance of HBM damage, the skin voltage of people wearing ordinary
shoes can be lowered by special floor materials. This partial control of skin voltage is a
"safety net" to protect ESD-sensitive items if they are accidentally touched by ungrounded
personnel. Also, there is less chance of fingers throwing sparks on dry days and upsetting
equipment in computer rooms. A practical test for candidate flooring is measuring the skin
voltage on twelve randomly selected people wearing their ordinary shoes and walking nor-
mally. The walker holds a probe connected to an electrically isolated plate monitored by a
static meter. 2
A static-limiting floor finish with a microscopically thin surface layer of antistat reduces
the skin voltage of a person on a tile floor to approximately 10 percent of that found for
conventional acrylic finishes or "waxes.''2 An alternative to this special finish, which must
be carefully applied and maintained, is conductive carpet. Two types of such carpet were
evaluated with the results shown in Table 1. Most of the shoe soles, and all the heels, hap-
pened to be rubber rather than leather. The relative humidity was 40-42 percent at 75°F.
As seen in Table 1, charging varies greatly among people in this test, so data must be
averaged. Also, the brand of acrylic finish, as well as the condition of the tile, including soil-
ing, affects the results. Carpet 1, which contained a proportion of conductive fibers, reduced
people's skin voltage to 42 percent of the level for the tile with conventional floor finish, while
Carpet 2, in which all the fibers were conductive, reduced the voltage to 25 percent. These
carpets did not perform as well as static-limiting floor fmish, but they may be suitable for
replacing high-charging conventional carpet in computer rooms or office areas.
Machine Model (MM)

In the MM a damaging current flows between a metallic object, such as a tool, and an ESD-
sensitive device. The MM is simulated by tests in which the resistance between a charged
capacitor and a probe touched to a device (or sensor representing a device) is nearly zero,
274 APPENDIX
Table 1. Walk Test Data.

SKIN VOLTAGE, kV
TILE WITH REGULAR TILE WITH REGULAR

PERSON ACRYLIC FINISH (WAX) ACRYLIC FINISH (WAX)
NO. CARPET 1 (CONTROL FOR CARPET I) CARPET 2 (CONTROL FOR CARPET 2)
I +0.5 -2.2 ...{}.OO5 ...{}.02

2 +0.5 ...{}.8 +0.1 "'{}.4
3 +0.8 ...{}.7 +0.2 "'{}.5
4 +0.25 -2.5 ...{}.OO5 ...{}.15
5 +0.45 ...{}.9 +0.15 "'{}.3
6 +0.8 "'{}.5 +0.1 ...{}.9
7 +0.8 ...{}.8 +0.05 "'{}.5
8 +0.5 ...{}.7 ...{}.01 "'{}.2
9 +0.4 -1.7 +0.02 ...{}.15
10 0 ...{}.3 +0.4 -1.0
11 +0.05 ...{}.6 +0.1 "'{}.5
12 +0.01 ...{}.5 +0.1 "'{}.I5
Average +0.42 -1.0 +0.10 ...{}.40
whereas HBM tests involve a resistance of 150-1500 n to represent that of a person. Since
the MM is the more severe condition, it is appropriate for worst-case tests of packaging
materials.
A convenient method for testing bag materials is to place a protected SED (Fig. I) inside
the bag, which rests on a grounded aluminum plate, and then to touch a charged probe to the
bag above the cap of the SED. A probe with a diameter of 0.025 inch represents a sharp
hand-held tooLS More severe tests6 use a rounded probe with a diameter of 0.25 inch to
represent a blunt tool. With the 0.2S-inch probe at 10 kV, the equivalence of the SED to an
oxidized wafer7 failing at an average of 117 V was demonstrated as shown in Table 2, and
the wafer test (Fig. 2) in turn correlated with MOSFET testsY
As seen in Table 2, only packaging films with an opaque metallic layer passed the severe
lO-kV, blunt-tool test with no failures in five trials. Such materials are rated "excellent."
Materials failing the blunt-tool test at 10 kV but passing at 6 kV are rated "good"; these
include MIL-B-S170SC Type III (see-through film laminate with vapor-deposited alumi-
num protected by O.S-mil polyester film). Materials passing only the sharp-tool testS at 9 or
10 kV are rated "fair," and materials failing all three tests are "poor."
"Poor" materials include see-through film laminates with the metal "out" (protected only
by a thin organic coating rather than polyester film). The spark thrown by a finger on a dry
day will create a hole in the metallization and trip the SED inside metal-out bags, even
though these bags pass the EIA 541 (Electronic Industries Association) shielding test. The
problem with the EIAtest is that it uses large, flat electrodes at relatively low voltage (I kV)
and does not include real-life spark discharges. This misleading test gives a good rating to
volume-conductive, carbon-loaded bags which perform poorly in our tests as well as in
tests recently conducted in Germany using the EIA 541 network with a krypton switch.s A
realistic approach demands that if spark discharges are a threat, they must be accounted for
in the test method.
Another example of the MM is discharge between a charged, isolated conductor and
an ESD-sensitive device. Isolated conductors may attain considerable voltages by
PAPER NO. 15 275
Table 2. Data for Packaging Materials Tested at 10 kV.

FAILURE RATE (FAILEDfTESTED)
MATERIAL 1lllCKNESS, MILS OXIDIZED WAFER SED
Antistatic polyethylene 6 4/5 5/5

(MIL-B-81705C Type n)
Black, conductive 4 5/5 5/5
polyethylene
Foil laminate 9 0/5 0/5
(MIL-B-81705B Type I)
Foil laminate, antistatic 9 0/5 0/5
Tyvek outside, antistatic
polyethylene inside
Metal-in, see-through 3 4/5 4/5
aluminum
(MIL-B-81705C Type m)
Metal-out, see-through 3.5 5/5 5/5
copper
Aluminum foil 1.3 0/5 0/5
Metal-in, opaque aluminum 4 0/5 0/5
(MIL-B-81705C Type I)
MIL-B-81705C Type m 198 total 2/5 3/5
lined with antistatic foam
Corrugated cardboard 130 5/5 5/5
space-charging in the vicinity of air ionizers, so tests were run with a 20-pF plate placed at
various distances from ceiling-mounted. pulsed DC emitters.
A potential relative to ground of 250 V was required for the 20-pF plate to trip the SED
(Fig. 1) with case grounded. This result is consistent with experience7 for low-capacitance
charged objects. When the plate touched the cap of the SED. the capacitance rose so that the
original 250 V fell to approximatelv the 92-V threshold voltage for tripping the sensor.
Packaging
Material
Discharge
Insulating Pivot Arm
Oxide
Conductive
P-Silicon
Adhesive
Fig. 2. Wafer test apparatus.

276 APPENDIX
For 250 V on the plate, cycle time was plotted against emitter-to-plate distance as shown
in Fig. 3. The normal operating range for emitter voltage and working distance lay in the
safe area of the graph and the conclusion was that space charging of 20-pF isolated conduc-
tors would not threaten devices, such as MOSFETs, mimicked by the SED.
Charged Device Model (COM)

In the CDM, an ESD-sensitive device (discrete or mounted on a circuit board) becomes
charged. For example, a dual in-line package (DIP) may be triboelectrically charged by
sliding inside a DIP tube. Then a lead or terminal discharges to a conductive surface. The
question arises: How conductive is "conductive" to cause the CDM discharge? Attempts
have been made to find an answer. 9•1O
One method of testing surfaces for CDM safety uses the SED as shown in Fig. 4. The
double-pole switch is closed, the power supply is turned on to charge the apparatus to I kV,
the switch is opened, and then the probe is touched to the surface under test. Capacitances
C 1 and C2 are 5-10 pF to represent circuit lines connected to a board-mounted MOSFET as
represented by the SED.
I
6
o
5
Danger Zone
(250 V or
More on Plate)
at 22 kV on
Emitters
.I....
...
~
GI
3
2
(No Danger
Zone at 10-
12 kV on
Emitters)
/0 fIj Nonnal Operating
1 , / R~
1 2 3 4 5
Distance from Emitters to Plate, feet
Fig. 3. Pulsed air-ionizer data.

PAPER NO. 15 277
Probe
Fig. 4. CDM test schematic.
The RC time constant for draining C2 at 8 pF is 8 x 10-7 second (rapid discharge) for
= =
probe-to-surface R 1()5 ohms and 0.8 second (slow bleed-off) for R 1011 ohms. In prin-
ciple, this constant could be calculated. However. in practice, R and C usually cannot be
measured with confidence, so an empirical test is necessary.
In general, materials with a surface resistivity of at least 109 ohms/square are CDM-safe
= =
at I kV and C, C2 5-10 pF. But keep in mind that surface resistivity readings vary with
the manner of measurement, including electrode configuration.
Curves of surface resistivity versus relative humidity (RH) can be derived by confining
samples with saturated chemical solutions that control the relative humidity in the air above
them. II As seen in Fig. 5, the resistivity of many materials varies greatly with RH, so the
same material may be CDM-safe at low humidity but unsafe at high humidity. At approxi-
mately 50 percent RH, vulcanized fiber was well above 109 ohms/square and passed the
above-mentioned CDM test with the SED; this material was selected for tweezers used to
handle ESD-sensitive dice. Similarly, a recently developed antistatic nylon for probe tools
for assembling the same dice passed the CDM test at 50 percent RH. A permanently anti-
static vinyl plastic, suitable for tote boxes, also passed. In contrast, a carbon-impregnated
cardboard packaging material had a surface resistivity less than lOS ohms/square at 50 per-
cent RH and failed the CDM test.
Field-Induced Model (FIM)

In the FIM, an electrical field (E-field) induces a charge on an ESD-sensitive device to
cause failure. The SED is unsuited for FIM simulations because it is tripped by voltage
surges rather than by overvoltage gradually accumulated as a device is moved into an
E-field. The oxidized wafer or actual MOSFETs must be used.
In the case of the FIM, or any ESD damage mechanism, actual devices or an SED may be
needed to sense ESD events because the risetime, e.g., 100 picoseconds, may be too brief
for any oscilloscope to measure. '2
278 APPENDIX
1013 .........~. . . ._ ..........~..........
v • Antistatic Vinyl
F • Vulcanized Fiber
N • Antistatic Nylon
C • Carbon-Illpregnated Cardboard
105 . .~__. .____. .~. .~__~~

o 10 20 30 40 50 60 70 80 90 100
RH,%
Fig. 5. Surface resistivity vs. relative humidity.
At assembly benches or shipping/receiving stations, the source of a damaging E-field

can be a nonconductive tool or container. Therefore, nonconductors should not be used in
handling ESD-sensitive items. Materials for this purpose must be sufficiently conductive
for charges to drain to ground, via a grounded operator's hand or a static-dissipative work
surface, within a few seconds. As a rule of thumb, the surface resistivity should be below
10 '2 ohms/square for a material to be antistatic and not retain charges, and curves like those
in Fig. 5 show the cut-off RH at which 10 '2 ohms/square is reached. Vulcanized fiber re-
mains antistatic down to approximately 3 percent RH and is judged to be ESD-safe under
normal working conditions. The other three materials in Fig. 5 remain antistatic even at 0
percentRH.
Also relating to the FIM is the propensity of a work surface or packaging material to
charge nonconductors rubbed against it. For example, a circuit board might be triboelectrically
charged by sliding against packaging material, and then the E-field on the board could
damage ESD-sensitive devices being installed or already installed.
Triboelectric charging propensity relates to the CDM as well as the FIM. For example,
minimal charging of the plastic cases of DIPs sliding out of a DIP tube is desired to reduce
the CDM hazard.
PAPER NO. 15 279
Table 3. Trlboelectrlc Charging Data.

VOLTAGE ON COUPON, kV'
FR-4 GF POLYIMIDE OVERALL

MATERIAL READINGS/AVG. READINGS/AVG. READINGS/AVG. AVG. RATING
Antistatic +1 +1 +3/+2 +1 +1 +2 /+1 0-1 0/-0.3 Medium Charging

vinyl
Cardboard +1 +1 +1/ +1 +0.4+2 +3 /+2 -1-1-1/-1 I Medium Charging
Conventional +4 +4 +4/ +4 +4 +2 +2 / +3 +2 +3 +1/ +2 3 Medium Charging
vinyl
Black conduc- +7 + I +3 / +4 +18 +11 +18/+16 +2+1 +1/+1 7 High Charging
tive plastic
Antistatic 0 0 0/ 0 o 0 0 / 0 0 0 0/ 0 o Low Charging
polyethylene
• Results for three tests and the average.
Experience shows that standard triboelectric charging tests with quartz and Teflon™ are
insufficient, and specific materials of concern must be evaluated. For example, a perma-
nently antistatic vinyl plastic, suitable for tote boxes. was tested (Table 3) using circuit-
board laminates (FR-4, GF, and poly imide) with the FlM in mind. Triboelectric charging is
notoriously erratic and unreproducible, so these tests were run in triplicate. Because of the
scatter of the data, the best that can be done in such testing is to make the rough classifica-
tions shown. Black conductive plastic is well known to produce high charges, whereas
antistatic polyethylene (MIL-B-SI705C, Type II) imparts very low charge because antistat
invisibly rubs off onto the surface of the other material. The antistatic vinyl, which had a
dry surface with no antistat layer, was a medium charger -like cardboard or conventional
vinyl.
Summary
Novel tests are appropriate when standard tests are insufficiently specific and realistic. The
object is to simulate worst-case ESD-hazard scenarios so that materials for tools, contain-
ers, work surfaces. and flooring can be selected with maximum confidence.
References
1. O. I. McAteer, Electrostatic Discharge Control, McGraw-Hili, 1989.
2. I. M. Kolyer and D. E. Watson. ESD from A to Z: Electrostatic Discharge Control for Electronics,
Van Nostrand Reinhold, 1990.
3. I. M. Kolyer. "Fundamentals of ESD Control," Technical Record of Expo '91, International
Conference on Electromagnetic Compatibility, EMC Technology Magazine, Reston, VA May
18-20,1992, pages 154-161.
4. I. D. Campbell, "ESD Process Control and Measurements," EMC Test and Design,
September/October 1992, page 63.
5. I. M. Kolyer and D. E. Watson, "Packaging for High-Voltage Discharge Protection," Evaluation
Engineering, March 1992, pages 96-100.
280 APPENDIX
6. I. M. Kolyer, "Toward an Ideal ESO-Protective Package", Proceedings of the 1994 EMC/ESO

International Conference, Anaheim, CA, April 12-19, 1994.
7. I. M. Kolyer and O. E. Watson, "ESO Testing of Silicon Wafers," Evaluation Engineering,
September 1991, pages 115-119.
8. I. B. Brinton, "Test Questions Non-Metallized Bags." EOSIESD Technology, Ianuary/February
1992. pages 12-13.
9. I. M. Kolyer and O. E. Watson, "COM and Work Surface Selection," Evaluation Engineering,
October 1991, pages 110-117.
10. I. M. Kolyer., ''Testing Surfaces for ESO Safety," Evaluation Engineering, November 1994,
pages S-36-5-40.
11. 1. M. Kolyer and R. Rushworth, "Humidity and Temperature Effects on Surface Resistivity,"
Evaluation Engineering, October 1999, pages 106-110.
12. The Pulsed EMf Handbook, Second Edition, KeyTek Instrument Corp., Wilmington, MA, 1991,
page 23.
Paper No. 16
This paper is reprinted, by pennission, from Evauluation Engineering,

November 1994, pages S-36-S-40.
TESTING SURFACES FOR ESD SAFETY
John M. Kolyer

Defense Electronics
Whether we want to admit it or not, we are our own worst enemies when it comes to elec-
tronic equipment. Human beings are the major source of direct injection (01), the most
straightforward and common cause of ESD damage.
Keeping ESD in check requires a good control program and even better testing proce-
dures. Today, several practical test methods simulate 01, the ESD damage mechanism that
includes the human body model and machine model, the charged device model (CDM), and
the field-induced model (FlM).'·2
The culprit in all these models is a conductive surface that participates in rapid dis-
charges to or from an ESD-sensitive (ESDS) device or circuitry leading to the device. The
conductive approach to ESD control, in which ESDS items are handled or approached only
with highly conductive grounded objects, encourages rapid discharges and must be less
safe than the rival static dissipative-approach. 2 The desirability of slow discharge is re-
flected in MIL-HDBK-773, which cautions against damage by spark discharges. 3
Where do you draw the line on conductivity? Resistance measurements vary with factors
such as electrode design, applied voltage, and relative humidity (RH), and results are diffi-
cult to relate to discharge rates and ESD damage. The carbon-loaded film in Table 1 is a
well-known sparker and, presumably, is too conductive. The laminate has a resistance to
ground in the 107 to 108 n range, previously judged to be CDM-safe for hard laminates. 4
But what about materials such as skin or a rubber mat?
Tests simulating damage models are necessary. A DI or CDM test has been proposed in
which a neon bulb connected to a 500-V power supply (PS) flashes when contacting con-
ductive surfaces.' This method is subjective and does not necessarily correlate with device
damage. 6
A CDM test uses actual devices (MOSFETs) for credibility, and this test now has been
expanded to represent DI and the FlM.6
The expense of inventorying MOSFETs has been avoided by replacing them with a simu-
lated voltage-sensitive device, the Static Event Detector™ (SED) from 3M.
For our testing purposes, a MOSFET (2N4351) stressed between the gate lead (G) and
the substrate-case lead (SC) is essentially a silicon-oxide layer, with a dielectric strength of
approximately 100 V, bypassed by a high-resistance leakage path. This path protects
MOSFETs during CDM discharges to high-resistance surfaces. 4
281
282 APPENDIX
Table 1. Materlala.
RESISTANCE (OHMS)
TOOROUND BBTWEEN
SURFACE 11IROUOH NFPA99
RESISnvrrY, 11IROUOHD NFPA99 EU!C'I1tODES
OHMS/SQUARB, (FlO. I) EU!C'I1tODB I IN. APART
NO. DESCRIPI10N AT 100 V ATIOV AT IOV AT 100 V
Antistatic 1 x 1()9 2 x 1()9 9 x 10' 8xlO'

polyethylene (5 x l()9at (8 x 1010 (1 X 1010 (4 x 1()9
(MIL-B-81705, 35%RH) at 35%RH) at 35% RH) at 35% RH)
Type ll), 6 mils
2 Carbon-loaded 3 x 10' 1 x 10' 3 x 1()4 3xl()4
polyethylene film, @IV @IV @IV @IV
4 mils
3 Conventional 3xl()9 1 x 1010 2xl()9 lxl()9
corrugated
cardboard,
110 mils
4 Hard benchtop 7 x 1()9 15 x 1010 4x 107 3 X 107
laminate with
buried conductive
layer
5 Rubber workbench 3 x 107 3xlO' 1 x 1()6 2xl()6
mat with buried
conductive layer
6 Bare finger 4x 10' 3xl()6 N/A N/A
at IOV
7 Carbon-loaded 1 x 10' 8 x 1()6 N/A N/A
plastic cot on at 10V
finger
8 Static-dissipative 4 x 10' 3 x 1()9* N/A N/A
plastic cot on at 10V
finger
• Measured through steel plate on wrist.
To cause damage, the rise time of voltage need not be rapid, because the oxide must
rupture whenever the voltage differential across it exceeds approximately 100 V. MOSFETs
subjected to the very slow rise time of 10 V/s failed at an average of 117 V-only slightly
above the 80 V to 100 V found at very fast rise times. A MOSFET with SC grounded will
fail when an antenna attached to G is moved slowly into an electrical field. 2
In the same situation, the SED is tripped only by a very intense field. creating a differen-
tial voltage far above 100 V, because normal sensor response requires voltage surges. When
rise times are fast, as they are in most ESD events, the MOSFET and the SED give equiva-
lent results.
An SED is visibly tripped at the moment of the ESD event. On the other hand, MOSFETs
can give false indications of failure by being damaged at other points in the test process.
PAPER NO. 16 283
The SED used for these tests had a capacitance of 150 pF and was tripped at 92 V. SED
and MOSFET failure rates were identical in a CDM test, and similar agreement has been
found in packaging tests with spark discharges (Table 2).
As shown in Fig. 1, the sensor (SED, MOSFET, thin-film resistor, or neon bulb) was held
between alligator clips AI and A2• The SED cap was connected by a fine wire (26-gauge) for
flexibility. Terminal Ts was connected to one of two probes: a dual in-line package lead (the
D probe) or an edge contact (EC) comprising a 0.25-in.-wide copper circuit line on glass-
filled epoxy board material mounted on a Teflon™ block (the EC probe).4
Table 2. Test Data.

c,
SIMULATED CONDmON MATERIAL 4 pF 204 pF 1028 pF
Charged DIP falls to I at 35% RH, 3, 4 0/10 0/10 0/10

grounded surface (MOSFETs: 0/5)
(Probe D) (CDM) 8 0/10 3/10 10/10
(See Reference 4) 2/10 9/10 10/10
5 10/10 10/10 10/10
(MOSFETs: 5/5)
2,6,7 10/10 10/10 10/10
Fingers grasp edge I at 35% RH 0/10 0/10 0/10
contact of board (wrapped on finger)
(Probe EC) (01, FIM, 8 4/10 9/10 10/10
CDM) 6, 7 10/10 10/10 10/10
Bag or box falls I in. I (6 x 10 in. bag) 0/10 1/10 2/10
onto edge contact of
board
(Probe EC) (CDM, 01) 3 (2 x 5 x 10 in. 0/10 2/10 2/10
box)
2 (6 x 10 in. bag) 10/10 10/10 10/10
Charged DIP falls to 2 (1.6 in! 1.4 pF) 0/10 2/10 2/10
isolated conductor on 0.25 in. Teflon
(Probe D) (CDM) 2 (6.3 in,2 5.6 pF) 10/10 10/10 10/10
on 0.25-in. Teflon
Charged circuit board
Material C, = 9414 pF
with lOO-ohm
(O.1K) or 300 kilohm Steel 0.1 K, ~.055%; 300K, ~.039%
(300K) 1/20-watt thin-
film resistor touched by 2 0.1 K,~.OOI%; 300K,tO.001%
pointed, grounded tool
(Probe D) (CDM) 5 0.1 K, to.005%; 300K, ~.OOI %
Material C, =4pF C, = 839pF

N-2 neon bulb as sensor
(Probe D) (CDM) 2 Slight flash in Aash
dim light
5 No flash in Slight flash in
dim light dim light
Note: Failure rates for the SED are expressed as fraction of failed/tested. Resistance shifts (highest of three tests) are
in percent.
284 APPENDtX
MOSFn
I: SED ~"
-= =-_-
.... -
~"'"- _~-0
-~
•
Fig. 1. Test apparatus.
Tenninals T T1, T, and T. were claw clips. Tl and Tl were grounded. With AI and Al
" C measured between Tl and a metal plate beneath the apparatus was 4
empty, capacilaJ'lCe I
pF. Similarly, capacitance Cl measured between T. and the plate was 8 pF with !he 0 probe
installed and to pF with the EC probe installed.
Capacitances CI and C2 represent those of circuit lines on a board. A similar measure-
ment between the plate and Ees on aerospace computer boards of various designs gave
from 3 pF to 31 pF. To raiseC a capacilorsomelimes was held byT, and Tl , if desired, C1
"
could also be l'3ised by a capacitor held by Tl and T•.
Testing was done at 1000 V. The SED sometimes was tripped by the surge when the PS
was turned on with switch Sclosed. The MOSFETs were always damaged, so the voltage
was increased slowly; for example. at 100 VIs.
After 10 s at 1000 V, S was opened and block B was pulled aside to let Dfall to the
surface. The SED was tripped by a I'3pid discharge through D. For the carbon-loaded film,
with 10' n resistance through D (Table t), the 8-pF capacitorC 2 dl'3ined with an RC time
constant of 8 x 10..1 s, the time constant for the laminate (.5 x 1010 n through D) was 0.4 s.
These times represent a I'3pid discharge vs a relatively slow bleed-off.
In theory, RC might be estimated in all cases. But, in pl'3ctice, R and C may be difficult
or almost impossible to measure, so an empirical test is more pl'3ctical and convincing than
a calculation.
Table 2 gives examples of simulated real-life handling conditions for ESDS items The
RH was approximately 50% except for tests with antistatic polyethylene at 35% RH. With
PAPER NO. 16 285
the caveat that generalization beyond the conditions of the tests is unjustified, conclusions
drawn from Table 2 are:
o The rule of thumb that a resistance to ground of 101 to lOS Q is desired for COM safety
of hard benchtop laminates is supported by the data. 4 This rule does not extend to other
materials, such as antistatic polyethylene.
o The most meaningful resistance measurement was through O. In the first test in Table
2, 1010 Q to ground gave no failures; 109 Q gave several failures, especially at high C I ;
and lOS to lOS Q gave all failures.
o Elevated RH reduces the general ESO problem but makes certain materials become
discharge hazards. Antistatic polyethylene was safe at 35% RH but not at 50% RH.
o ESO-protective finger cots should be in the antistatic range (109 to 1011 Q/sq) rather
than static-dissipative (lOS to lOS Q/sq).

o Even cardboard, with its high resistance through 0, gave failures at high C I when a box
fell onto an EC. This result emphasizes that discharge safety depends on the mode of
contact.
o Isolated (ungrounded) conductors such as the carbon-loaded film are COM threats
even at low capacitance. The converse situation (conductor at 1000 V, device uncharged)
gave similar results in past work. A I pF conductor was harmless, a 3 pF conductor was
damaging.1
o Even with a very conductive surface and an unrealistically high C I , the resistance of
thin-film resistors shifted only slightly in a COM simulation. 4

o An N-2 neon bulb as a sensor gave a subjective result rather than a clear pass/fail.
Incidentally, conductive surfaces can be hazardous to personnel handling high voltage.

The carbon-loaded polyethylene film could carry a calculated current of 33 rnA at 1000 V
between NFPA 99 electrodes I inch apart. An AC current of 21 rnA to 40 rnA causes mus-
cular inhibition. 8
In conclusion, ESO-protective materials high in the antistatic range (1010 to 1011 Q/sq)
may be necessary for worst-case 01, COM, or FIM hazards. However, a rigorous antistatic
approach requires finger cots and restricts the normal use of conductive tools. Tests, such as
those in Table 2, are recommended to indicate real needs and avoid unnecessary and im-
practical precautions.
References
1. McAteer, 0.1., Electrostatic Discharge Control, McGrawHilI, 1989, pp. 173-188
2. Kolyer,l. M., and Watson, O. E. ESD from A to Z: Electrostatic Discharge Controlfor Electronics,
Van Nostrand Reinhold. 1990, pp. 5, 6, 9,157,160,169, and 205.
3. MIL-HDBK-773, April 1988.4.3.1.
4. Kolyer,1. M., and Watson, O. E., "COM and Work Surface Selection," Evaluation Engineering,
October 1991, p. 110.
5. "New Test Proposed for Checking ESO Safety of Materials," Compliance Engineering, Fall
1990, p. 77.
6. Kolyer,l. M., "Is Your Work Surface COM Safe?" EOSIESD Technologv, February/March 1992,
p.27.
7. Kolyer,l. M., and Watson. O. E., "ESO Testing of Silicon Wafers," Evaluation Engineering,
September 1991, p. \15.
8. OoO-HOBK-263, May 1980,7.3.1.3.
PaperNo. 17
This paper is reprinted, by pennission, from Evaluation Engineering,

October 1990, pages 106-110.
HUMIDITY AND TEMPERATURE EFFECfS ON SURFACE RESISTIVITY
John Kolyer and Ronald Rushworth

Defense Electronics
This investigation evaluated antistatic materials used in ESD control for electronics. The
objectives were to define curves for surface resistivity (a measure of the antistatic property)
vs relative humidity, detennine the temperature effect, and set an effective lower limit for
controlled relative humidity in assembly areas.
Twelve materials were studied: four commercial antistatic films, including a polyethyl-
ene film with a radiation-cured coating, three cellulosic (cellophane) films, two kinds of
paper (present in static-controlled areas in the fonn of documents such as work instruc-
tions), a detergent (representing topical antistats), a static-limiting floor finish, and leather
(representing leather shoe soles).
With reference to Fig. A, antistatic polyethylene I met MIL-B-81705, Type II, and con-
tained an amide antistat; antistatic polyethylene 2 contained an ethoxylated tertiary amine
antistat; the antistatic nylon also contained an ethoxylated tertiary amine antistat; and the
paper was 20-lb bond made by the bleached kraft process. Also, the filter paper was Whatman
No.1 made from cotton linters; the detergent was Joy (Proctor & Gamble); the floor finish
was a high-solids acrylic; and the leather was split, vegetable-tanned, and unfinished. To
facilitate resistance measurements, both sides of a nonconductive film (cellulose acetate)
were dipped in the detergent and floor finish.
Strips of film or sheet were held between clamps to make a square, as in ASTM D 257,
and sealed inside jars with saturated reagent-grade salt solutions (MgCI 2, NaN03, KCI,
KN0 3, K2SO.) to control the relative humidity (see Table 1).1-3. A drying agent (anhydrous,
calcium sulfate with indicating blue color) afforded 0% RH.
The jars were stabilized at three fixed temperatures, 41 to 46°F, 75 to 80°F, and 108°F,
until the samples reached equilibrium moisture content and the resistance readings became
constant. Although no acidic or basic vapors that might affect resistance were found, confir-
matory detenninations were made at 75 to 80°F with an alternate set of salt solutions (KF,
K2C0 3, NaHSO., NaN0 3, NH4HlO.).
The apparatus is shown in Figs. 1-6. Figure 1 is an individual jar. Figure 2 shows a
leather sample in the clamps. The jars were placed in a refrigerator (Fig. 3) or a low-heat
oven (Fig. 4). Figure 5 shows the interior of the oven with a light bulb (bottom of picture) as
the thennostatted heater, then the fan, then the samples. In Fig. 6, a jar is tested under
ambient conditions.
286
PAPER NO. 17 287
Code: o 75· 80 ° F, alternate solutions

- - - - - 41· 46°F -0- - - -0- 108 0 F
10' ...................................... 10' .........................._ ...._
AntJetatlo PoIyeIIIyIene 1 Antlltetlo PoIyethy..... 2
10' ..................................... 10' ..........................................

o 10 20 30 40 50 eo 70 eo 90 100 o 10 20 30 40 50 eo 70 eo 90 100
.... % .... %
10' ................................. 10' .....................................
Polyethylene with
RId.- Curld COetIng
10·' ..............." -....~............~

o 10 20 30 40 50 eo 70 eo 90 100 10·' .................." - - - -..........
o 10 20 30 40 50 eo 70 eo 90 100
.... % .... %
Fig. A. Plots of resistance data

288 APPENDIX
Code: -,--...... 75 - 80° F ° 75 - 80° F, alternate solutions
- - - - - 41· 46°F -0- ---0- 1080F
10' ...~~~-!-~~~~~~ 10· .................- .........~......

o 10 211 30 040 50 eo 70 eo 110 100 o 10 211 30 040 50 eo 70
AH,% AH,%
10' .............................- .....
10' .................- .........-!o~~~ 10' ............................................

o 10 211 30 040 50 eo 70 eo 110 100 o 10 211 30 040 50 eo 70 eo 110 100
AH,% AH,%
Fig. A. continued.
PAPER NO. 17 289
Code: - - _ 75-80 o F 75 - 80 ° F. alternate solutions

-a-_
o
41 - 460F -0----0- 108 °F

-1"....""'1'.......1"""1
107 . . . . . . . . . . . . . . . .
Detergent
10·' ...&..&..&............................... 10' ...~...............~...............

o 10 20 30 40 110 eo 70 eo 80 100 o 10 20 30 40 110 eo 70 eo 80 100
""'' ' ""'' '
10·' ...~~........~~~~~1;.I 10' ...&..&..&.............................

o 10 20 30 40 110 eo 70 eo 80 100 o 10 20 30 40 110 eo 70 eo 80 100
""'' ' ""'' '

Fig. A. continued.
290 APPENDIX
Table 1. Relative Humidity over Saturated

Aqueous Solutions.
TEMPERA11JRE RELATIVE
SOLUTE of HUMIDITY,'"
MgCI 2·6Hp 43 35
75 33
108 32
NaN0J 43 78
75 75
108 70
KCI 43 88
75 85
108 82
KNOJ 43 95
75 94
108 89
K2S04 43 98
75 97
108 96
KF·2Hp 75-80 27
K 2COJ·2H2O 75-80 43
NaHSO.. HP 75-80 52
NH4~P04 75-80 93
As seen in Fig. A, agreement of results with the standard vs alternate solutions generally
was good. Incidentally, sulfuric acid solutions, sometimes used to control relative humidity,
did evolve acidic vapor and were rejected for that reason.
The resistance across the films was measured at 100 V with a Beckman L-12 megohmmeter
with the cables shielded to prevent RF interference. Time of electrification was 5 s. Runs
were made in triplicate with the averages plotted in Fig. A.
Standard deviations generally were 5 to 60% of the mean. No data were rejected. The
clamps contacted both sides of the specimens, so R in Fig. A is the same as surface resistiv-
ity (in M Q/sq) only for the volume-conductive materials (paper, leather, and cellophane
films). For the other materials, the surface resistivity is twice the value of R in Fig. A.
The data provided several conclusions:
• Surface resistivity is controlled by two factors. First is the moisture content which, in
tum, is controlled by the relative humidity; second is the temperature. A similar tempera-
ture effect has been reported for charge decay time,4 which would be expected to corre-
late with surface resistivity for materials such as antistatic polyethylene. At constant
relative humidity, the resistance of the antistatic materials decreased exponentially with
temperature, as occurs with dielectrics' and semiconductor.6 Metals show the opposite
effect.
• The reality of the temperature effect is proved by the dry-air data. At high relative
humidity, the temperature effect becomes small or nil as the materials move from the
semiconductive toward the conductive state.
• Because of the significant thermal effect, the temperature should be given, along with
the relative humidity, in reporting surface resistivity data.
PAPER NO. 17 291
• Surface resistivity commonly is assumed to vary exponentially with relative humid-

ity, S but the plots of log resistance vs relative humidity (Fig. A) often were curves depart-
ing far from linearity. The cellulosic materials (paper and cellophane) and leather constitute
one family of curves; all are fairly close to straight lines of approximately the same
slope.
3 4
5 6
Figs. 1-6. Resistance apparatus.

292 APPENDIX
The antistatic polyethylenes and polyethylene with radiation-cured coating constitute

another family of fairly similar curves, while the curves of antistatic nylon are somewhat
similar but shifted much upward. The detergent and static-limiting floor finish form an-
other family, with curves rapidly rising, especially for the floor finish, as 0% RH is
approached. Therefore, the floor finish is very sensitive to slight changes in relative
humidiy at the low end.
• Antistatic polyethylene is sometimes assumed not to function in dry air, but the two
brands tested were effective. The surface resistiviy remained less than 1012 n,tsq so that
charges could drain quickly from the surface.
• The graphs provide cutoff relative humidities below which materials lose their anti-
static propery (that is, the surface resistiviy exceeds 1012 n,tsq). At room temperature, the
cutoffs are 0% RH (no cutoff) for polyethylene with either of two extruded-in antistats or
an antistat-impregnated, radiation-cured coating; 0% (no cutoff) for a detergent "sweat
layer"; 3 to 4% for a plasticized cellophane film with or without a coating; 9% for a
static-limiting floor finish; 18% for leather or an unplasticized cellophane film; 23 to
26% for paper; and 50% for antistatic nylon .
• These data support the requirement for a minimum controlled relative humidity of
30% to keep most antistatic materials functional. This agrees with experience. So-called
antistatic nylon is an exception, but this film is really not antistatic in practical terms and
has been the subject of a GIDEP Alert7 for that reason.
The plasticized, coated cellulosic film Ecostat,8 contains moisture and a hygroscopic
softener or plasticizer. No antis tat is concentrated at the surface where it can be removed by
rubbing.
After vigorous rubbing with this film, nonconductive surfaces remained nonconductive.
However, they were made antistatic by being rubbed with, or merely pressed against, the
antistatic polyethylenes or the polyethylene with the radiation-cured coating.
In contrast to other materials, Ecostat remained antistatic after brief rinsing with water.
Soaking at 81°F for 24 h extracted the plasticizer so the film reverted to unplasticized cello-
phane (Fig. 7) and the room-temperature cutoff relative humidity rose from 4 to 18%.
In preliminary static decay tests (FED-STD-101, Method 4046) at 8 to 10% RH and
75°F, Ecostat showed a decay time of 5 to 11 s. It was noncorrosive to copper or Sn62 solder
when exposed in contact with these metals at 100% RH and 75 to 80°F for one year or
120°F for one month.
This film appears to be suitable for a noncontaminating intimate wrap; e.g., inside MIL-
B-81705, Type I, to meet the intent ofMIL-HDBK-773. It also is biodegradable.
PAPER NO. 17 293
CI Cellophane, pla.t•
• Cellophane, plaat. • coatad

o Cellophane, pla.t.. aner extraction
e Cellophane, plast•• coated, after extraction
1~pp~-r~--. .~~--p-~-r. .
10'
•
~
10'
10' ....~~~. .~. .

o 10 20 30 «)
--6.--.....
50 60 70 80 90 100
RH,%
Fig. 7. Effect of water extraction on cellophane films.
References
1. International Critical Tables, Vol. I, p. 67.
2. Stokes and Robinson, Industrial and Engineering Chemistry, Vol. 41,1949, p. 2013.
3. Handbook o/Chemistry, edited by N. A. Lange, McGraw-Hill, Revised 10th Edition, 1967, p.
1432.
4. Evaluation Engineering, April 1982, p. 74.
5. ASTM D 257-78, Appendix XI.
6. Encyclopaedia Britannica, 15th Edition, Vol. 18, 1989, p. 241.
7. GlDEP Alert C6-A-87-09, November 12, 1987.
8. Evaluation Engineering, March 1990, p. 96.
PaperNo. 18
This paper is reprinted. by permission. from Evaluation Engineering.

March 1992. pages 96-100.
PACKAGING FOR HIGH-VOLTAGE DISCHARGE PROTECTION
John M. Kolver and Donald E. Watson

Defense Electronics
San Bernardino. California
The EIA-541 electrostatic shielding test for packaging materials uses the relatively low
voltage of 1.000 V on a large flat electrode. This may simulate fields or low-voltage dis-
charges. but possible damage from ESD includes high-voltage discharges--often as visible
sparks-from people or conductive objects. This article describes a test method for simulat-
ing such discharges and outlines packaging strategies to resist them.
The test is conducted at -9 kV. the minus sign meaning that electrons flow from a charged
capacitor toward the package. This is not a worst-case human voltage; the machine-model
resistance of nearly zero is worst-case. and the ESD sensitivity of 100 V is nearly worst-
case. The test conditions constitute a reasonable requirement. passed by Types I and III of
MIL-B-SI705.
Figure I represents an idealized. nearly zero-resistance person with 200-pF capacitance
discharging from a sharp-tipped. hand-held tool after being triboelectrified to -9 kV (for
example. by walking on a carpet at 20% RH). The material under test was pulled tight over
the capacitive probe. Then the tip of the O.025"-diameter capacitor lead was touched to the
material for 1 s. The vertical shield prevented the E field of the capacitor lead from inducing
false voltages in the circuitry.
The sensor was an oxide-coated silicon wafer modeling a MOSFET gate. I A pulse of 100
V or greater across the disks of the capacitive probe punctured the oxide under a steel probe
pressed against the surface. The resistance of the oxide measured through the steel probe
n n
was over 10 12 initially and usually less than 109 when the oxide had been damaged. so
the failure criterion was the upper limit of detection of 2 x 109 n bv a digital multimeter.
The tests were run at approximately 50% RH.
Wafer-oxide failure rates for the 14 commercial materials in Table 1 are consistent with
MOSFET data in past work. 2 Results of the tests indicate that a conductive layer was re-
quired for passing the test (failure rate = 0/5). Foil bags (Materials 4 and 5) are prescribed in
MIL-HDBK-773 and have been recommended for highly ESD-sensitive items and worst-
case discharges. 2•3
It was also concluded that semitransparent vapor-deposited metallization is effective only
when buried under a layer of plastic (Material 6). The screen bag (Material 8) is very
discharge-protective2 and relatively transparent. but expensive.
The S-mil skin of Material 14 essentially is plain paper on the outside but carbon-
impregnated on the inside. This resembles the successful shielding combination of Material 2
294
PAPER NO. 18 295
Capachor (200 pF)
C • Capachive Probe whh Brass Discs
Fig. 1. Apparatus.
Table 1. Commercial Materials.

FAILURE RAlE
NO. MATERIAL CONDUcnVE LAYER (FAlLURES/l'ESTS)
Antistatic polyethylene (pink poly). None 4/5

6 mils (Mll.-B-81705.TYpe U)
2 Cellophane. 1.4 mils None 4/5
3 Black conductive. 4 mils Carbon-loaded polyethylene 5/5
4 Foil laminate. antistatic TYvek outside. Aluminum foil. 0.25 mil 0/5
antistatic polyethylene inside
5 Foil laminate. aluminized Aluminum foil. 0.25 mil 0/5
outside. carbon-loaded polyolefin
inside (Mll.-B-81705.TYpe I)
6 Metallized with 0.5-mil polyester cover Aluminum. partially 0/5
layer. antistatic polyethylene inside transparent
(Mll.-B-81705.TYpe m. Class 1)
7 Metallized with 0.16" paperboard Aluminum. 150 A 0/5
cover, biopolymer inside
8 Screen sandwiched between Metallic window screen 0/5
Mll.-B-81705, TYpeU
9 Externally metallized Aluminum, partially 5/5
(thin organic coating) transparent
10 Externally metallized Copper, partially 4/5
11 Externally metallized Nickel, partially 5/5
12 Antistatic foam, 0.25" None 4/5
13 Corrugated cardboard, 0.09" None 4/5
14 Corrugated cardboard, 0.16" Carbon impregnated on 0/5
inside of skins
15 Air space None 0/5 (3")
4/5 (1S')
296 APPENDIX
Table 2. Material Combinations.

FAILURE RATE
OUTER MATERIAL INNER MATERIAL (FAILURESfJ'ESTS)
1 0/5
12 1/5
2 12 5/5
lor2 3,9 or 10 0/5
1 11 2/5
2 11 0/5
3 or 11 5/5
9 3/5
10 4/5
12 1/5
12 2,3,9, 10, 11 or 12 0/5
3,9,10 or 11 12 0/5
3 or 10 2 5/5
9 or 11 2 4/5
lor 13 15 (0.5") 0/5
outside and Material 3 inside in Table 2. The 0.14" gap (largely air) inside Material 14 adds
to the protective effect.
The conductive layer should be buried (Materials 4, 6, 7, 8 and 14) rather than exposed
(Materials 3, 5, 9, 10 and 11). This not only improves shielding but also discourages ESD
damage by the charged device model.4
Sparks did not correlate with ESD failure. Material 3 failed with a visible arc, but Mate-
rial5 passed; Materials 1,2,12 and 13 failed without an arc. The arc to Materials 9,10 and
11 may have contributed to failure by creating bare spots approximately 0.01" to 0.02"
diameter in the metallization. (Fig.2).
Choices are limited for single materials pressed against the ESD-sensitive item, but op-
tions multiply for combinations including double bagging and air gaps (Table 2). For ex-
ample, a pink poly bag (Material 1) or an externally, metallized bag (Materials 9 or 10) is
ineffective alone, but when combined correctly with pink poly on the outside-not the
inside-the metallization is buried and protected from spark damage.
Cellophane (Material 2) also is an effective cover layer. Materials 3, 9, 10 or 11 succeed
when combined with foam (Material 12).
Table 3, derived from Table 2, provides examples of six construction types with the
principles of their success: Faraday-cage shielding, a dielectric barrier of relatively heavy
plastic, and protective spacing. Even inexpensive, unmetallized materials are adequate with
air gaps, which need be only OS'.
Commercially recyclable and biodegradable containers use construction types 5 and 6
Other commercial packaging utilizes construction type 4, in which a transparent bag is
inflated with the least expensive material-air.
Common packaging materials in the right configurations can prevent the damage of high-
voltage discharge. The techniques of buried metallization and protective spacing offer many
possibilities for cost-effective package design.
PAPER NO. 18 297
Fig. 2. Bare spot 0.017 inch in diameter created on surface of Material 9 by spark discharge.
Table 3. Package Constructions.

NO. CONSTRUCTION MATERIAL NOS. (TABLE I) PRINCIPLE
Single bags or thin-walled 4,5,6,7,8 Faraday-cage effect

boxes in which ESD-sensitive (buried metal-
leads, terminals or contacts lization dissipates
press the inside surface discharge)
2 Double-bagging systems lor 2 over 3, 9, or 10; 2 over 11; Faraday-cage effect
with conductive layer 12 over 3, 9, 10, or 11
3 Double-bagging systems Two layers of I, totaling 12 mils; Dielectric barrier
without conductive layer two layers of 12, totaling OS' blocks discharge
4 Soft packages with air gap 1 or 2 with gap (OS' minimum) Space weakens
provided by antistatic bubble-wrap; E field
transparent film bag inflated with air
5 Rigid packages with 14 alone, or 13 with metallic layer, Faraday-cage
conductive layer e.g., aluminum foil, bured under effect
a film such as Material 2
6 Rigid packages with air gap 13 with a protective gap (OS' Space weakens
minimum) provided by antistatic E field
bubble-wrap or spacers and parti-
tions; may have double-pane view-
ing window of I or 2 with OS' gap
298 APPENDIX
References
I. Kolyer, J. M., and Watson, D. E., "ESD Testing of Silicon Wafers," Evaluation Engineering,
September 1991, pp. 115-119.
2. Kolyer, J. M., and Watson, D. E., ESD from A to Z: Electrostatic Discharge Control for
Electronics, Van Nostrand Reinhold, 1990, pp. 173, 175-181.
3. Holmes, O. C., Huff, P. J., and Johnson, R. L., "An Experimental Study of the Screening
Effectiveness of Antistatic Bags," Reliabiliy Analysis Center EOSIESD Symposium Proceed-
ings, Philadelphia, PA, 1984, pp. 78-84.
4. Kolyer, J. M., and Watson, D. E., "The Charged Device Model and Work Surface Selection,"
Evaluation Engineering, October 1991, pp. 110-117.
Paper No. 19
This paper is reprinted, by pennission, from Evaluation Engineering,

HIDDEN CHARGES ON ESD-PROTECTIVE PACKAGING

Defense Electronics
A static charge is a static (immobile) excess or deficit of electrons on a surface. This charge
emanates an electrical field (E field) usually easily detected by a static meter.
However, the E field is weak and difficult to detect when suppressed by proximity to a
nearby conductor as in packaging laminates with nonconductive outer layers sandwiching a
conductive or static-dissipative interJayer. The hard-to-find charges on the laminates are
called cryptocharges (meaning hidden or secret).
A cryptocharge, being immobile, cannot flow into a device lead or tenninal to cause
ESD damage, and the associated E field generally is too weak to cause damage by the field-
induced model (FIM). Buried-metal ESD-shielding laminates can be antistat-depleted and
nonconductive without being an FIM threat. Cryptocharges, however, can be brought out of
hiding by induction.
A conductor brought within an E field is polarized by induction. If the outer end or side
of the conductor is momentarily grounded before the conductor is removed from the field,
the conductor will have an opposite charge to that of the field. This is compound induction. I
Although cryptocharge fields are weak, when a metal plate is pressed to the charged
surface, momentarily grounded, and then lifted from the surface, potentials of up to several
kilovolts appear on the plate. This effect is contact compound induction (CCI).
To study CCI, several laminated and unlaminated films or sheets were tested (Table I).
Material I was a clamshell pack with the film stretched over a 3.8" x 3.8" x 0.3" cardboard
frame with 2.3" x 2.3" windows. Materials 2, 3 and 5 were bags stretched over cardboard in
the same configuration as Material I. Material 4 was a 4" x 6" x 1.1" box.
Samples were electrically cryptocharged for I min with the window area, or the side of
the box, resting on a charged 2" x 2" x 0.03" aluminum plate. A 0.2-lb weight on the sample
facilitated contact. See Fig. I.
Alternatively, samples were tribocharged by stroking them on a nylon carpet at 70-75°F
and 55-65% relative humidity (RH).
CCI was accomplished by pressing a 2" x 2" x 0.03" aluminum induction plate with a
Teflon™ handle against the cryptocharged surface with a force of 0.5 lb for 5 s. Then the
plate was grounded momentarily by being touched with a grounded person's finger before
being lifted from the surface (Fig. 2). The voltage on the plate was read with a Trek Model
512 Field Meter calibrated with a known -7 kV on the same plate.
299
300 APPENDIX
Table 1. Materials.
SURFACE RESlmVITY,
OHMS!SQUARE@ 100 V
AND 65% RH (75°F)
MAn:RIAL TOTAL THICK- SURFACE
NO. DESCRIYI10N NESS, MIL LAYERS INn:RLAYER 0un:R INNER
Clean-skin 1.4 Polyethylene Static- >1013 >1013

shrink film 2 dissipative,
nonmetallic
2 Mil-B-81705, 3.1 Polyester Aluminum 3 x 1010 2 x 1()9
Type III, outside, metalliza-
Class I polyethylene tion,approx.
inside 100 A,
semitrans-
parent
2W Same. but outer surface rinsed with water for 15 s >1013 2 x 1()9
3 Similar to 3.6 Polyester Aluminum 3xl()9 5 x 1()9
Material 2 outside, metalliza-
but opaque polyethylene tion,opaque
inside
3W Same. but outer surface rinsed with water for 15 s >1013 5 X 109
4 Paperboard 16 Paperboard Aluminum I x 1010 3 x lOS
laminate outside, metalliza-
transparent tion, 150 A
biopolymer
inside
5 Polyethylene 2.5 None None >1013 >1013
film
6 Polychloro- 7 None None >10 13 >10 13
trifluoro-
ethylene
(Aclar) film
+OR-
7kV
POWER
SUPPLY
Fig. 1. Deposition of cryptocharge.

PAPER NO. 19 301
Fig. 2. Contact compound induction (CCI).
Table 2 shows that antistatic surfaces were unfavorable for cryptocharging. Material 2
lost its charge relatively rapidly while Materials 3 and 4 held no charge. Thus, cryptocharges
would seem to reside on the outer surface, not inside the interlayer.
Cryptocharges also were localized in the window touching the electrified plate (Fig. 1),
whereas a charge within the interlayer should have spread throughout the sample. This led
to the conclusion that the role of the interlayer is suppressing voltage, not holding charges.
Like other static charges, cryptocharges decayed exponentially because the rate of decay
(-liVldt) was directly proportional to the diminishing potential or voltage (V) that drives
Table 2. CCI Data.

CCI DECAY CONSTANTS
(SEE TEXT)
CHARGING CCI INITIAL DECAY TIME FROM
MA1ERIAL NO. METIlOD VOLTAGE (VO>, kV B,DAr' HVTO I kV
Electrical (-) +5 0.176 0.970 9.1 days

Electrical (+) -5 0.196 0.964 8.2 days
Tribo (+) -7 0.223 0.953 7.2 days
2 Electrical (-) +7 93.6 0.872 25 min
2w Electrical (-) +7 0.421 0.982 3.8 days
Tribo (-) +9 0.657 0.982 2.5 days
3 Electrical (-) 0
3W Electrical (-) +5 0.0871 0.996 18.5 days
Tribo (+) -4 0.0946 0.960 17.0 days
4 Electrical (-) 0
5 Tribo (+) -3 5.71 0.999 6.8h
6 Tribo (-) -5* 0.316 0.963 5.1 days
• Field measured directly. not by CCI.

302 APPENDIX
drainage of charge from the surface and, by means of the E field, attracts oppositely charged
air ions. Then
BV=-dVidt
where B is a constant. Integrating the expression yields:
InV = InVo-Bt
where V is voltage in kV at time t in days, In is the natural logarithm, and Vo is the initial
voltage in kV.
The values of B in Table 2 were obtained by exponential regression analysis of several
data points, r being the correlation coefficient. The decay time in days from 5 kV to 1 kV is
calculated by dividing In5 by B. The half-life of any cryptocharge is In2 divided by B.
The effect of RH on the decay rate is uncertain. Decreasing humidity lowers the rate of
charge drainage over surfaces but may slightly increase the mobility and neutralizing effec-
tiveness of air ions. 2 In these tests, the temperature was 75-80°F and the RH was 55-65% as
measured by psychrometry. The concentration of air ions (conductivity of the air) was un-
known, so the decay rates in Table 2 are only comparative.
Unlike static charges that are not voltage-suppressed, cryptocharges resisted neutraliza-
tion by ionized air from a blower because the E fields were very low (100 V to 300 V on the
surface as sensed by the field meter). Thus, cryptocharges are not only hard to detect but
also hard to remove.
CCI must be distinguished from tribocharging. Uncharged Material 6 strongly tribocharged
the induction plate with a positive voltage by separation, but uncharged Materials 1,2,3,4,
and 5 gave negligible voltages by separation. Also, CCI for Material I was not inhibited by
cementing Material 1 to the face of the induction plate, whereas tribocharging would have
been minimized.
In all cases, momentary grounding was necessary for CCI but would not have been for
tribocharging. With Materials I, 3W, and 5, the polarity of the plate would have been posi-
tive for tribocharging, as it was for a negative cryptocharge; but the plate was negative for
a positive cryptocharge (Table 2). Incidentally, weak CCI charges on the plate occurred
without momentary grounding and might be explained by bleed-off into the air.
The CCI process apparently involved negligible electron flow between the sample and the
induction plate because covering the plate with volume-nonconductive film (Material 1)
did not inhibit CCI. Electron exchange, as opposed to induction, also would have made the
plate the same polarity as the cryptocharge instead of the opposite polarity.
Discharges from the induction plate, with a capacitance of approximately 1 pF when
held by the handle, damaged wafer oxide. 3 The oxidized wafer was the kind from which IC
chips are cut and represents the gate of a MOSFET. Failure occurred when the discharge
created a voltage differential across the oxide in excess of its dielectric strength of approxi-
mately 100 V. See Figs. 3 and 4.
The inverse situation is the charged device model (COM) in which a charged DIP with
1-3 pF capacitance discharges to ground with resulting ESD damage. I
In terms of the l-pF plate, the last column of Table 2 gives the dangerous period for an
initial CCI voltage of 5 kV. This was approximately 3 days for Material 2W, a week for
Material I, and 2.5 weeks for Material 3W.
The persistence of the cryptocharge on Material 3W might be due to effective voltage
suppression by the relatively thick metallic interlayer or to very low conductivity of the
PAPER NO. 19 303
Fig. 3. Wafer oxide test.
!!IS
3/S
2IS
115
VOLTAGE,kV
Fig. 4. Wafer oxide damage from charged induction plate.

304 APPENDIX
washed polyester surface. Antistat depletion by volatilization, absorption by contact with

paper products, or accidental rinsing could transform Material 3 to 3W in the workplace.
In summary, cryptocharges can be deposited electrically or triboelectrically, are some-
times very persistent and, under the contrived conditions of CCI testing, can emerge from
hiding to induce MOSFET-damaging voltages on a conductor pressed to the surface. But
are cryptocharges ever deposited under realistic conditions? And is CCI an ESD hazard or
just a curious phenomenon?
Electrical cryptocharging is possible when packages are handled. A 7-kV discharge from a
200-pF capacitor or the fmger of an ungrounded person was found to charge Material 1 to
several kilovolts as detected by CCI, and ungrounded personnel are easily charged to 7 kV
by walking on a carpet at 25% RH. As for tribocharging, Materials I, 2W and 3W attained
4-9 kV in rather humid air (fable 2) and would charge much higher under dry conditions.
Assuming that a cryptocharge is present, one ESD-hazard scenario resembles field-induced
double jeopardy in which compound induction and possible FIM damage are followed by
CDM damage. I As a hypothetical example, a dual in-line package (DIP) or circuit-board
module is placed in a clamshell pack with a negative cryptocharge on the interlayer.
A few days later, a grounded operator opens the pack and touches a lead of the DIP or an
edge contact of the board, accomplishing CCI and possibly causing FIM damage. Then the
operator removes the positively charged DIP or module from the clamshell pack and again
touches a lead or edge contact to cause a rapid discharge with resulting CDM damage. Such
a sequence of events may be improbable but is not impossible.
Is cryptocharging really a threat? To answer a question with a question: Is the tribocharging
propensity of packaging really a threat? Or are carbon-loaded bags or tote boxes whose
conductive surfaces promote the CDM by rapid discharge really a threat as MIL-HDBK-
773 suggests?
No one knows. So every hazard, albeit remote, must be considered in a rigorous ESD-
control program that fights the invisible to prevent the unmeasurable (subtle damage caus-
ing premature failure).
Since cryptocharges cannot be ignored, laminates such as Materials 1,2, and 3 must be
exceptions to the rule (made with the FIM in mind) that static charges of up to 300 V
measured by field meter are tolerable on nonconductors. 4 Suspect packaging in use should
be checked by CCI and a risk assessment made and precautions proposed.
For relatively insensitive or low-value items, special procedures for removal from the package
without touching leads or contacts might suffice. For items that are highly ESD-sensitive,
expensive, or critical in defense or life-support systems, alternate packaging might be prudent.
A cost-effective choice for avoiding cryptocharges or any form of stored energy is a
volume-conductive shielding laminate such as Material 4. This material never becomes
antistat-depleted, and charges cannot be retained because they bleed not only across the
surface but directly through the layers.
References
1. McAteer, O. 1., Electrostatic Discharge Control, McGraw-Hili, 1989, pp. 58,176 and 179.
2. Jonassen, N., ''The Physics of Electrostatics," distributed at the 1984 EOS/ESD Symposium.
3. Kolyer, J. M., and Watson, D. E., "ESD Testing of Silicon Wafers," Evaluation Engineering,
September 1991, pp. 115-119.
4. Kolyer, J. M., and Watson, D. E., ESD from A to Z: Electrostatic Discharge Control for Elec-
tronics, Van Nostrand Reinhold, 1990, pp. 160, 161, 172 and 210-215.
PaperNo. 20
This paper is reprinted, by pennission, from the Proceedings of the 1994 EMCIESD
1nternational Conference, held in Anaheim, CA, April 12-19, 1994.
TOWARD AN IDEAL ESD-PROTECTIVE PACKAGE
John M. Kolyer

Autonetics Electronic Systems Division
Anaheim, California
Abstract
An ideal electrostatic discharge (ESD )-protective package would shield against worst-case
external discharges and be permanently static-dissipative and volume-conductive. Also, it
would have non-ESD properties such as recyclabi/ity, biodegradability, and flame-resis-
tance. This paper discusses state-of-the-art packaging and proposes superior designs ofthe
future that may be achieved, without major breakthroughs, by consolidation and extension
ofpresent techniques. Basic to this development will be realistic ESD tests, several of which
are described with examples of data.
Introduction
The title of this paper says "toward" because an ideal product is, by definition, impossible. It
would have to include conflicting properties as in "biodegradable, heat-sealable, low-cost,
transparent, stainless steel." Furthennore, it would have to be universally suitable for count-
less applications with specialized requirements. However, laminations of paper, plastics, and
metals can combine diverse properties and approach an ideal package for specific purposes.
This paper discusses key electrostatic discharge (ESD) properties and realistic tests for
them. Non-ESD properties are also reviewed. Then promising packaging designs that are
now available or on the horizon are discussed, with pros and cons of the two major catego-
ries, plastic and paper.
ESD Properties Desired

ESD properties have been discussed' in tenns of damage mechanisms: Human Body Model
(HBM), Machine Model (MM), Charged Device Model (CDM), and Field-Induced Model
(FlM). See Refs. 2 and 3 for explanations of the models.
The key ESD properties desired are high-voltage discharge resistance, CDM safety, low
current-carrying capability to protect personnel as well as devices, a static-dissipative surface
even at low relative humidity (RH),permanence of the static-dissipative property (including
305
306 APPENDIX
lack of undesirable antis tat transfer to other surfaces), abrasion resistance (no loss of static-
dissipativeness and no sloughing of conductive particles), minimal triboelectric charging
propensity for nonconductive surfaces touched by the packaging material, and volume con-
ductivity to prevent stored energy. These properties will be further explained in the follow-
ing discussion of tests.
Realistic Tests for Key ESD Properties

Nine tests have been developed as practical criteria for the above key properties. In the
following discussion, the tests are referenced, briefly described, and illustrated with data on
commercial materials.
High-Voltage Discharge Resistance. External discharges can damage ESD-sensitive

items, through the wall of a package, by the HBM or MM. A realistic tesr uses a 3M Static
Event Detector (SED) (Fig. 1) inside a bag or box (Fig. 2). A charged steel probe touched to
the package represents either a sharp toolS or a blunt tool held by a person (Table 1); the
blunt tool throws a longer, more damaging arc. A ceramic capacitor and a person serving as
the capacitor give similar results. Discharge protection ratings are assigned according to
Table 2. This test is realistic because a spark is thrown, as often happens in reality. The
Electronic Industries Association standard EIA 541 test, with large, flat electrodes, low volt-
age (1 kV), and no spark, allows metal-out, partially transparent bags to pass, but the metalli-
zation is destroyed by a spark from a fmger (Fig. 3). A potential of 8 kVon a person, as shown
in Fig. 3, commonly results from walking on a nylon carpet at 30% RH.
CDM Safety. In the CDM damage mechanism, a charged ESD-sensitive device discharges
to a conductive surface. Tests have been devised to simulate this process,6-9 and Fig. 4
shows a method using an SED.· An indirect method for judging CDM safety is to measure
the resistance to ground of materials with a buried conductive layer (Fig. 5) or resistance
between electrodes for other materials (Fig. 6). The resistance reading can be related to tests
with metal oxide semiconductor field-effect transistors (MOSFETs) in simulated CDM
events.6 Table 3 gives examples.
Current-Carrying Capability. The resistance is measured between two 5-lb, 2.5-inch-

diameter National Fire Protection Association (NFPA) 99 electrodes on the surface of the
Fig. 1. Protected SED.

PAPER NO.2O 307
Fig. 2. High-voltage discharge test.
Table 1. Test Conditions.

SHARP TOOL BLUNT TOOL
Voltage,tV 10 6 10
Capacitance, pF 200 350 350
Probe diameter, inch 0.025 0.25 0.25
Table 2. Discharge Protection Ratings.

BLUNT-TOOL 'I1!ST
SHARP-TOOL
'I1!ST 6kV 10kV RA11NO
F Poor
p F F Fair
P p F Good
P P P Excellent
Code: p =Pass, F =Fail
Fig. 3. Discharge tests on metal-out bags.

308 APPENDIX
Package
Surface
'--
Discharge
Probe
Tel
Fig. 4. CDM test schematic.
material (Fig. 6). Resistance for a paperboard packaging material versus voltage and RH is
shown in Fig. 7. Table 4 gives current calculated from resistance for several materials.
Obviously, low current-carrying capability is desired for safety for people as well as de-
vices. An alternating current of 4-21 rnA causes reflex action (which could result in injury),
21-40 rnA causes muscular inhibition, and 40-100 rnA causes respiratory block.1O there-
fore, only the fIrst material in Table 4 is a current-carrying hazard for personnel at Ito V.
Surface Resistivity versus RH. Surface resistivity values are only approximate because
the measurement depends on factors such as the pressure of the electrodes against the sample.
Nevertheless, this property is useful as a guideline. At approximately 1012 ohms/square,
Megohmmeter
~
LL2
Sample
Buried Conductive Layer
Fig. 5. Resistance to ground.

PAPER NO. 20 309
ohmmeter
~
LD
Fig. 6. Current-carrying capability.
materials become nonconductive and hold static charges for several seconds or more. Re-
sistivity rises with falling RH, and the RH at which 1012 ohms/square is reached is the cutoff
for retention of the antistatic property. Surface resistivity at various RH levels is measured
by sealing samples above humidity-regulating saturated chemical solutions (or anhydrous
calcium sulfate for 0% RH) as diagrammed in Fig. 8 and explained in detail in Ref. 11.
Examples of curves of surface resistivity versus RH are shown in Fig. 9. A low-RH cutoff is
desired for packaging materials because dry air may be encountered in shipping and han-
dling. During unpacking, for example, triboelectric charges not drained to ground (through
the work surface or operator's fingers) could cause damage by the FIM.
Permanence of the Anmtatic Property by Solvent Extraction. Materials depending on fu-

gitive antistatic additives (antistats) can become nonconductive and hold static charges when
the antistat is lost by volatilization, absorption by contact with materials such as paper
products, or solvent extraction. Table 5 gives examples of surface resistivity measurements
Table 3. ResIstance to Ground and COM Danger.

RESISTANCE PREDICTED MOSFET
MATERIAL RH(%) (OHMS AT 10 V) FAILURE RAl1l (%)
Conductively coated paperboard 50 3xl(}l 87

with thin antistatic varnish, Brand A
Conductively coated paperboard 40 8 x 1()3 80
with thin antistatic varnish, Brand B
Paperboard impregnated with carbon 50 2 x 1()6 37
Paperboard with buried 43 2xlOS 0
conductive layers
Conductively coated paperboard 36 6 x 109 0
with heavy antistatic varnish
310 APPENDIX
10 8
en
E
.c
0
cC 10 7
• 10V
106
• l00V
0 _ _ _ 500V
o 10 20 30 40 50 60 70 80 90 100
R!i%
Fig. 7. Resistance to ground for cardboard with buried conductive layer.
before and after exposure of samples to two common solvents. The antistatic coating failed
because the antistat in it was easily leached.
Antistat Transfer. Transfer of antistat to the surface of packaged items can cause unde-
sirable effects such as lowering of surface impedance of resistors, stresscracking of poly-
carbonate plastic, and discoloration of epoxy paint. 2 In a simple test (Fig. 10), samples are
pressed against nonconductive plastic (cellulose acetate) sheets, which become antistatic if
antistat is transferred. In Table 6, antis tat is seen to have been transferred from an antistatic
Table 4. Current-Carrying Capability at 110 V.

RESISTANCE CURRENT
MATERIAL RH(%) (OHMS) (MA)
Conductively coated paperboard 43 3 x l()1 32

with thin antistatic varnish (measured at I v)
Paperboard impregnated with carbon 50 I x 1()6 0.11
Paperboard with buried
conductive layers 43 9x 10' I X 10-3
Antistatic viny I 54 Ix 10" I x IQ-4
Conductively coated paperboard 36 3xl0" 3 x 10-5
with heavy antistatic varnish
PAPER NO. 20 311
coating. Superior antistatic materials are inherently antistatic, without depending on an ad-
ditive, or they contain a polymeric or chemically bound additive that cannot migrate.
Abrasion ResIstance. This property is measured with a Thber Abraser in which an abra-
sive wheel presses the surface as shown in principle in Fig. 11. The difference in cycles
required to wear holes through the liner (skin) of the four conugated paperboard samples in
Thble 7 is caused by varying liner thickness. Superficial conductive coatings were quickly
removed, whereas a buried conductive layer was not reached for hundreds of cycles. Abra-
sion is of concern because it might cause loss of BSD shielding and, more importantly,
result in sloughing of conductive particles that could bridge circuit lines in open devices.
Triboeledrle Charging. This property is notoriously erratic and unreproducible. Rather

than attempting to bracket all materials with the extremes of the triboelectric series, quartz
and TeflonTN, testing with the materials of concern in practice is recommended. A method
has been described' in which coupons of circuit-board laminates (epoxy or polyimide) are
stroked against a packaging material and the voltage on the laminate is read with a static
meter.1b prevent buildup of contamination, the least amount of which can give false charges,
the coupons are cleaned with isopropyl alcohol between determinations. Because of the
scatter of the data, as seen in Thble 8, the best that can be done is to sort samples into the
rough categories of low, high, and medium chargers. Very low chargers, e.g., antistatic
~
~
Fig. 8. Surface resistance vs. relative humidity.

312 APPENDIX
.
.......
~
>
...~
cI
I
1=PIp8I'boenI
2 • ConducIIvely coated papeltloerd with
heavy antIatatlc vamlah
3. hpeltMMlrd with bUried conductive
layers
4. Paperboercllrnpragnated with carbon
105 . . . .~~__. .__. .____. .~. .

o 10 20 30 40 50 60 70 80 90 100
RH,%
Fig. 9. Surface resistivity vs. relative humidity.
polyethylene, may owe their success to the undesirable transfer of antistat, causing liquid to
separate from liquid on a microscopic level. For rigid containers, which do not rub the
packaged item on loading or removal as bags do, a medium-charging rating should suffice.
Crypotocharges. These hidden charges are an example of stored energy. They are diffi-
cult to detect with a static meter and occur on packaging material with a buried conductive
layer that suppresses the voltage on the charged (electron-rich or electron-depleted) non-
conductive surface. 12 Cryptocharges are brought out of hiding by contacting the surface
with a conductor, which becomes polarized, and then momentarily grounding the conduc-
tor, e.g., by touching it as shown in Fig. 12, so that electrons flow on or off the conductor to
PAPER NO. 20 313
Table 5. Solvent Extraction Tests.

SURFACE RESISTIVITY
(OHMS/SQUARE, 100 V. 31-41% RH)
AF"lER ONE WEEK
ISOPROPYL
WATER RINSE WATER SOAK ALCOHOL SOAK
MATERIAL ORIGINAL (30 SECONDS) (30 MINUTES) (30 MINUTES)
MIL-B-SI705C. Type I or 10'0 >1013 >1013 >1013

TypeID
Plastic with antistatic coating 9 x lOS >1013 >1013 >10'3
Paperboard with buried 3 x 109 2 X 109 4 x lOS 3 x lOS
conductive layers
Conductively coated 1 x 10'0 1 X 1010 9 x 1Q9 2 X 1010
paperboard with heavy
antistatic varnish
give it a net charge. This charge then manifests itself as an E field (electrical field) when the
conductor is lifted from the surface. This special case of compound induction. called con-
tact compound induction (CCO. can impart several kV to the conductor, which could dam-
age ESD-sensitive items by a direct discharge according to the MM or, if the conductor is
part of an ESD-sensitive item, by the CDM. Scenarios for damage to devices by cryptocharges
are unlikely but must be considered, especially for critical hardware such as defense or Iife-
support systems.
Non-ESD Properties Desired
Transparency. This property is easily obtained with plastics but not with paper products.
Cellophane, like paper. is cellulosic, and the correct grade, e.g .• plasticized and nitrocellu-
lose-coated. is permanently antistatic, noncorrosive, and heat-sealable but has low tear
strength and is best used as a window, liner, or intimate wrap. '3
Samples
Fig. 10. Antistat transfer.

314 APPENDIX
Table 6. Antlstat Transfer Teste.

SURFACE RESlSnvITY (OHMSISQUAItE,
100 V. 36-59'1 RH) OF CEU.ULOSE
ACETA11! (ORlOlNAUY 1014)
MA11!RIAL AFI1!R ONE DAY AFI1!R ONS WEEK
Plastic with antistatic coating 3 X 1010 1 X 1010

M1L-B-8170SC. TYpes I or III; antistatic vinyl; paperboard >10" >10"
with buried conductive layers; conductively coared
paperboard with heavy antistatic varnish
Low Corrosivity. A stringent test13 is to expose copper or Sn63 solder in contact with a
packaging material at 100% RH for three months at 7S-80"F or one month at 1lO°F. There
are many standard tests. e.g.• FED-STD-IOI. Method 3005. Paper products should have a
low reducible sulfur content by the Technical Association of the Pulp and Industry (TAPPI)
test T406 or FED-STD-IOl, Method 3005.
Low Toxicity. By Committee of Northeastern Governors (CONEG) regulations, heavy

metals, e.g., lead, mercury, cadmium, and hexavalent chromium, cannot exceed low levels
in packaging. These metals can be minimized in both plastic and paper.
Strength and DurabiUty. The American Society for Testing and Materials (ASTM) has
many standard tests for properties such as tensile strength (ASTM D 882), tear resistance
(ASTM D 1004, D 1922). puncture resistance (ASTM D 2582), impact resistance (ASTM
D 3420, D 4272), and resistance to being dropped (ASTM D 959, D 4(69). Less general
properties, e.g., score resistance, are important in specific applications.
tr,\ Rotating
-
8 Sample
Fig.ll. Abrasion resistance.

PAPER NO. 20 315
Table 7. Taber AbraserTest.

CYCLES 1'0 ABRADE CYCLES 1'0 CREATE
RESISTANCE (OHMS)·
(CAUSE DULL HOLES IN 3% OF
SURFACE) 50% OF UNABRADED AREA UNDER
MATERIAL AREA UNDER WHEEL AREA ABRADED AREA WHEEL
Conductively coated 10 3 x l()l Ix10"-3x1OS llOO

paperboard with thin
antistatic varnish
Paperboard impregnated 140 4 x l()l 9 x l()l 540
with carbon
Paperboard with buried 470 >2 x 109 >2 x 10" 700
conductive layers (3 x 10" inside
holes)
Conductively coated 2 I x 10'0 5 X 109 1050
paperboard with heavy
antistatic varnish
• AI 1V. measured between 0.063-inch-diameler conical probes 0.2 in. apart.
Recyclability. Plastics are recyclable but require sorting for best results; a single type of
plastic can be added as regrind to virgin plastic for molding, but mixed plastics can be used
only for noncritical items such as low-pressure sewer pipe. Another approach is to depoly-
merize mixed plastic waste to basic oil feedstock. 14 Germany is especially strict about recy-
cling plastics, with a goal of 80% of plastic packaging to be recycled by July 1995. Of
course, reuse of plastic items is preferred to recycling; some companies reuse plastic trays,
for example, to lower costs and protect the environment. Paper packaging has an advantage
in recycling because it need not be sorted (unless, for example, it has a high carbon content)
and can be mixed with general paper waste, such as newspapers, for repulping. The solid-
waste debate is discussed in a recent book" that challenges many conventional positions.
For example, conversion by U.S. fast-food chains from disposable to reusable packaging
would require consumption of huge arnounts of water, detergent, and energy for dishwashing.
Biodegradability. This controversial property requires moist conditions. Dry landfills

contain intact paper products after decades, but this is no surprise (witness Egyptian pa-
pyri), and it might be argued that landfill material beneath permanent building sites need
never degrade. In the United Kingdom, abandoned coal mines provide low-cost, permanent
disposal sites for all ordinary waste. IS The biodegradability of paperboard packaging is
superior to that of wood or leaves and is at least useful in preventing long-term littering of
the landscape.
Flame Resistance. There are many tests for plastics, e.g., Underwriters Laboratories test
UL94 for heavy sections or ASTM D 568 or D 1433 for sheet and film, and for paper
products, e.g., ASTM D 4433 for fire-resistant, treated paperboard. A new fire-retardant,
antistatic, recyclable, and biodegradable treatment for liner board for packaging was an-
nounced late in 1993. Some plastic materials are inherently fire-resistant, for example, one-mil
316 APPENDIX
Table 8. Trlboelectrlc Charging Test.
RATING
OVERAU (PROPENSITY ro
VOLTAGE ON COUPON (tV). 11fREE
AVERAGE (kV). CHARGE CIRCUIT
READINGS AND AVERAGE
NEGLECTING BOARD
MATERIAL EPOXY I EPOXY 2 POLYIMIDE SIGNS LAMINATES)
Plain paperboard +1 +1 -7
(corrugated cardboard) +1 +2 -1 3 Medium
±l ±ll ::2
+1 +S -3
Carbon-loaded polyethylene +4 +18 +3
+1 +11 +1 8 High
ill ...n ±2
+S +16 +2
Antistatic polyethylene 0 0 0
(Mll..-B-8170SC, Type II) 0 0 0 0 Low
Q Q Q
0 0 0
Paperboard with buried +2 +2 -4
conductive layers +1 +4 -2 3 Medium
1:3. +7 =l
+2 +4 -2
Paperboard impregnated +1 +10 -1
with carbon +O.S +12 -2 4 Medium
-.Jl ill ::l.
+1 +11 -1
Paperboard with heavy +14 +14 +1
antistatic coating +11 +11 +1 8 High
+11 ±ll .J!
+12 +12 +1
polyvinyl chloride (PVC) packaging film that is made permanently antistatic by a propri-
etary polymeric additive. Aluminum trihydrate and magnesium hydroxide are environmen-
tally friendly flame retardants because they function by endothermically decomposing to
release harmless water rather than to toxic gases. 16 Other inorganic, nonhalogenated flame
retardants include intumescents like ammonium polyphosphate.1 7 All these additives can be
used in (or on) both plastic and paper packaging.
Other Properties. The many other properties of importance in packaging include perme-
ability, markability, heat-sealability, and gluability with both resin and hot-melt adhesives.
It is important to realize that just one "Achilles heel" in some special property can dis-
qualify an otherwise ideal package. In fact, it is questionable how many applications a
single "ideal" package design could satisfy. The "ideal package" may in fact be a family of
designs with intrinsic advantages, e.g., permanent antistaticity and volume conductivity, in
common.
PAPER NO. 20 317
Fig. 12. Contact compound induction (CCI).
Promising New ESD-Protective Package Designs
Plastic. Conventional plastic packaging is typified by MIL-B-81705C (Table 9), which

depends on antistatic additives and is volume-nonconductive. However, a commercial poly-
meric antistatic additive confers permanence (with no antistat transfer to other surfaces)
and volume conductivity (to avoid stored energy such as cryptocharges) to a variety of
thermoplastics including acrylonitrile-butadiene-styrene (ABS), acrylic, polycarbonate, PVC,
nylon, polyester, thermoplastic polyurethane, and polypropylene. PVC and nylon versions
were found to be antistatic at 0% RH when measured by the method of Fig. 8. New products
of this type have appeared as both sheets and thin films. Another approach to permanent
antistaticity involves dispersed powdered metal, e.g., copper, or metallic or carbon fibers.
Some plastic packages use protective spacing (air gaps or antistatic foam between the wall
and the packaged item) to mitigate the effect of external discharges. s In one design, a trans-
parent, antistatic bag is simply inflated with air. Spacing in conjunction with a fair shielding
layer can raise the discharge rating to good (Table 2), and spacing in conjunction with a
good shielding layer can raise the rating to excellent.
Paper. Paperboard has the advantages of being inherently volume-conductive and biode-
gradable, as well as reusable and recyclable. Its weaknesses are permeability, opacity, flam-
mability, and limited durability. Several commercial corrugated paperboard designs, which
are stiff, strong, and lightweight, were tested with the following results. Paperboard with
carbon impregnated throughout the skin was too conductive (Table 3, Fig. 1), but conductively
coated paperboard with a heavy antistatic varnish was CDM-safe (Table 3), safe to person-
nel (Table 4), and permanent (Tables 5 and 6). Though the varnished paperboard lacked
high abrasion resistance (Table 7), a highly conductive surface was never exposed.
Conductively coated paperboard with a thin varnish was undesirably conductive (Tables 3
and 4). A particularly effective design has a conductive carbon layer on the inside of each
318 APPENDIX
Table 9. Plastic vs. Paper Packaging.

PAPERBOARD WITH BURIED
CONDUCTIVE LAYERS
MIL-B-81705C METALLIZED
(PLASTIC) CARBON PAPER RJTURE
PROPERTY TYPE I TYPE III LAYER LAYER PLASTIC PAPER
High-voltage Excellent Good Fair* Excellent Excellent Excellent

discharge resistance (good with
(Table 2) O.5-in. anti-
static foam
between
wall and
packaged
item)
CDMSafety Excellent Excellent Excellent Excellent Excellent Excellent
Volume conductivity No* No* Yes Yes Yes Yes
(to prevent stored
energy)
Antistat transfer No No None None None None
Permanence No* No* Yes Yes Yes Yes
(solvent extraction)
Triboelectric Low Low Medium Medium Low- Low-
charging Medium Medium
Biodegradability No* No* Yes Yes Yes Yes
RecycJability Yes Yes Yes Yes Yes Yes
Transparency No* Yes No* No* Yes Yes
(cellophane
window)
Vapor barrier Excellent Fair Poor* Poor* Excellent Good
(permeability)
Flame resistance Poor* Poor* Poor* Poor* Good Good
Toxicity Low Low Low Low Low Low
Durability High High Fair* Fair* High Medium
Costlow Low Low Medium Medium- Medium-
High High
Overall comments Definite deficiencies, Definite deficiencies, In general, approaches
e.g., permanence, e.g., opacity, permeability the ideal, but cost
stored energy to vapors will not be low
(reuse will help)
• Deficiency. In some cases, e.g., biodegradability, criticality is arguable.
liner (skin) in contact with the medium (fluted paper). Thus, the conductive layer is safely
buried and cannot be easily abraded or be a CDM or personnel hazard. This design has
generally excellent ESD properties (Tables 3-7 and 9, Figs. 7 and 8) but gives only fair
shielding against external discharges. Excellent shielding is achieved in this design if
metallized paper is substituted for the carbon layer, while all the other desirable proper-
ties are retained (Table 9); aluminum is, in this context, nontoxic, and the amount used
(approximately 1000 A) is too small to interfere with repulping or biodegradation. Two
layers of metallized paper, one inside each liner, are needed for the excellent shielding
PAPER NO. 20 319
rating; a spark discharge creates a hole in the metallization on the first layer but leaves the
second layer intact and is intercepted by it. Metallized cellophane is a barrier to gases but is
more expensive than metallized paper and, not being fibrous, can cause fish-eyes in re-
cycled paper. Aluminum foil, incidentally, gives excellent high-voltage discharge protec-
tion but presumably interferes with repulping.
Future ESD-Protective Package Designs
Plastic. The antistat permanence and volume conductivity achieved for plastics in recent
years has been a major breakthrough. This corrects the deficiencies in these properties found
in MlL-B-81705C, Types I and III, which are used as a reference point in Table 9, and
leaves only lack of biodegradability and lack of flame resistance. These may not be major
deficiencies, but efforts can be expected to correct them, at least for special packaging
applications. In fact, plastic packaging already available meets most of the "future" ratings
in Table 9--{)r can meet them if the right materials are combined. For example, aluminum
metallization buried between layers of the PVC film mentioned above might give a perma-
nently antistatic, flame-resistant version of MIL-B-81705C, Type I. Biodegradability is a
more "blue sky" property because, for one thing, the plastic might degrade prematurely on
the shelf or in use, but new products are appearing on the market. IS Also, transparency and
excellent (rather than good) high-voltage discharge resistance are incompatible properties
with present methods of metallization, and an extremely good vapor barrier requires metal
foil, an opaque layer of vapor-deposited metal or, possibly, a layer of expensive
polychlorotrifluoroethylene (PCTFE) transparent film.
Paper. Paper has the inherent limitations of being opaque and highly permeable. Coated
cellophane is transparent and a good vapor barrier and is volume-conductive and biode-
gradable, 13 but it is not recyclable for repulping (only for producing more cellophane). Flame
retardants can be added to paperboard, but corrosivity must be avoided. A flame-retardant
treatment for liner board recently announced claims to be colorless, odorless, nontoxic,
biodegradable, and recyclable. Durability of paperboard packages can be improved by us-
ing heavy sections, but there will be penalties in weight and cost. The paperboard with
buried metallized paper proposed in Table 9 remains to be commercialized and would be a
good future product on which to improve with flame retardants, etc.
Plastic versus Paper. In general, plastic is stronger and more durable than paper, but
paper is more "natural"; it is made from a renewable resource, is biodegradable, and is
easily recycled. Also, in regard to ESD, paper has the advantage of being permanently
antistatic and volume-conductive, but state-of-the-art plastic compounds also have these
qualities. Therefore, paper and plastic are moving toward the ideal package from differ-
ent positions. In the foreseeable future, neither material will preempt the other, and paper
and plastic constructions will coexist along with mixed constructions, e.g., paperboard
boxes lined with antistatic plastic foam. Of course, combinations like the latter could be
recycled only by separating the materials. A foam-lined paperboard box would be reused
as many times as possible and then peeled apart for separate disposal of the foam and
paperboard.
320 APPENDIX
Conclusion
Much progress toward an ideal ESD-protective package has been made in recent years.
Advanced materials are available, realistic tests have been devised. and ingenious, cost-
effective packages can be created by metallization, lamination, incorporation of cushion-
ing such as foams, rigid mounting of items to prevent triboelectric charging, protective
spacing with air gaps, and other design features. Biogradability for plastic packages will
be difficult to achieve but may be unimportant, while flame retardance for either plastic
or paper can be attained to a considerable degree if demanded. Certainly, overly conduc-
tive surfaces, less than good high-voltage discharge resistance, fugitive antistats, and
stored energy will be - or should be - problems of the past.
Without major breakthroughs, consolidation and extension of present techniques should give
products approaching the "ideal" packages in Table 9. These may be more expensive than
current packages, but greater durability as well as emphasis on reuse will lower the effective
cost. Meanwhile, less toxicity and more recycling and reuse will protect the environment.
References
I. J. M. Kolyer, "Fundamentals of ESD Control," EMC Technology Magazine 1992 Expo Techni-
cal Record, May 1992, pages 154-161.
2. 1. M. Kolyer and D. E. Watson, ESD from A to Z: Electrostatic Discharge Controlfor Electron-
ics, Van Nostrand Reinhold, 1990.
3. 0.1. McAteer, Electrostatic Dischange Control, McGraw-Hili, 1989.
4. 1. M. Kolyer, "Realistic Testing of ESD Materials," EMC Test and Design, September 1993,
pages 28-31.
5. 1. M. Kolyer and D. E. Watson, "Packaging for High-Voltage Discharge Protection," Evalua-
tion Engineering, March 1992, pages 96-100.
6. 1. M. Kolyer and D. E. Watson, "COM and Work Surface Selection," Evaluation Engineering,
October 1991, pages II 0-117.
7. D. C. Anderson, "A Simple Approach to ESD Damage Prevention," EMC Technology, Marchi
April 1991, page 38.
1990, page 77.
9. J. M. Kolyer, "Is Your Work Surface COM-Safe?," EOSIESD Technology, February/March 1992,
pages 27-28.
10. DoD-HDBK-263, May 2,1980, page 46.
II. 1. M. Kolyer and R. Rushworth, "Humidity and Temperature Effects on Surface Resistivity,"
Evaluation Engineering, October 1990, pages 106-110.
12. J. M. Kolyer and D. E. Watson, "Hidden Charges on ESD-Protective Packaging," Evaluation
Engineering, September 1992, pages 94-100.
13. News notes in EOSIESD Technology, OctoberlNovember 1989, page 9, and Evaluation Engi-
neering, March 1990, page 96.
14. P. L. Layman, "Advances in Feedstock Recycling Offer Help with Plastic Waste," Chemical
and Engineering News, October 4, 1993, pages 11-14; News note: "Group Formed for Recy-
cling Plastics into Feedstocks," ibid, November 29,1993, page 41.
15. 1. H. Alexander,ln Defense of Garbage, Praeger Publishers, 1993.
16. S. Ainsworth, "Magnesium Oxide Finds New Applications," Chemical and Engineering News,
October 25, 1993, pages 15 and 16.
17. "Flame Retardants: Processors Learn to Work With Halogen-Free Systems," Modern Plastics,
18. "New Players Emerging in Biodegradable Polymers," Modern Plastics, October 1994, pages
33-37.
Index
Abrasion resistance, 311; illus., 314; grit-blasting, 68

table, 315 paper forms and labels, 58
Aclar shipping containers, 71
charged by heat gun, 65 vinyl chair, 58
in cryptocharge test, table, 300 categorization of, table, 23
in operator charging test, 245 definition of, 23
in special test on chain link fence, discussion of, 26
69 Antistat, topical. See 1bpical
in special test on walls, 58 antistat
usually negatively charged, 43 Antistatic
Air approach to ESD control, 28, 196
space for ESD protection, 296; bags and tore boxes, permanence
table, 295; table, 297 of, 175-180; table, 179; table
test for triboelectric charging by, 313
12,66 box, transparent, 101
Air guns, nuclear and electrical, bubble-wrap,
·63-64 criteria for selection, 100
Aluminum foil decay of charge on, by room
for conductive masking tape, 193, ionization, illus., 193; table,
198 192
for shielding, 39 for cushioning, 48, 173
for SSP, 48; table, 262; table, 275; categorization of materials, table,
table, 295 23
tape for holding shields on module, clean nylon, criteria for selection,
68 99
Amber, charging of, 3 cubical container, 101
Analysis definition of, 23, 83, 110
applied to discussion of, 27
automated process, 72 foam,
brushes, 60 criteria for selection, 100
charged windows, 59 shunt, 104
cleanroom gloves, 62 testing of, table, 275; table, 295
321
322 INDEX
Antistatic (continued) measured in disposition of

intimate wrap in SSp, 48 mishandled hardware, 145
materials, on circuit-board laminates, by
disadvantages of, 28, 159, sliding on stainless steel, 66
238-243 on walls, 91
in future, 159 ARTG
pennanence and shelf life of, categorization of, table, 23
175-180; table, 179 definition of, 24
resistance measurements on, 170 discussion of, 27, 207, 210, 250;
plastic liner for trash cans, 92 table, 157
polyethylene, in model handling and assembly
cleaning of, 89-90 specification, 81
criteria for selection, 99 lower limit of one megohm, 81
discoloration of paint by, 242 lower limit violated with wet shoes
effect on adhesive bonds, table, on conductive floor, 91
241 upper limit of 10 megohms, 81,
fogging of instrument mirrors by, 210,250
239,241 Assemblies, definition of, 89
precautions with, 89-90 Assembly and handling, model
testing of, 263, 286; illus., 287; specification for, 77-106
illus.295 Audits and reviews, 87-88
smocks, 100; table, 141 Automated process
surfaces, 253 control of hazards in, 72, 88
vinyl, 75-76, 315-316 coupons for, 72, 233-237
workbench top, 185 monitoring of, 237
Antistatic lotion recertification of, 88
categorization of, table, 23
criteria for selection, 105 Badges, identification
definition of, 23 among necessary nonconductors in
discussion of, 27, 245, 250 SSW, 84
Antistats hazard of, 57-58
corrosion and contamination by, treated with topical antistat, 58,
238-243 84
effect on adhesive bonds, table, 241 Bag materials
transfer of, 310-311; table, 314 commercial, 201; table, 202
Apparel of operators. See Clothing, conclusions from tests on, 170-174,
operator, and Smocks, 205-206,294-296
ESD-control criteria for selection, 99-100
Apparent charge in future, 159-160, 174, 319
categorization of, table, 23 properties of, table, 202; table, 205;
definition of, 24 table, 295; table, 300; table,
discussion of, 27; table, 157 318
in special test of air guns, 64 screen layer for excellent shielding,
in special test on chain link fence, 206; table 202; table, 205
69 tests on, 274; illus., 204; illus.,
INDEX 323
275; illus., 295; illus., 307; for boxes for SSP, 48, 311; table,
tables, 195; table, 203; table, 275; table, 295; table, 300
205; table, 227; table, 262; in packaging specifications, 111
table, 275; table, 295; table, tote box, cost of, table, 142
307 Carpet
Bags, Faraday-cage. See Faraday antistat-treated, 18, 83, 91
cage charge from walking on, 14, 18, 73;
Basic Rule illus., 9; mus., 17; table, 36
categorization of, table, 23 charge generation on plastic by, 299
central in operator training, 105, with conductive fibers, criteria for
106 selection, 73, 102
definition of, 24, 80, 109 Carriers, requirements for, 111
discussion of, 28 Cathode ray rubes (CRrs)
importance of, 1-2 charges on, 59
in model handling and assembly screens for, 59, 105, 141
specification, 80 Caution tag, 87
in model packaging specification, CCI (Contact Compound Induction),
109 299,313; mus., 301; illus.,
Benchtops. See Workbenches 317
Bins, criteria for, 101 CDM
Biodegradability, 315 categorization of, table, 23
Bleed-off time (decay time), 41, 264, definition of, 24
277 discussion of, 7-8, 16, 29,44-46,
Blow-off nozzles, 37-38, 73; table, 115, 162, 253, 254, 260,
190 263-271, 281, 285, 305, 306;
Boxes, criteria for selection, 101 illus., 7; mus., 15; illus., 277;
Boxes, tote. See Tote boxes allus., 308; table, 309; table,
Brushes 318
Criteria for selection, 105 involving operator, 45, 162, 186,
evaluation of four types, 60-61 285
Bubble-wrap, antistatic. See Anti- test on tote boxes, 231; table, 229
static bubble-wrap CD Rule
Buyers' guides for packaging mate- applied to
rials, 49 brushes, 61
clothing, 57
Capacitance gloves, 85-86
categorization of, table, 23 screwdriver handle, 61
definition of, 24 walls, 58, 91
discussion of, 29 windows, 59
of capacitor (150 pF) used in as primary defense, table, 158
shielding/discharge test, 183, categorization of, table, 23
294; table, 307 definition of, 24, 198
probe used in shielding/discharge discussion of, 29,45, 116
test, 48, 262; illus., 295 experimental derivation of,
Cardboard 188-189; illus., 189
324 INDEX
CD Rule (continued) static charges on, 10, 57, 162

for 12-volt-sensitive devices, equa- Complementary equipment/materials
tion and table, 161-162 and techniques, table, 197
in model handling and assembly Components, definition of, 89
specification, table, 84 Conductive
in operator training, 105 approach to ESD control, 28, 196
Cellophane, 74, 286, 292; mus., 288; boxes, 101; table, 309; table, 310
mus., 293 categorization of materials, table,
Certification 197
of ionization equipment, 88 chairs, 102, 139, 141, 156,254;
of operators, 55, 86, 88, 105-106 table, 197
of packaging, 113 definition of, 24, 83, 109
of SSWs, 51, 80, 88, 93-94 discussion of, 30
Chain link fence fibers in carpet, 73, 102
for shielding, 69-70 floor mat, 82, 102, 139, 141
near SSZ, 86, 129 floor tile, 75, 102; mus., 74
Chair, vinyl, static charge on, 58; floors,
table, 187 hazard of, 91, 132, 143
Chairs, conductive. See Conductive in model handling and assembly
chairs specification, 91
Charged "finger" test, mus., 171 foam for shunt, 40,104
Charged Plate Monitor, 67, 151 footwear, 72-74; table, 139; table,
Charges, hidden. See Cryptocharges 141
Check lists, 148, 150, 151 hazard by carrying lethal current,
Chloride ion 143,306,308; mus., 309;
corrosion of silicon wafers by, 42, table, 227
52 heel strap, 100
in workbench laminates, 63 masking tape, 193, 198
Circuit-board laminate, charging of materials,
by grit-blasting, 68 disadvantages of, 30, 143, 178,
by rubbing with antistatic packag- 180,226
ing, 176; mus., 177 in future, 160, 319
by sliding on stainless steel, 66 when to use, 53, 143, 196-197
Classification of ESDS Items, 79 packaging materials, high tri-
Cleaning of antistatic polyethylene, boelectric charges generated
89-90 by, 178, 226, 279; table, 228;
Cleanliness table, 279; table, 316
in model packaging specification, plastic, table, 262
lll, 114 polyethylene bags, 263; table, 275;
of SSW, 85, 90 table, 282; table, 295
Cleanroom gloves. See Gloves, polyolefin, criteria for selection, 99
cleanroom seat covers, 102
Clothing, operator shunt for conectors, plastic, 40, 62,
as hazard, table, 197 104
rules for, 85, 135 surfaces, 253
INDEX 325
tape for holding shields on module, required in SSW, 55, 94, 157,250
68 safety of, 95, 143
threads in smocks, 61 springs for safety, 143
trays, 101 two-conductor (resistive) type, 214,
Conductive tote boxes. See Tote 250; illus., 215; illus., 247
boxes, conductive versus periodic monitoring, 213,
Conductor 250
categorization of, table, 23 Contracting, 136
checking of grounding with Corona discharge, 9, 11
ZapOasb,247 Corrosion
definition of, 24 by workbench laminates, 63
discussion of, 30 of silicon wafers, 42, 52, 63
isolated, charging by ionizers, of solder by antistat containing
66-67 n-octanoic acid, 238-239,
when to allow in SSW, 53 242; table, 240
Connector dust cover, special test and special test for, in SSp, 49
standard test for, 59-60 Corrosivity in packaging, 110, 114,
Connectors 314
dust cover for, 59-60 Cost-effectiveness of ESD-control
electronic box, packaging of, 70 program, 134-135, 153
power turned off before inserting Cots. See Finger cots
ESDS item, 86 Coupon
shunt for, plastic, 39-40, 62, 104 categorization of, table, 23
Contact charging, 10 definition of, 24
Containers for storage or in-plant discussion of, 30, 234
transfer, 112 for automated process, 72,
Contamination 233-237; illus., 236
by antistatic polyethylene, for testing SSP, 47
238-243 CRrs. See Cathode ray rubes
of packaging, 111, 114 Cryptocharges, 299-304, 312-313
Continuous wrist-strap monitor Current-carrying capability, 143, 306,
alert signals, 95 308; illus., 309; table, 227
as primary defense, table, 158 Curve-tracer, 146, 151, 182
band,95 Cut-off relative humidity, 292, 309
calibration of, 94
capacitance type, 213; illus., 214 Damage mechanisms
categorization of, table, 23 check list of, 148
cord,95 in automated process, 236-237
cost-effectiveness of, 216 relation to defenses and hazards,
criteria for selection, 94-95, 100 table, 157
definition of, 24 Decals, on certified operators' badges,
discussion of, 30,157,207-217, 106
250 Defenses
procurement requirements for, primary and secondary, table, 158
94-95 relation to hazards and damage
326 INDEX
Defenses (continued) Ius., 204; illus., 275; illus.,

mechanisms, table, 157 295; illus., 307; table, 203;
Desiccator cabinets table, 307
criteria for selection, 104 pulse attenuation measured by,
static charges in, 69 194-195; table, 195
Detergent (water scrub) test, illus., results of, table, 195; table, 203;
221 table, 205; table, 227; table,
DI 262; table, 275; table, 295;
allowable voltage on conductors for table, 318
12-volt-sensitive devices, Disposition of mishandled hardware,
161-162 88, 145-147
categorization of, table, 23 Documents, reference, check list of,
definition of, 24 149-150
discussion of, 9-10, 16, 30,209, Double-bagging
281; illus., 14 cost of, 174
from people as principal ESD haz- method, 110, 169
ard,244 Drag test, 40; illus., 222
grouding not always needed for Drain time, 52, 184-185, 231; table,
damage by, 68-69 230
in automated process, 236 Dual-in-line package (DIP), 8, 44,
Discharge 263,283,304
air,8-9 Dust cover for connector, special test
categorization of, table, 23 and standard test for, 59-60
defmition of, 24
discussion of, 31 E Field (see also Field)
guarding against, 155 categorization of, table, 23
high-voltage, protection from, definition of, 24
294-298 discussion of, 16, 31; illus., 15;
rate increased by conductive sur- table, 157
face, 29, 285 Electron, 3-10,45, 105-106; 160;
rate slowed by antistatic surface, illus., 5
31,264,277 Electronic box, packaging of, 70
relation to damage mechanisms and EMIIRFI shielding, 115, 201, 206
defenses, table, 157 Enforcement problem
spark, 12, 31, 196, 253, 262, 265, check list, 154
268,296 in ESD-control program, 138-139
Discharge test Engineering function, 137
categorization of, table, 23 Equipment, ESD-control
definition of, 24 and materials,
discussion of, 31, 110; illus., 204; approved, 95-106
table, 203 complementary, table, 197
in model packaging specification, criteria for selection, 96-106,
procedures for, 113-114 196-198
on tote boxes, 229-230; table, 227 what to buy, check list, 150
procedures for, 171; illus., 184; il- approved, as primary defense, table,
INDEX 327
158 FFB
elimination of unnecessary, 249 categorization of, table, 23
in future, 160 definition of, 25
in SSW, 54-55 discussion of, 33, 45, 256; ilIus.,
ESDS 15; iIlus., 45
categorization of, table, 23 in automated process, 236
definition of, 24 Field
discussion of, 32 categorization of, table, 23
ESDS item controlled within SSW, 50
categorization of, table, 23 definition of, 24
definition of, 24 discussion of, 4-9, 11, 31, 116, 162,
discussion of, 32 299, 313; ilIus., 6; ilIus., 7;
in future, 161-162 ilIus., 17
selection of packaging materials emission, 11
for, 169-174, 294-320 from lights and light fixtures, 92
Evaporation, charge formation by, 12 from transformers or electric
motors, 91
Facilities and Industrial Engineering, from various objects, MOSFE1S
136-137 damaged by, 187
Failure analysis, iIlus., 20 guarding against, 155
Faraday cage hazard at the work station,
categorization of, table, 23 188-194,247
definition of, 24-25, 109 penetration of walls by, 91
discussion of, 32-33, 48-49, 116, read by rIeld meter to determine
170, 249, 251 charge on operator, 208
foil laminate relation to damage mechanisms and
bags, table, 170; tables, 195; defenses, table, 157
table, 202; table, 205; table, strategy of keeping weak, 155
262; table, 275; table, 295; Field meter
table, 318 as noncontact voltmeter, 34, 191;
criteria for selection, 99-100 table, 191
in model packaging specifica- categorization of, table, 23
tion, 109, 129 checks with, to enforce CD Rule,
MIlrB-81705, 1YPe I as, 170, 56; table, 197
201; table, 318 cost, 140
part of conservative approach, criteria for, 102
156,249 definition of, 25
for SSP, 48-49 discussion of, 9-10, 12, 33-35,
tote boxes, 49, 228, 331 140,141
Faraday cup for monitoring of necessary non-
categorization of, table, 23 conductors, 84
definition of, 25 in future, 160
discussion of, 33, 114 measuring apparent charge with,
in triboelectric charging tests, 27,181-182
230; iIlus., 221 use of, check list, 152
328 INDEX
FlM Freezing, charge generation by, 10, 38

categorization of, table, 23 Future
definition of, 25 equipment, 160
discussion of, 4-7, 10, 256, 260, ESDS items, 161-162
278, 279, 281, 299; illus., 6; materials, 159-160
illus. 15; table, 283 properties of bags, 174, 319-320
in automated process, 236 standards, 118, 122, 161
in special test of connector dust
cover, 60 Gloss test, illus., 220
Finger cots, 45, 85-86, 100, 162, Gloves
285; table, 282 antistatic, 85-86, 100
Flame resistance of packaging, cotton, 85, 188
315-316 latex, 62-63, 85-86
Floor finish. See Static-limiting Door nonconductive for cleanrooms,
finish 62-63,85-86
Floor mats, conductive, 82, 102; vinyl,62
table, 139; table, 141 Grid, ionizing, 103
Floors Grit~blasting
as necessary nonconductors in discussion of, 11
SSW,84,255 of heat sink on module, 68
conductive, of module to remove conformal
hazard of, 91, 132, 143 coating from components, 68,
in model handling and assembly 193-194
specification, 91 Ground fault circuit interruptors
in SSW, 51 (GFCls), 82, 143-144
materials, criteria for, 102 Grounding
mats for. See Floor mats, as primary defense, table, 158
conductive categorization of, table, 23
tile or concrete, treatment of, 91 check, in SSW, 93-94
PM (Floating Model), illus., 15; table, cord, criteria for, 100
261 definition of, 25
Foam, antistatic. See Antistatic foam discussion of, 35, 162; illus., 17
Foam, conductive. See Conductive in SSW, 80-81
foam lug, for workbench tops, 35, 52,
Foil laminate bags. See Faraday cage, 81, 185; illus., 82
foil laminate of personnel, 81, 245, 250; illus. 82
Footwear resistor needed for, 52, 81; illus.,
hazard of, on conductive floors, 91, 82
132,143 versus DI damage, 68-69
special, needed for conductive
floors, 74-75,82,91 H fields. See Magnetic fields
with leather soles, low charging, Hair
224 arm, as ESD hazard, 56; table, 197
with vinyl soles, high charging, charge induced on, 8
224 head, control of charges on, 50, 57,
INDEX 329
85; table, 197 compound, 10,299

Handling and assembly, model speci- contact compound. See CCI
fication for, 77-106 discussion of, 4-8
Hazards, ESO Installation site, procedures, 88
check list of, 148-149 Insulator. See Nonconductor
of static charges and fields at the Ionization
work station, 181-200 as primary defense, table, 158
relation to defenses and damage categorization of, table, 23
mechanisms, 157 definition of, 25
HBM discussion of, 19, 36-38, 92, 144,
as principal ESO hazard, 244 146-147,224,250,251,302;
categorization of, table, 23 ilIus., 17; table, 197
definition of, 25 equipment, criteria for, 103
discussion of, 4-10, 15, 161, 256, for controlling charge on silicone
272-273; illus., 5; illus., 15 rubber probe, 72
Heat guns, 64-65 grid,103
Heel mark test, illus., 220 hazard of nuclear, 37, 144, 190
Heelstra~conductive, 100 in future, 160
Hidden charges. See Cryptocharges in grit-blasting, 68
High voltage, nature of, 8-10 in model handling and assembly
History of ESO control, 31 specification, 83
Housekeeping in SSW, 85, 90 pros and cons of, 38
Humidification room,
as backup, 155 applicability of, 97
as secondary defense, table, 158 caution recommended, 249
categorization of, table, 23 charge decay rates, illus., 193;
definition of, 25 tables, 192
discomfort from "mugginess," 224 criteria for equipment, 103
discussion of, 35-36, 147, 249, effect of humidity on, 191-192
250 testing of systems, 183, 191-193,
in model handling and assembly 198
specification, 83 when to use, 38, 53
when to use, 52-53, 92 Ionizers
Humidity AC,
control of, 83, 234 for controlling charge on silicone
effect on cellulosics, 255, 292; rubber probe, 72
illus., 288; ilIus., 293 old,67
effect on room ionization system, safe distance from, 66-67; table
191-192 190
relative, 116,277,286-293,302, air guns, electrical and nuclear,
308-309, iIlus., 278; ilIus., 63-64, 73
310; illus., 311; illus., 312 blowers, criteria for, 103
voltages at high and low, table, 36 certification of, 83, 88
charging of isolated conductors by,
Induction 66-67
charging, 10
330 INDEX
Ionizers (continued) 92
cost, 139, 141 Manufacturing function, commitment
electrical vs. nuclear, 37, 189-190; to ESD-control program
table, 190 required, 137
evaluated with MOSFET board, Marking, caution
189 in handling, 87
fanless, 67 of packaging, 111, 115
nozzles, criteria for, 103 Masking tape, conductive, 193, 198
ozone generation by, 38, 89 Materials, ESD-control
"piggyback" for grit-blasting, 68, and equipment,
193-194 approved, 95-105
precautions, 89 complementary, table, 197
pulsed DC, criteria for selection, 96-105,
criteria for, 103 196-198
safe distance from, 66-67; table, what to buy, check list, 150
190 approved, as primary defense, table,
shadowing effect, 89 158
space-charging by, 75, 275-276; in future, 159-160
illus., 276 in model packaging specification,
Items, definition of, 89 109-110
in SSW, 53-55
Labeling. See Marking, caution Materials, packaging. See Packaging
Labels, paper, allowed in SSW, 58 Mats, floor. See Floor mats
Laminate, foil. See Faraday cage, foil Mats, table. See Table mats
laminate Megohmmeter, 93,104,151; illus.,
Laminates, workbench. See Work- 311
benches, laminates Metallized (see-through) bags, limita-
Latent fallure tions of, 31, 48-49; illus.,
categorization of, table, 23 297; table, 205; table, 262;
definition of, 25 table 275; table, 295
discussed in video training tape, Metals, criteria for selection, 104
106 Microscopes
discussion of, 38-39, 49, 138, 147, in SSW, 54, 247
234,244,256; illus., 15 ungrounded, with rubber feet, 37
possible, of mishandled hardware, MIlrSTD-1686, replacement of,
146 122-129
Lights and light fixtures, ftelds from, Mishandled hardware, disposition of,
92 88, 145-147
Logistics, 136 MM
definition of, 25
Magnetic fields (H fields) discussion of, 8, 9, 16, 39,
controlled by soldering standard, 273-276; illus., 15
54 MOSFET
shielding from, by iron foil, 115 categorization of, table, 23
Maintenance of ESD-protected areas, damage tests, procedure, 182-183
INDEX 331
damaged by, in special test with UV light, 69

charged operator touching lead to role of grounding in test with,
bench tops, table, 185; table, 68-69
186 test procedure with, 182-183;
fields, illus., 15; illus., 189 illus., 182
faelds from ionizers, 190 Mottoes, check list of, 152
fields from various objects, 187 Myths
grounded operator above ARI'G, check list of, 153-154
246 discussion of, 134-135,251-257
grounded operator touching lead
to charged tote boxes, table, Neon bulb, 8, 270, 281, 283; illus.,
186; table, 187 270; illus., 306
unbalanced room ionization Nitrogen, ionization of, 73
system, table, 191 Nonconductive
damaged in CDM test on tote categorization of materials, table,
boxes, 231; table, 229 23
definition of, 25 definition of, 25, 110
discussion of, 39, 265 discussion of, 39
in coupons for automated process, plastics, as hazard in work station,
234; illus., 236 137,187,197-198
in demonstrations for training tote box, cost of, 142
personnel, 106 Nonconductor
in discharge test, 113 categorization of, table, 23
in packaging tests, illus., 171 definition of, 25
in shielding!discharge test for bag discussion of, 4, 8, 278
materials, 196 excluded from SSW, 85, 247
in shielding test for SSP, 48-49 exclusion from SSW as secondary
in special test defense, table, 158
for ARI'G, 246 fields from, damage to MOSFETs
for mishandled hardware, 146 by, 187; table 187
on air guns, 64 necessary, in model handling and
on connector dust cover, 59-60 assembly specification, 84, 91
on heat gun, 64-65 not charged by nitrogen in desicca-
on packaging of electronic box tor cabinet, 69
with connector, 70 not charged in test with UV light,
on pulsed DC ionizer, 66-67 69
on screwdriver handle, 61 Nylon, antistatic, 75, 87, 99, 277,
in various tests, 4-8, 12, 14-15, 317; illus., 278; illus., 287
44,252,258,260,264,266,
267,272,273,274,281-283, One-meter rule in DoD-STD-1686, 51
294,302 Operator disciplines
punch-through of gate oxide of, as primary defense, table, 158
173, 260; illus., 15 categorization of, table, 23
MOSFET board certain, as secondary defense, table,
in shipping test, 70-71 158
332 INDEX
Operator disciplines (continued) Permanence of the antistatic property

check list of, 151-152 of bags and tote boxes,
compatible with work-place para- 175-180,309-310; table 231;
phernalia, 135 table 318
cornerstone of cost-effective pro- Personnel Voltage 'lester
gram, 134, 156 categorization of, table, 23
definition of, 25 data in ARTG test, illus., 246
discussion of, 39, 138-139, 188, definition of, 25
194,196-197,199,245,250 discussion of, 39, 151, 161
in certification of personnel, 105 for walk test, 83, 248
in model handling and assembly in disposition of mishandled hard-
specification, ware,l46
inspection of, 88 Photoelectric effect, 11
list of, 84-86 Piezoelectric effect, 11-12, 15, 38
in SSW, 55 Plant Services, 137
regarding chairs, 58 Plated wire shipping container,
retraining needed, 86,245 analysis of, 71
Ozone, generation by ionizers, 38, 89 Polarization, 4
Polonium-210, 37, 64, 83, 144, 190
Packaging Polyester fabric, for trlboelectric
antistatic bags for, shelf life of, charging in special tests, 56,
176, 178; table, 179 61, 67, 69, 100, 101
biodegradable, 292, 315 Polyethylene, antistatic. See Anti·
configurations of, 71 static polyethylene
flame-resistant, 315-316 Polystyrene foam, 8-9
for delivery, 88 Probe tools, antistatic nylon, 75, 277
history of, 169-170 Program management, 136
materials in model specification for, Program organization and
109-110 implementation
materials, relative costs, table, 174 administration, 249
model specification for, 107-115 advice to small companies,
new developments in, 305-319 139-142
paper, 317-319 check list for, 152-155
plastic, 317, 319 cost-effectiveness, 134-135, 250
recyclability of, 315 discussion of, 130-144,249,250
selection of materials for, 169-174 enforcement problem in, 138-139
summary of test results on mate- objectives, 131-132
rials, table, 174; table, 227; plan of action, 136-138
table, 262; table, 275; table, role of program coordinator in,
296; table, 297; table, 318 142-143
Paper safety, 143-144
in packaging, 317-319 variables in, 156
labels and forms, allowed in SSW, weakened by myths, 134-145
58 what, when, and how, 132-133
Particle beams, 11 Proposals, 136
INDEX 333
Puncture resistance of bags, 314; by bag materials, table, 174

table, 174 categorization of, table, 23
Purchasing, 136 definition of, 25
of CKI's by screen, 59, 105, 141
Quality assurance provisions required for cable, 60
in ESD-control program, 137-138, test for. See Discharge test, pro-
143 cedures for
in model handling and assembly test on tote boxes, 229-230; table,
specification, 88-89 227
in model packaging specification, Shielding/discharge test
113-115 apparatus, mus., 184; mus., 204
QA representative notified when flaws in, 31
hardware mishandled, 145 test conditions, table, 203
test results, table, 205; table, 227
Radar Shipping container, test on, 71
shielding against, 206 (see uncon- Shipping test, 70-71
densed paper for data) Shoes. See Footwear
test with packaged MOSFE1S, Shunt, conductive
mus., 172 foam, 40, 104
Radiation effects, 121 plastic bar, 40,62, 104
Radioactive decay, II Shunting
Rails, requirements for, 111 antistatic plastic for, 104
Relative humidity. See Humidity categorization of, table, 23
Resistor conductive foam for, 40,104
for grounding, 52, 81; mus., 82 conductive plastic for, 40, 62, 104
simulated, mus., 236 definition of, 25
thin-film, 44, 267; table, 283 discussion of, 39-40
Roller test, 40; mus., 221 foam for, antistatic or conductive,
Room ionization. See Ionization, 39-40,104
room in model packaging specification,
111, 114
Safety lead-shorting devices for, 104
check list, 155 material, criteria for selection, 104
in ESD-control program, 143-144 to protect ESDS items in SSp, 47
of personnel by grounding, 81-82 Silicon wafers, corrosion of, 42, 52,
Screen for CKI's, 59, 105, 141 63
Screen, metal, 205-206; table, 202; Silicone rubber probe, triboelectric
table, 295 charging of, 72
Screwdriver handle as rleld hazard, Skin
61; table, 187 causing charging of nonconduc-
Scuff test, 40 tors, 34-35
Seat covers, conductive, 102 control of voltage on, 245-249,
Shelf life of antistatic packaging ma- 250; mus., 246
terials, 175-180; table, 179 voltage suppression by, 56, 57
Shielding Slip test, 151, 161; mus., 220
334 INDEX
Small companies, advice to, 139-142, for elecronic box as package, 70

154 for fanless ionizer, 67
Smocks, ESD-control for footwear, 74-75
antistatic, 100 for grit-blasting, 68
as basic requirements in future, 162 for heat guns, 64-65
conductive threads in, 61 for ionizing nitrogen, 73
danger of stainless-steel fibers in, for materials and equipment, 97
62,194 for old AC ionizers, 67
evaluation of three designs, 61-62 for plain polyethlene bag as pack-
in SSW, 50-51, 135, 207 age, 70-71
pros and cons of, 57 for screwdriver handle, 61
Soldering irons for selecting materials, 53
magnetic fields from, 54 for shunt bar, 62
voltage on tips, 54 for smocks, 61-62
Solvent extraction test, 309-310; for space-charging, 75
table, 313 for static charging, illus, 177; illus.,
Space-charging. See Ionizers, space- 221; illus., 222
charging by for tribolelectric cbarging by air,
Special test 66
accelerated aging of bags, illus., for UV light, 69
179 for workbench laminates, 63
categorization of, table, 23 for wrist-strap bands, 63
corrosion and contamination test in disposition of mishandled hard-
for SSp, 49, 238-239 ware, 146
discbarge test, illus., 171; illus., Specifications
204 model, handling and assembly,
discussion of, 40, 118 77-106, 124, 126-127
for air guns, 63-64 model, packaging, 107-115, 125,
for automated process, 72, 128-129
234-237 referenced in model handling and
for brushes, 60-61 assembly specification, 97
for chain link fence, 69-70 referenced in model packaging
for charges on specification, 109
arm hair, 56 Spraying of liquids, charge generation
clothing, 57 by, 11
CKfs, 59, 105 SSP
head hair, 57 categorization of, table, 23
identification badges, 57-58 corrosion, special test for, 49
walls, 58-59 definition of, 25
for c1eanroom gloves, 62-63 discussion of, 40, 47-49, 116, 117
for connector dust cover, 59-60 in model packaging specification,
for cordless wrist strap, 72 108-109
for DC ionizers, 66-67 requirements, check list of, 149
for desiccator cabinet, 69 SSW
for DI and grounding, 68-69 categorization of, table, 23
INDEX 335
certification of, 51, 80, 88, 93-94 specification, 78, 97-98

cleanliness requirement, 85, 90 in model packaging specification,
decal for certification of, 93, 94 108
defmition of, 26, 80, 93 Static charge
designation of, 90 apparatus for producing, illus., 117;
discussion of, 40, 50-55, 117 ilIus., 221; illus., 222
for highly sensitive devices in fu- categorization of, 23
ture, 162 definition of, 26
housekeeping in, 85, 90 discussion of, 40-41
identification of, 93, 94 hazard at the work station, 184-188
in model handling and assembly Static-dissipative
specification, 80-87 categorization of materials, table,
inspection of, 88 23
minimal, check list for, 149 definition of, 26, 83, 110
operator skills vital in, 106, 245 desoldering handpiece, 96
work surface in, 90, 184-185, 198 discussion of, 41
SSZ table mat, 102
categorization of, table, 23 tote box, 101
definition of, 26 workbench tops, advantage of,
discussion of, 40 184-185,268,285; illus., 282
in model handling and assembly Static Event Detector (SED), 12, 44,
specification, 80 272, 281, 283; illus., 273;
in model packaging specification, illus., 284; illus., 306; illus.,
109 307
maintained in SSW by operator, Static rleld (see also rreld)
50-51,55,84-85 categorization of, table, 23
maintained inside SSP, 47 definition of, 26
Stainless-steel fibers, hazard of discussion of, 41
in smocks, 61-62, 194 Static-OmitinR noor finish
in wrist-strap bands, 63, 194 as "safety net" or backup, 52-53,
Standard test 155,225
categorization of, table, 23 as secondary defense, table, 158
definition of, 26 categorization of, table, 23
discussion of, 40 cost, 140; table, 223
for brushes, 60 cost-effectiveness of, 219, 221, 224
for connector dust cover, 59 criteria for, 102
for floor finish, ilIus., 200 definition of, 26
for packaging materials, table, 174 discussion of, 18, 21, 41, 73,
for paper forms and labels, 58 218-225,248,273; illus., 17;
for various packaging configura- table, 158
tions, 71 effectiveness after wear, 250
Standards in maintenance of ESD-protected
industry, 118-121, 123 areas, 92
in future, 161 in model handling and assembly
in model handling and assembly specification, 84
338 INDEX
Static-limiting (continued) in SSW, Tesla coil test, 228; illus., 171; illus.,
52-53 204; table, 203; table, 205
test methods for, mus., 220; mus., Test equipment
221; mus., 222; illus., 223 check list of, 150-151
Statistical process control, 123 criteria for selection, 104
Steel, painted, 84, 263; mus., 270 Testing of ESDS items
Steel work surfaces, 91, 93 precautions during, 86-87
Stresscracking of polycarbonate by procedures, inspection of, 88
antistats, 238, 239, 242 Test, multiple-choice, for certifying
Summary of book, 158 personnel, 115-117
Surface resistivity Thermionic emission, 11
categorization of, table, 23 1bpical antistat
definition of, 26 applied to plastic parts of equip-
discussion of, 41-42 ment, 54, 96
effect of relative humidity and tem- applied to screwdriver handle, 54,
perature on, 277, 286-293, 61
310, 317-319; illus., 311; categorization of, table, 23
illus., 312 chloride-free, 42, 104
measurement of, for packaging, 113 cost, 110
meter, for static-limiting Door fIn- criteria for selection, 104
ish, 83,150 definition of, 26
meters in future, 160 discussion of, 42; table, 197
method of measuring, 182; illus., for treating
178 antistatic tote boxes, 179
of brush handles, 60 identification badges, 58, 84
of smocks, 61 necessary nonconductors in
requirements for packaging, no SSW, 84, 91, 198
versus time for aged antistatic walls or windows, 58-59
bags, illus., 179 possible contamination effects of,
Symbols, 112 238-243
Tote boxes
Taber abraser, 311; illus., 314 antistatic, need for antistat treat-
Table mats, 101, 117, 263; illus., 270; ment of, 179
table, 282 cardboard, 142,226-229, 285;
Tape table, 231
antistatic, 74, 255-256 conductive,
charging by stretching, 12 cost of, 142; table, 231
conductive, for holding shields on disadvantages of, 180, 185-186,
module, 68 226
conductive, masking, 193, 198 Corshield, 227-228; table, 227;
Techniques vs. materials and equip- table, 231
ment, table, 197 cost of, 142; table, 231
Teflon criteria for, 101, 185-187
charged by heat gun, 65 current carried by, 227
in triboelectric series, 43, 279 discussion of, 4, 226-232
INDEX 337
drain time of, 231; table, 230 resistivity for static-limiting

Faraday-cage construction, 49, Door finisb, 41
228-229 limiting of, in packaging, 110
hazard of static cbarges on, not caused by pure gases, 12,59,
179-180; table, 186; table, 228 66
lid used in special test, 66 of operator at work station, 248,
materials, comparison of, table, 231 250; illus., 246
new materials for, 160,226-229 polyester fabric for, 56, 61, 67, 69,
sbielding/discbarge test on, table, 100,101
227 reduced by increased relative
static-dissipative, 101 humidity, 36, 83
techniques for using, table, 197 test, in model packaging
triboelectric cbarging of, 228 specification, 114
vulcanized fiber, 232; illus., 227 tests for liner of SSp, 47
Toxicity of packaging, 314 tests for packaging, 278, 311-312;
TQM (Total Quality Management), table, 279; table, 316
19,123; illus., 20 Triboelectric series
Training of operators, 55 discussion of, 3
Transient personnel in SSWs, 86 example of, table, 42-43
Transparency of packaging materials, limitations of, 43
313 Troubleshooting
Trays, criteria for, 101 illustrative examples of, 56-76
liiboelectric cbarging major tools for, check list, 151
Aclar for, 27, 43, 58, 65, 69, 245 1\veezers, 73-74
avoidance of by preventing
packaged item from sliding, Ultraviolet (UV) light for inspecting
71, 110 ESDS items, 69
bag shaker test for, 43 Unglazed cardboard
by air, 12, 59, 66 for packaging, 111
by body movement, 85, 248 for tote boxes, 83
by conductive packaging materials, Units, defmition of, 89
178, 226, 279; table, 228; Unrealistic testing, example with
table, 279; table, 316 example with plastic, 31
by grit-blasting, 68, 193-194 example with wire, 53-54
by spraying with conformal
coating, 188 Vehicular bounce test for packaging,
by stainless steel, 66, 69 illus., 172
by tote box materials, table, 228 Vinyl, antistatic, 75-76, 315-316
categorization of, table, 23 Voltage suppression
definition of, 26 by conductive masking tape, 198
discussion of, 42-43, 278; ilIus., by skin, 56-57
17 categorization of, table, 23
in blow-off operations, 38 definition of, 26
in operator training, 105-106 discussion of, 43
lack of correlation with surface hiding field on tote box, 52
338 INDEX
Voltage suppression (continued) criteria for, 101

in brushes, 60-61 testing of, 184-185, 263,
in cleanroom gloves, 62-63 269-271, 281-285
in packaging, 110, 299 surface,
of charged nonconductive plastics corrosivityof, 52
on workbench, 187 drain time of, 52, 184-185, 231,
of necessary nonconductors 269, 277, 285; table, 230
(painted or plastic-coated metal (stainless steel or aluminum),
grounded metals) in SSW, 84 91,93
Vulcanized fiber for tote boxes, 233; with buried conductive layer, 93
table, 227 Wrist straps
V-Za~35, 132, 137, 153 bands, metal fibers in, 63, 194
bead-chain, 194, 198; table, 212
Wafers, silicon carbon-loaded plastic as question-
corrosion by chloride ion, 42, 52, able material, 194
63 cloth band, looseness and soiling
ESD testing of, 258-262; mus., of,212
259; illus., 264; mus., 303; cordless, 72, 272
table, 275 discussion of, 251
Walk test expansion band, table, 212
apparatus for, mus., 222 factor in determining resistance to
apparent vs. actual voltage, 219; ground of operator, 212
illus., 223 increased AKfG by loose bands,
applied to connector dust cover, testing and monitoring of, 216, 250;
59-60 table, 212
categorization of, table, 23 various designs, 194
definition of, 26
discussion of, 43, 73, 74 X-rays, charging of particles by, 11
Walls
charged, field hazard, 58-59 Zapflasb
in SSW, 51 as basic equipment for operators,
penetrated by fields, 91 245
precautions, 91 brightness of light vs. resistance,
Water analogy, 4-8 211
Windows, charged, 59, 91, 197 categorization of, table, 23
Woolen fabric, 34, 50 cost, 140, 141
Workbenches criteria for selection, 104
conductive surfaces, 91,93 definition of, 26
damage to MOSFE'lS by touching discussion of, 43,54,55,93
surface of, table, 185 for checking grounding of conduc-
design and grounding of, 51-52, tors, 247
93-94,184-185 for locating AC power leakage, 52
laminates, various uses of, 43
chloride ion in, 63

John M. Kolyer, Donald E. Watson (Auth.) - ESD From A To Z - Electrostatic Discharge Control For Electronics-Springer US (1996)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

John M. Kolyer, Donald E. Watson (Auth.) - ESD From A To Z - Electrostatic Discharge Control For Electronics-Springer US (1996)

Uploaded by

Copyright:

Available Formats

ESD

Distributors for all other countries:

tt Electronic Services <http://www.wkap.nl>

Library of Congress Cataloging-In-Publication

A C.I.P. Catalogue record for this book is available

ISBN-13: 978-1-4612-8499-4 e-ISBN-13: 978-1-4613-1177-5

Printed on acid-free paper.

Preface To Revised Edition v

of the antistatic property of antistatic materials. These publications are referred

How to Use This Book

triboelectric charging by body movement are secondary precautions. However, all

Purpose of This Chapter

Static Electricity and Electrons

The Water Analogy

Particle Beams. Particles can be charged by bombarding them with other

X-Rays. Particles also can be charged by bombardment with X-rays, which

Grit-Blasting. This instance of charging by relatively large particles is only a

Spray. By the "balloelectric effect," atomizing a liquid by spraying gives a

Photoelectric Effect. Photons ("particles" of light) may strike a surface to

Deformation and Cleavage (Piezoelectric Effect). Piezoelectricity was dis-

charges because of reorientation of polar molecules (when crystal symmetry is

Evaporation. "Cleaning or drying fluid by evaporation" is listed as a "typical

"fluid" lies in "stagnant pools" (negative charge) or is conspicuously lacking in

Fundamentals of ESD Control

m OXIDE LAYER w,.PUNCTURE IN z

Fig. 3-1. Threats.

The screwdriver-to-IC-Iead ESD mentioned above is an example of the MM

::~lIl -orA~ DISSIPATIVE

Fig. 3-2. Defenses.

safety comes ftrst in any operation. Alternate methods of grounding personnel,

Fig. 3-3. Documentation/Organization.

precepts, operator experience. A "living list" of approved ESO-control materials

Real and Conceptual Tools from A to Z

Table 4-1. Categorization of Tools.

REAL (AN OBJECT

Primarily pertinent to fields Antistatic (materials) Apparent charge

Primarily pertinent to dis- Antistatic lotion ARTG

Equally pertinent to both Coupon Analysis

Real and Conceptual Tools: Definitions

Apparent Charge: The charge in volts of a surface or object as read by a

SSW: Static-safe workstation. This is a volume in space which is designated

Real and Conceptual Tools: Discussion

Antistatic: DoD-HDBK-263 defined "antistatic" as having a surface resis-

demonstrate by wiping a hand over the surface of an uncharged plastic sheet.

ACTIVITY 20%RH 80%RH

Walking across vinyl floor 12kV 250 V

However, the fact remains that triboelectric charging becomes especially

In contrast to humidification, which reduces triboelectric charges by 'nipping them in

Electrical ionizers have the possible problem of charging isolated conductors

that antistatic or static-dissipative shunt materials, e.g., foams, are insufficient,

The Charged Device Model (CDM)

Fig. 5-1. Field-from-board (FFB) mechanism.

The Static-Safe Package (SSP)

The Static-Safe Workstation (SSW)

How the SSW Holds the SSZ

Workbenches: Design and Grounding

grounded metal" is one of the "necessary nonconductors" allowed in paragraph

Continuous wrist-strap monitors (see Papers No.5 and 10 in the Appendix)

General Design, Construction,

Troubleshooting, by which we mean the solving of unique problems as they arise,

6. Paper Forms and Labels

9. Cathode Ray Tubes (CRTs)

10. Connector Dust Cover