I I
IEEE TRANSACTIONS ON EDUCATION, VOL. 33. NO I . FEBRUARY 1990
The Teaching of Electromagnetics
JOHN R . WHINNERY, FELLOW,IEEE
(Invited Paper)
Abstmct-Maxwell’s formulation of electromagnetic theory is one of
the most profound and beautiful of physical theories, and is the basis
of all electrical engineering. This paper discusses the history of its
teaching in universities, some questions of content and order, useful
analogies and teaching aids to clarify the concepts, and the growing
importance of computational approaches to the subject.
INTRODUCTION
R ICHARD FEYNMAN, in his brilliant book, Lectures
on Physics [l], introduces his treatment of electro-
magnetic radiation with the following statement:
“Perhaps the most dramatic moment in the develop-
ment of physics during the 19th century occurred to J. C.
Maxwell one day in the 1860’s when he combined the
laws of electricity and magnetism with the laws of the
behavior of light . . . . Maxwell could say when he fin-
ished with his discovery, Let there be electricity and mag-
netism and there is light!”
Certainly Maxwell’s achievement is one of the most
beautiful and yet profound of the formulations of classical
physics. The set of four equations we know as “Max-
well’s equations,” in modem notation [2], is simple
enough to imprint on a T-shirt, and yet rich enough to
provide new insights throughout a lifetime of study. Some
students grasp the clarity, power, and excitement on a first
introduction while others have a very rough time with the
concepts. This special issue is dedicated to helping more
and more of the students move into the former class.
HISTORY
Maxwell built upon centuries of experimentation and
codification of electric and magnetic interactions, and
upon theories of and experiments with light. His achieve-
ment, at about the same time as the U . S . Civil War, seems
like ancient history to most of us. Certainly this author
thought of it as such upon a first introduction a half-cen-
tury ago, and yet only a half century before that Hertz had
carried out his incisive experiments, verifying Maxwell’s
predictions. Another surprise to most of us was in finding
the breadth and depth of analysis of waveguiding, reso-
nance, scattering, and diffraction by Lord Rayleigh,
Oliver Heaviside, Arnold Sommerfeld, J. J. Thomson,
and others [ 3 ] , [4] in the late 1800’s. Many applications
Manuscript received February 3, 1989.
The author is with the Department of Electrical Engineering and Com-
puter Sciences, University of California, Berkeley, C A , 94720.
IEEE Log Number 8929914.
0018-9359/90/0200-0003$01.OO 0 1990 IEEE
3
of these analyses had to wait for the microwave sources
of World War 11, and in some cases the original analyses
were redone without any realization that they were al-
ready in the literature [ 5 ] .
Another surprise to those who have grown up since
World War 11: Maxwell’s equations, the basis of all elec-
trical engineering, were seldom, if ever, mentioned in a
typical undergraduate EE curriculum of the 1930’s. The
texts and courses stressed the derived concepts of circuit
analysis-certainly a very powerful tool but one that had
to be strained to explain the behavior of such things as
rotating machines or antennas [6]. All of this changed with
the development of microwaves during the war, and the
necessity to return to the fundamental base. Physicists and
the relatively few electrical engineers with doctorates had
to provide the analyses for the waveguides, cavity reso-
nators, aperture radiators, and other elements of this new
regime. The effect upon the post-war curriculum was im-
mediate, with courses on field and wave concepts at both
the undergraduate and graduate levels available in nearly
all curricula and required in most.
The exciting characteristic of electrical engineering has
been its rapid change, and the ability of our profession to
adapt to this change. During the last few decades there
have been many new specialties developed and absorbed
by the profession-some of these several stages removed
from the charged particles and their fields of influence that
enter into all electrical phenomena. With this expansion
and diversity, the need for electromagnetics is less clear
to some of the specializations, and there is a less uniform
pattern of requirement of field courses. More will be said
about this later but for the present we only point out the
lesson of past history: there will be more change, and with
this change a continuing need to return to fundamental
laws rather than to derived concepts.
ORDERA N D CONTENT
In discussing the question of content and order, it is
important to say at the outset that there is no unique so-
lution as to the optimum. Different texts and different in-
structors with the same text can give successful courses
using differing selection and ordering of material. The
typical pattern for an undergraduate course is that of pro-
gression from static field concepts to the dynamic laws,
followed by application to such things as waveguiding and
resonant systems, radiating systems, and optoelectric ele-
ments. Review or introduction of the mathematical tools
 4 IEEE TRANSACTIONS ON EDUCATION, VOL. 33, NO. I . FEBRUARY 1990
such as vector analysis, orthogonal functions, and Fourier
analysis may be introduced where needed, given as one
or more introductory chapters or placed in appendixes.
This author prefers the introduction of the mathematical
tools at the point they are to be used. But in any of these
orderings, there should be flexibiity for the instructor in
the attention to be given them, as there is typically a wide
variation in background on some of these matters among
entering students.
A major alternative to the above ordering is that of in-
troducing Maxwell’s equations at the beginning, allowing
everything, including statics, to follow from these. This
approach works if students have had a sufficiently rigor-
ous prerequisite physics course, but for most undergrad-
uates a beginning with the simpler static laws helps to
give a feeling for both the physics and mathemtics of field
concepts before proceeding to the more general dynamic
laws. Too much time with statics, however, may imbed
static concepts into the mind so deeply that they provide
difficulty in moving on to more general concepts. (The
misuse of scalar potential in time-varying problems is a
common example.)
Much of the development will be done with simple di-
electric, magnetic, or conducting materials, both for rea-
sons of simplicity of concept and practical importance of
such materials. Nevertheless, it is important to recognize
at the outset that many materials of modern electronics
and optoelectronics are inhomogenous, anisotropic, non-
linear, or time varying-and sometimes a combination of
all these things. It is especially important in presenting
general theorems such as uniqueness or reciprocity to
make clear the limitations on material properties that are
assumed in any given proof.
Moving systems are not emphasized much in a typical
undergraduate course. Two or three pages out of a few
hundred appears to be typical coverge for the most-used
undergraduate texts. This is in part because so few of our
graduates work with electrical machinery for which such
concepts are essential, and also because such systems pre-
sent added difficulties. At a minimum, it is important to
recognize the importance of motion to Faraday’s law [7].
In turning to the graduate level, one can use the same
order of presentation as for undergraduates, with every-
thing done at a higher level of sophistication and mathe-
matical development. Since the students have already met
Maxwell, it does seem very desirable to start with his
equations even if the next subject is to be the special case
of statics [8]. An intereting but optional extension is that
of deriving all the laws from Coulomb’s experimental law
by starting with relativity as a fundmental [9].
Much emphasis in the graduate course is typically
placed upon advanced methods of solution of boundary-
value problems-use of Green function techniques, vari-
ational methods, integral equations, etc.-but it is still
important to place the major emphasis on understanding
and the development of physical pictures. The role of nu-
merical methods will be discussed in a separate section.
- -
7 1
SPARKING UP FIELDS
As mentioned above, some students see with great clar-
ity the physics behind the necessary symbolism of elec-
tromagnetics, and find the subject easy because of the
logic with which everything develops from the basic laws.
Other students do not [lo]. For both groups, but espe-
cially the latter,it is important to use analogies and teach-
ing aids to the extent possible to help develop the needed
physical pictures. Some suggestions will be given in the
following, many of these developed and used by col-
leagues. Indeed the section title “sparking up fields” was
composed by a colleague [ 1 11 for a recent talk on the sub-
ject.
Analogies and Analog Demonstrations
There are many exact analogs to static fields because of
the universality of Laplace’s equation. Some of these are
familiar enough to the students that mention alone is help-
ful, as in the analogy between electrostatic and gravita-
tional potential. (But even here, bringing in a topographic
map can help.) Others need demonstration or hands-on-
work by the student. Some of the analogies useful for
demonstration of electrostatic fields in charge-free, sim-
ple media are as follows:
1) Electrolytic tank or resistance paper analog
2) Liquid crystal paper [12]
3) Soap bubbles or rubber membrane models
4) Fluid flow field mappers [ 131.
The resistance paper is applicable to Cartesian two-dimen-
sional problems, but the electrolytic tank can be tipped to
model axially symmetric problems also. The rubber mem-
brane, pushed by elements to heights proportional to elec-
trode potentials, has the advantage that small ball bear-
ings roll down the slopes with paths approximating the
paths of electrons in the corresponding field. The fluid
flow models can be extended to regions with sources
(Poisson’s equation) or to axially-symmetric regions [ 131.
It is harder to find equally good analogs for time-vary-
ing problems, but many exist for wave motion in general.
A film on similarities of wave motion, prepared by the
Bell Laboratories [ 141, is particularly good in showing the
effect of mismatching in producing reflections. Spring-
mass systems, as analogs to resonant systems, can be use-
ful, especially when coupling is added to show the trans-
fer of energy between coupled resonant systems. (Cou-
pled pendulum demonstrations are commercially
available, but home-made versions are easily put to-
gether.) This list is not exhaustive, but once an instructor
begins to think in these terms, he or she will soon develop
new ideas or variations of old ones.
Classroom Demonstrations and Minilabs
Short classroom demonstrations, such as voltage induc-
tion by Faraday’s law, reflection measurements on trans-
mission lines, or with quasi-plane waves, are very helpful
if well planned [15]. Some of these are better for hands-
WHINNERY: TEACHING ELECTROMAGNETICS 5

on experience, and may be set up as a minilab. A few

years ago one of our colleagues [ 161 set up such a minilab
with six demonstration experiments to be completed in a
15-week semester. Each demonstration experiment was
planned for about an hour, with short preparation time.
Experiments were scheduled flexibly through a sign-up
procedure. These continue to rank high in student evalu-
ations. Recent topics for the minilabs were
1) electrostatics (resistance-paper analog and Mac-
Intosh models)
2) magnetic measurements and Faraday’s law (Fig. 1)
3) pulse propagation (experiments on coax line and
MacIntosh models)
4) microwave measurement instruments Fig. 1. One of the minilabs in which Faraday’s law and the Hall etfect are
5 ) reflection and matching of coaxial transmission lines used to measure the magnetic field of a solenoid.
6) plane wave experiemnts (Fig. 2).
The minilabs are not intended to replace a full laboratory
experience, but that is often better scheduled after the first
course, with more sophisticated experiment and measure-
ment techniques.

Films and Videos

A number of excellent films and videos on electromag-
netics are available, the format to be used dependent upon
class size and the equipment available for clear viewing.
If video equipment is practical, one can record with a
video-camera short experiments or demonstrations in the
research laboratories to further enrich the presentations.
When using presentations prepared by others, it is impor-
tant to preview them to be sure they are worth the time
and are of the right level.
After trials with both long and short films and video,
we have found that short ones ( 5 to 10 min) are more
effective, inserted in the lecture at a position to illustrate
a particular point. The film on wave phenomena 1141, al-
ready mentioned, did hold student attention for all of its
half-hour running period. A colleague’s film on wave ve-
locities [17] was also useful, but has been even more
helpful when broken into 10-min video segments [See Fig.
3(a), @)I.
Computer Modeling
Computer models have the advantage over the above in
that they can be interactive, so that parameters may be
changed and the consequences immediately observed.
Static field problems and problems of wave reflections are
subjects especially suitable for such programs. Programs
suitable for PC’s with standard software are described in
the literature [18], and commercial programs are also
available 1191. As noted in the earlier list of minilabs,
these are especially suitable for that format.
Show und Tell-und More
It is also helpful to bring to class samples of wave-
guides, resonators, and other components utilizing elec-
tromagnetic principles. Comparisons among elements for
Fig. 2 . A minilab illustrating Snell‘s law. total internal reflection. and
Brewster’s angle. using a He-Nc laser as the source of quasi-plane
waves
different frequency ranges, and especially between micro-
wave and optical ranges. are especially useful. Term pa-
pers, individual projects, and class visits to research lab-
oratories can be especially effective. Once the decision is
made to look for such “sparks,” each instructor will have
fun in generating new ones suitable to his or her outline
and format. Good ideas will be found by following this
TRANSACTIONS and the American Journal of Physics [20].
Is ELECTROMAGNLTICS FOR EVERYONE?
It has been noted that electromagnetics was an ignored
subject before World War I1 and a required one in the
period immediately after. Now there is a mix. and a good
deal of discussion as to its relevance to certain EE spe-
cialties. Students interested in microwaves, plasmas, op-
toelectronics, advanced solid-state devices, and electro-
mechanical energy conversion can clearly see the
necessity of understanding the subject. Others may not
see the need so clearly, but bioelectronics engineers con-
cerned with modem measurement systems can see the ne-
cessity for understanding of fields in the body and of
transmission phenomena. Circuit and computer engineers
must be concerned with delay phenomena, dispersion, and
6
ib)
Fig. 3. Excerpts from the film and video\ o n v,a\e velocities I17j. ( a ) I!-
lustration o t a wa\eforni in time. its transform in the frequency domain.
and an ci - B diagram of a tran\tiii\sion circuit h i t h dispersion. ( b ) The
o\cilloscope picture showing experimental confirmation of pulse
broadening because of dispersion in the tran\mission circuit.
unwanted coupling between elements through radiation or
evanescent fields.
For each of the above groups, it is important to have
problems which utilize field theory for their special inter-
ests. A colleague 1211 who recently completed a special
course in fields for computer engineers found it rough
going in establishing the preliminaries, but did find real
interest in the subject of magnetic storage and greatest
interest in the question of delays and dispersion when data
trains are transmitted over various transmission systems-
especially optical fibers.
THE ROLEOF COMPUTATIONAL M~.I.HOIX
As with all aspects of engineering. numerical compu-
tational methods are playing an increasingly important role
in the solution of electromagnetic problems and this fact
obviously has impact on the presentation of the subject.
At one level, one simply finds it more practical than be-
fore to evaluate series solutions and other complicated so-
lutions that result from classical methods of analysis.
Problems can then be assigned that would have once been
considered as cruel and unusual punishment. But more
fundamentally, one can apply finite element and other nu-
merical techniques to the basic partial differential equa-
tions with prescribed boundary conditions [22]. The
memory capacity of modern computers now makes pos-
sible modeling in the time domain [23]. There is a con-
tinuing evolution and it is important to follow this in order
to decide upon the optimum way of presenting the mixture
of analytical and numerical techniques.
Since most of the analytical techniques we use are de-
veloped for linear systems, numerical methods become
increasingly important as one faces nonlinear elements or
materials. But even here-as in the study of solitons [24]-
approximate analytical solutions can play an important
role.
CONCLUSIONS
We can think of our goal in teaching electomagnetics
as one with at least four channels:
1) development of physical pictures and a good intui-
tive understanding of field and wave phenomena;
2) development of the ability to use the most imporant
analytical tools for solution of problems in electromagne-
tism:
3) development of an understanding of the role of nu-
merical techniques as these become increasingly impor-
tant:
4) presentation of practical problems from a variety of
specialties of electrical engineering.
It is important in all of this to recognize the increasing
importance of materials which may be inhomogeneous (as
an ion-implanted semiconductor circuit), anisotropic (as
a lithium-niobate modulator), nonlinear (as a magnetic
storage element), or time varying (as a surface-acoustic-
wave device). Or (as with a plasma) there may be a com-
bination of all these characteristics!
The base in Maxwell’s equation will remain the same
but there will be the continuing excitement of new sys-
t e m to which these must be applied, new techniques of
analysis and synthesis. and new methods for presenting
the material.
We conclude with another quote from Feynman [ 2 5 ] :
“From a long view of the history of mankind-seen
from, say, 10 000 years from now-there can be little
doubt that the most significant event of the 19th Century
will be judged as Maxwell’s discovery of the laws ofelec-
trodynamics. The American Civial War will pale into pro-
vincial insignificance in comparison with this important
scientific event of the same decade.”
ACK~WWLEDGMEN~
With deep appreciation I acknowledge the influence of
S. Ramo-my first teacher of Maxwell’s Equations. I also
greatly appreciate the insights. both in content and teach-
ing methods, given me by many colleagues at Berkeley:
 I
WHINNERY: TEACHING ELECTROMAGNETICS
D. J . Angelakos, C . K. Birdsall, A. Dienes, T. K. Gus-
tafson, M. A. Lieberman, A. J. Lichtenberg, K. K. Mei,
S. E. Schwarz, T. Van Duzer, R. M. White, the late S.
Silver, and by friends in other institutions too numerous
to list.
REFERENCES
[ l ] R. P. Feynman, The Feynman Lectures on Physics, Vol. 1 Reading.
MA: Addison-Wesley, 1963, p. 28-1.
121 The familiar vector notation, now commonly used, was introduced by
Willard Gibbs and Oliver Heaviside ca 1895. Maxwell expressed his
work in terms of Hamilton’s quaternions. See M. J. Crowe, A History
of Vecror Analysis. New York: Dover, 1986.
[3] Sir E. Whittaker, “A history of the theories of aether and electric-
ity,” Amer. Insr. Phys., vol. I, 1987.
[4] A. A. Oliner, “Historical perspectives on microwave field theory,”
IEEE Trans. Microwave Theory Tech., vol. MTT-32, pp. 1022-1045,
Sept. 1984.
[5] K. S. Packard, “The origin of waveguides: A case of multiple redis-
covery,” IEEE Trans. Microwave Theory Tech., vol. MTT 32. pp.
961-969, Sept. 1984.
[6] This writer still remembers his struggles with the concept of subtran-
sient synchronous reactance of electrical machines, and with the “de-
rivations” given in popular texts of radiation from a dipole antenna
ca 1937.
[7] Many graduate-level texts give more complete coverage. An excellent
review of the approach to moving systems is given by C . T. Tai, “On
the presentation of Maxwell’s theory.” Proc. IEEE, vol. 60, pp. 936-
945, Aug. 1972.
[8] J. A. Stratton makes this point clearlyin the Preface to his classic text
Elecrromagneric Theory. New York: McGraw-Hill, 1941.
[9] For example, R. S . Elliott, Elecrromagnerics. New York: McGraw-
Hill, 1966.
[lo] M. N. 0. Sadiku, “Problems faced by undergraduates studying elec-
tromagnetics,” IEEE Trans. Educ., vol. E-29. pp. 31-32, Feb. 1986.
[ l l ] R. M. White, Univ. California, Berkeley.
[12] Obtainable, for example, from Edmund Scientific Company, 101 E.
Gloucester Pike, Barrington, NJ 08007. Electrodes at different tem-
pertures are placed on the paper and color differences show variations
of potential.
[I31 A. D . Moore, “Fields from fluid flow mappers,” J . Appl. Phys., vol.
20, pp. 790-804, Aug. 1949; also Trans AIEE, vol. 71, Part I, pp.
1-5, 1952.
[14] Shive, “Similarities in wave behavior,” Now in video format. For
information on availability, requests through your library to Library,
AT&T Bell Laboratories, Whippany, NJ 07981.
[ 151 The recent book, Elecrromagneric Fields and Energy, by H . A. Haus
and J. R. Melcher, contains 37 detailed descriptions of possible dem-
onstrations!
[I61 M. A. Lieberman, Univ. California. Berkeley.
7
[17] T. Van Duzer, “Wave velocities, dispersion, and the w - 6 dia-
gram,” originally produced by the Educational Development Center
for the National Committee on Electrical Engineering Films, 1969.
[ 181 For example, T. T . Crow, “Solutions to Laplace’s equation using a
spreadsheet on a personal computer,” Amer. J . Phys., vol. 55, pp.
817-823, Sept. 1987.
[ 191 The programs used in our minilabs were obtained from Kinko’s Ac-
ademic Courseware Exchange. 255 West Stanley Ave., Ventura, CA
93001.
[20] Amer. J . Phys., a journal of the American Association of Physics
Teachers, published by American Institute of Physics, 335 E. 45th
St., New York, NY, 10017.
[21] A. Dienes, Univ. California, Davis.
[22] V . Vermuri and W. J . Karplus, Digital Computer Treatment of Par-
rial Diferenrial Equarions. Englewood Cliffs, NJ: Prentice-Hall,
1981.
[23] A. C. Cangellaris, C. C. Lin, K. K. Mei, ”Point-matched time do-
main finite element methods for electromagnetic radiation and scat-
tering,” IEEE Trans. Antennas Propagar., vol. AP-35, pp. 1160-
1173, Oct. 1987.
1241 H. A. Haus, Waves and Fields in Optoelectronics. Englewood Cliffs,
NJ: Prentice-Hall, 1984, Sec. 10.2.
[25] R. P. Feynman, The Feynman Lectures on Physics, vol. I1 Reading,
MA: Addison-Wesley, p. 1-11.
John R. Whinnery (A’41-SM’44-F’52) was born
in Read, CO, on July 26, 1916. He received the
B . S . degree in electrical engineering in 1937 and
the Ph.D. degree in 1948, both from the Univer-
sityof California, Berkeley.
From 1937 to 1946 he was with the General
Electric Company, Schenectady, NY, working on
various microwave problems. and during that pe-
riod coauthored, with Simon Ramo, the first edi-
tion of the text Fields and Waves in Modern Ru-
dio. Since 1946 he has been at the University of
California, Berkeley, as faculty member, Chairman of the EE department.
and Dean, returning to teaching in 1964 and was named a University Pro-
fessor in 1980. Leaves from his Department have been spent at the Hughes
Aircraft Company, Bell Laboratories, and several universities. His re-
search interests since 1964 have been concentrated on photonic devices.
Dr. Whinnery received from the IEEE the Education Medal (1967). the
Microwave Career Award (1976), Centennial Medal (1984) and Medal of
Honor (1985); he received the Lamme Award of the American Society for
Engineering Education (1975) and the Founders Medal of the National
Academy of Engineering (1986). He is a member of the National Academy
of Engineering, the National Academy of Sciences and is Fellow of the
American Academy of Arts and Sciences, the Optical Society of America
and the American Association for the Advancement of Science.