Page 99 - Analog Circuit Design Art, Science, and Personalities

Propagation of the Race (of Analog Circuit Designers)

Subjects that provide the background essential for design are offered by all of our
departments. An excellent example is the widely acclaimed Introduction to Design
architected by Professor Woodie Flowers of the department of Mechanical Engi-
neering. The culmination of this subject is a spirited contest that finds which student-
designed machine best accomplishes a specific task. Although Introduction to Design
and many other M.I.T. courses would provide interesting examples of approaches
to teaching design skills, 1 will limit this discussion to those subjects with which 1
am involved and which I have taught several times. The sub.jects described have
evolved to their current form and content via the contributions, suggestions, teach-
ing, and inspiration of many colleagues, including Professors Hae-Seung Lee,
Leonard A. Gould, Winston R. Markey, and Campbell L. Searle, and Drs. Chathan
M. Cooke, Thomas H. Lee, and F. Williams Sales, Jr.
Engineering education at M.I.T. and elsewhere started a fundamental change in
the 1950s because of the pioneering effort led by Professor Gordon S. Brown, then
head of M.T.T.’s department of Electrical Engineering. Prior to that time, engineer-
ing education was generally quite specific, with options that channeled the student
into a narrow area early in his or her educational process. However, the technolog-
ical explosion that followed the second world war made it impossible to predict
areas likely to be of interest even at graduation, much less a few years later. The
approach that evolved from this dilemma was to provide an education broadly
based in mathematics and physics. Regardless of the opportunities available to the
graduate. the truths discovered by Fourier, Maxwell, and Laplace would be essen-
tial. The new engineer would have a background that permitted easy assimilation
of the specifics of any particular area, including design.
A further justification for this approach is that analytic skills are the ones most
difficult to acquire through self-study, with the discipline and structure provided
by the classroom almost required for success. Few high school students study
vector calculus on their own because of the joy it provides. Conversely, hobbies
or acquired interests often lead students to “thing”-oriented pursuits such as circuit
or computer hacking long before they get to college.
It is impossible to argue with the general success that this approach to education
has enjoyed. However, the potential negative impact on the propagation of the race
of designers comes when a student spends 4 or 6 or 8 years (depending on the final
degree obtained) in an academic program devoid of hardware and design experi-
ence. While this student could become an innovative and productive design engi-
neer with a very short internship in a specific area, he or she may not ~’clnt to. This
bias is particularly likely when none of the academic role models practice design.
The three subjects to be described provide a degree of balance by exploring de-
sign-oriented specifics and philosophies. It really doesn’t matter if an occasional
specific is obsolete when the student leaves M.I.T. The generalized background
acquired from other subjects allows easy adaptation for the student whose career
has been motivated in this direction. The subjects are electives and thus acquire
their enrollment only because of residual interest from earlier, often pre-M.I.T.,
experiences or because of a favorable student grapevine.
These subjects share a number of features. All have an associated laboratory,
with students averaging approximately 2 hours per week in this endeavor. While the
details of the laboratory vary depending on the subject. all reflect our belief that it is
essential to attempt actual design in order to learn how to do it. None demands
much literary effort in the final write up (not because we don’t think this aspect is
important-but we prefer to exercise other skills in the limited time available). All
laboratory exercises require close interaction between the student and a teaching

80