E 480: Reliability and Failure Analysis: Presented By: Nur Hamidah Abdul Halim

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 49

EMT 480 : RELIABILITY

AND FAILURE ANALYSIS

Presented by: Nur Hamidah Abdul Halim

Lectures edited from:


Dr Ramzan Mat Ayub
Dr Hasnizah Aris
Dr Noraini Othman
COURSE OUTCOMES
CO1:
• Ability to analyse and evaluate the reliability
based problems
CO2:
• Ability to explain types of semiconductor failures
and select the most relevant techniques and
tools to perform the Failure Analysis.
CO3:
• Ability to recommend and apply appropriate
techniques and tools available in Failure Analysis.
Course Synopsis
In the first section of Reliability, students will learn the
concept of Reliability, its terms & definitions, the different
types of Reliability Distributions and also the different types
of Reliability Prediction Techniques such as Failure Modes
and Effects Analysis (FMEA) & Fault Tree Analysis (FTA).

In the second section of Failure Analysis, students will be


exposed to the different types of FA techniques commonly
conducted on a failed semiconductor device and the test
instrumentation associated with each technique.
Teaching Plan
Week Topics
OVERVIEW OF RELIABILITY
Appraise and understand reliability, probability, failure rate
1
and other terms on reliability for application in electronics
system
RELIABILITY PREDICTION TECHNIQUES &
REPORTING (FMEA and FTA)
2
Ability to choose and select the best problem solving tools
for failure analysis
References
} 1. Patrick O’Connor (2002). Practical Reliability Engineering, Wiley

} 2. Ebeling, C. E. (1997). Reliability and Maintainability Engineering,


McGraw Hill

} 3. Lawrence C. Wagner, (1999). Failure Analysis of Integrated Circuits:


Tools and Techniques.: Kluwer Academic Publishers.

} 4. Perry L. Martin (1999). Electronic Failure Analysis Handbook.: McGraw


Hill

} 5. E. Ajith Amerasekera and Farid N. Najm (1997). Failure Mechanisms in


Semiconductor Devices. 2nd Ed.: John Wiley & Sons

} 6. Friedrich Beck (1998). Integrated Circuit Failure Analysis: A Guide to


Preparation Techniques. : John Wiley & Sons
Lecture 1: Overview of Reliability
(for electronic systems)
What is Electronic ?
The science of controlling electrical energy electrically, in which the electrons
have a fundamental role.
Where the electrical energy come from?

Electronics deals with electrical circuits that involve active electrical components
such as vacuum tubes, transistors, diodes, integrated circuits, associated passive
electrical components, and interconnection technologies.
Vacuum Tubes – one of electronic components that built early computer
Forefather of today's all-electronic digital computers named ENIAC (Electronic Numerical Integrator
and Calculator). Was built at the University of Pennsylvania between 1943 and 1945. ENIAC filled a
20 by 40 foot room, weighed 30 tons, and used more than 18,000 vacuum tubes.
What is Microelectronics ?
Microelectronics is a subfield of electronics. As the name suggests, microelectronics
relates to the study and manufacture (or microfabrication) of very small electronic
designs and components. Usually, but not always, this means micrometre-scale or
smaller.
These devices are typically made from semiconductor materials.

Many components of normal electronic design are available in a microelectronic


equivalent. These include transistors, capacitors, inductors, resistors, diodes and
insulators and conductors can all be found in microelectronic devices

They are designed and fabricated by monolithic approach – Monolithic Integrated


Circuits (thus the cheapest and most popular approach).
Business Model Of Semiconductor
Industry
Foundry (TSMC, UMC, Silterra, Infineon, 1st Silicon)
– only manufacture

Design House (Alterra, MyMS)


- only design

Integrated Design Manufacturing (Intel, Motorola, IBM, MIMOS)


- design and manufacture
Semiconductor Manufacturing Processes
• Design (Digital, analog, full custom, semi custom (gate array) etc)
– Tape out (design to Mask Shop)
• Mask making
• Individual process development • Backend Process
• Front-end Processes (individual transistor) • Deposition (oxide, nitride etc)
• Metalization
– Deposition • Rapid Thermal Process
– Oxidation • Lithography & Etch
– Diffusion •
– Photolithography
– Etch (wet and dry) • Test (Parametric and Functional)
– Implantation • Packaging
Review on Semiconductor Material
Solid Materials Classification Insulator
(based on electrical conductivity) Semiconductor 1. Band gap evergy
Metal
The ability to control material’s conductivity is the key to modern electronic devices.

Simple cubic 2. Energy states density


Lattice of Unit Cell Body centered 3. Carriers mobility
Face centered 4. Phonon scattering
Diamond and Zincblende 5. Traps density (defect)
Si density = 5 × 10!! 𝑎𝑡𝑜𝑚𝑠/𝑐𝑚"

Unit Cell Arrangement Amorphous


Poly crystalline 111
(based on atomic arrangement) Single crystal Crystal Orientation 100
Miller Indices
Intrinsic Semiconductor Electrons / holes
Semiconductor Energy Bands concentration
Extrinsic Semiconductor

ni = pi and

nipi = ni2 (MASS ACTION LAW) – the product of n p is always a constant for a
given semiconductor material at given temperature

n-type Semiconductor Typical doping 1016 – 1019 / cm3


Doping 5 × 10!! 𝑎𝑡𝑜𝑚𝑠/𝑐𝑚"
p-type Semiconductor 1 per 50,000 atoms Si
Elementary semiconductor Silicon
Germanium

II-IV SiGe
Compound Semiconductor Electron mobility
III-V GaAs, InP etc
Optical properties
II-VI CdS, ZnTe etc

Offer high performance devices (optical characteristics, higher frequency and power)
Binary SiGe, GaAs
Ternary AlxGa1-xAs
Quaternary InxGa1-xAsyP1-y
Review on Semiconductor Process Technology
Design rules
IC Logic design
µP, Flash Analog design
Device model
Bipolar Tech Layout
Technology Library CMOS Tech
of certain foundaries Mask
BiCMOS Tech
Inverter
Other logic block
Memory (volatile, nonvolatile)
P/N junction Capacitor CMOS
Diode NMOS BJT Device Physics
Resistor PMOS Memory Cell
CMOS Process Technology (integration)
Well Technology Gate Technology Multi level Inter
Isolation Technology Contact Technology Connect Tech Process Tech

Cleaning Ion implantation Photolithography


Oxidation PVD CVD Deposition

Solid material classification Periodic arrangement Intrinsic and extrinsic Material


Lattice Crystallic orientation semiconductor
Unit cell Energy band & diagram Material doping
• Reliability
TERMS AND • Failure Rate
• Mean Time Between Failure (MTBF)
DEFINITION • Mean Time to Repair (MTTR)

PROBABILITY •

Exponential Distribution
Binomial Distribution
DISTRIBUTION • Poisson Distribution
• Weibull Distribution
FUNCTION
• Serial Reliability Model
• Parallel Reliability Model
SYSTEM MODEL • k of n Reliability Model
• Combination Reliability Model
1.1 TERMS & DEFINITION :

• “Reliability is defined as probability of a device


performing its purpose adequately for a period of
time intended under the specified condition”

• “Reliability refers to the probability that a


component or a system will perform a required
function for a given period of time when used
under stated operating condition”
Operating temperature
SUBJECT + OUTPUT+ TIME + CONDITION
“iPhone found to be 300% more reliable than
Samsung smartphones”
How does the score been calculated ?

IMPRESSION RATIO

𝐌𝐚𝐧𝐮𝐟𝐚𝐜𝐭𝐮𝐫𝐞𝐫 𝐩𝐫𝐨𝐛𝐥𝐞𝐦 𝐢𝐦𝐩𝐫𝐞𝐬𝐬𝐢𝐨𝐧𝐬


𝐒𝐚𝐦𝐩𝐥𝐞 𝐩𝐫𝐨𝐛𝐥𝐞𝐦 𝐢𝐦𝐩𝐫𝐞𝐬𝐬𝐢𝐨𝐧𝐬
1.2 TERMS & DEFINITION :
1.2 TERMS & DEFINITION :

MEAN TIME TO REPAIR (MTTR)


} MTTR is the mean (average) time between failures and being
returned to service.
!"#$% &"'(#)*+
MTTR=
," "- -$)%./+0

AVAILABILITY
} Availability of a module is the percentage of time when
system is operated.

12#)*+ 5!67
Availability= =
3"'(#)*+412#)*+ (5!!945!67)
EXAMPLE 1
Evaluate the Failure Rate and MTBF for washing machine that
has accumulated 5 failures that resulted in five service calls
during 1,200 hours of operation.

SOLUTION

FR(λ) = 5 Failures/ 1,200 hours


= 4.167 x 10-3 failures/hour
= 4167 x 10-6 failures/hour
MTBF = 1/ λ = 1/(0.00417) = 240 hours

Conclusion : The machine is expected to break down after 240


hours of usage, and this information is crucially needed to establish
spare part requirement
2.1 Exponential Distribution
When a device is subjected only to failures that occurs at discrete interval.

Failure rate (λ) is expected to be the SAME or CONSTANT over long


operating period.

The failure density function


𝒇 𝒕 = 𝝀𝒆!𝝀𝒕

For t >0, where λ is the failure rate;


The Reliability function is
– Reliability 𝑹(𝒕)= 𝒆−𝝀𝒕
– Unreliability (failure) , 𝑸(𝒕)= 𝟏 −𝑹(𝒕) = 𝟏 − 𝒆−𝝀𝒕

Where
λ : failure rate
t : operating time
EXAMPLE 2
Evaluate the reliability of an electrical device that has an
exponential reliability distribution with a failure rate of 9 x 10-6
over time t = 1,000 hours.

SOLUTION

Reliability or functional
𝑹(𝒕)= 𝒆−𝝀𝒕 = 𝒆−(𝟗 𝐱 10−6𝐱 𝟏𝟎𝟎𝟎)=𝟎.𝟗𝟗𝟏𝟎𝟒

Unreliability or failure
𝑸(𝒕)= 𝟏 −𝑹(𝒕)=𝟏 − 𝒆−𝝀𝒕
= 1 – 0.99104 = 0.000896
2.2 Binomial Distribution
• Binomial Reliability Distribution is a discrete distribution, where there are only
two possible outcomes such as success or failure.

• (𝑹+𝑸)𝒏=𝟏 , n : total number of trial

• Example : flipping coins , n = 2 ( head or tail) (𝑹+𝑸)𝟐=𝑹𝟐+


𝑸𝟐+𝟐𝑹𝑸=(𝟎.𝟓+𝟎.𝟓)𝟐=𝟏

• The probability density function is given by;

𝒇 𝒙 = 𝒏 … . . 𝒙 𝒑𝒙 𝒒(𝒏$𝒙) where

𝒏!
𝒏….𝒙 =
𝒏 − 𝒙 ! 𝒙!

• (+) : suitable for parallel system


2.3 Poisson Distribution
• Poisson Reliability Distribution is a discrete distribution by taking
consideration of time interval / time domain.

• The extension of binomial function for the large n (n >20) i.e. large system.

• If the failures are Poisson distributed, they occur at a constant average rate
and the number of events occurring in any time interval are independent
(parallel system) of the number of events occurring in any other time
interval.
• The number of failures in a given time would be given by;

𝒂𝒙𝒆!𝒒
𝒇 𝒙 =
𝒙!
• Where x is the number of failure and a is the expected number of failures.

• Simplified form for reliability analysis

(𝝀𝒕)𝒙 𝒆"𝝀𝒕
𝒇 𝒙; 𝝀, 𝒕 = 𝒙!
where λ is the failure rate, t length of time under
consideration and x is the number of failure.

The probability of zero failures in time t which is the Reliability function is


given by;

(𝝀𝒕)𝟎 𝒆"𝝀𝒕
𝑹 𝒕 = = 𝒆!𝝀𝒕 which is the exponential distribution.
𝟎!
For parallel system, R(t) can be expressed in terms of probability of number of
failure in time t.

𝒓
(𝝀𝒕)𝒙𝒆!𝝀𝒕
𝑹 𝒕 = H
𝒙!
𝒙+𝟎
(𝑹+𝑸)𝒏=𝟏

𝑹=𝟏−(𝑸𝟏+𝑸𝟐+ 𝑸𝟑+ …)

• Qi is the probability of exactly i failures occurring during time period, t.


• Example :
• Probability of exactly 1 FAILURE during time period, t :
𝑸𝟏(𝒕)=(𝝀𝒕)𝒆−𝝀𝒕
• Probability of exactly 2 FAILURES during time period, t :
𝑸𝟐(𝒕)=(𝝀𝒕)𝟐𝒆−𝝀𝒕/𝟐!
EXAMPLE 3
What is the probability of :
i. One failure
ii. Two failures
iii. Three or more failures occurs during a time period of 1000 hours for an
electronic device with failure rate of 100 x 10-6
SOLUTION
Given t = 1000 hours and λ = 100 x 10-6
} i. Probability of exactly ONE failure
𝑸𝟏(𝒕)=𝝀𝒕𝒆−𝝀𝒕
=(100 x 10−6)(1000)e−(100 x 10−6)(1000)
= 0.0905

} ii. Probability of exactly TWO failures


𝑸𝟐(𝒕)=𝝀𝒕𝟐𝒆−𝝀𝒕/𝟐!
−𝟔
=(𝟏𝟎𝟎×𝟏𝟎−𝟔×𝟏𝟎𝟎𝟎)𝟐𝒆−(𝟏𝟎𝟎×𝟏𝟎 ×𝟏𝟎𝟎𝟎)/𝟐!
=𝟎.𝟎𝟎𝟒𝟓
Probability of THREE or MORE failures
𝑹 = 𝟏−(𝑸𝟏+𝑸𝟐+ 𝑸𝟑 …)
𝑸3.. = 𝟏−𝑹−𝑸𝟏−𝑸𝟐
= 𝟏−𝒆−(𝟏𝟎𝟎×𝟏𝟎−𝟔×𝟏𝟎𝟎𝟎) -0.09048-𝟎.𝟎𝟎𝟒𝟓𝟐
= 1.625 X 10-4
2.4 Weibull Distribution
} Complex continuous random reliability distributions, failure rates are not
CONSTANT all the time.

} Weibull reliability distribution is similar to exponential reliability function with


additional shaping parameter.
𝜷
𝑹𝒕= 𝒆−(𝜶𝒕)

} where as α : failure rate


β : shape parameter (it indicates whether the FR
decrease/increase)

} If β < 1.0 , failure rate is decreasing


β = 1.0 , failure rate is constant
β > 1.0 , failure rate is increasing
Serial Parallel

SYSTEM MODEL
(assessing multiple
components)

K of n configuration Combination
3.1 SERIAL RELIABILITY MODEL

• Simplest reliability model


• EACH COMPONENT of the system NEED to be
WORKING for overall system success

• Reliability : 𝑅𝑠= 𝑅1 × 𝑅2 × 𝑅3 × 𝑅4 × …….× 𝑅𝑛 =


(1−𝑄1)(1−𝑄2)(1−𝑄3)(1−𝑄4)….(1−𝑄𝑛)
• Failure rate : 𝜆𝑠= 𝜆1+ 𝜆2+ 𝜆3+ 𝜆4+⋯+𝜆𝑛 ∴MTBF=1/𝜆𝑠
3.2 PARALLEL RELIABILITY MODEL
3.3 κ of n CONFIGURATION RELIABILITY
MODEL
3.4 COMBINED RELIABILITY MODEL
• Assume that a particular subsystem is made
up of three independent and identical units in
parallel and at least two units must operate
normally for the subsystem success. Each
unit’s probability of success is 0.8.
• (i) Calculate the subsystem reliability.
• (ii) Give your interpretation of the result
obtained.

You might also like