IEEE Power Electronics Magazine - June 2020 - 104

White Hot

by Robert V. White

My First Power Supply Design

s this issue of IEEE Power
Electronics Magazine goes to
press, it will be 40 years
since I graduated from the Massachusetts Institute of Technology with
my B.S. degree in electrical engineering. This has me waxing a bit reflective, so I thought I would share with
you the story of my first power supply design.
Technically, my first power supply
design was for the General Electric
Power Systems business unit in Philadelphia. During the summer of 1979,
between my junior and senior years,
I worked there as a summer engineer. My task for the summer was to
design a dc-dc converter that operated from -48 V and with ±15-V outputs
rated at roughly 1 A. This was to be
the power supply for a new line of
solid-state relays the power systems
business unit was developing for the
utility market. What I designed, but
not did not know it at the time, was a
primary-side-regulated, peak-current-mode-control flyback converter
operating in discontinuous conduction mode with a fixed off-time control. The control IC was a single
LM339 quad comparator driving a
Hitachi power MOSFET through a
discrete totem pole. The prototype
worked well. I wish I had kept a copy
of the schematic. I have, from time to
time, made some sketches, but I don't

Digital Object Identifier 10.1109/MPEL.2020.2986844
Date of current version: 16 June 2020

104

IEEE POWER ELECTRONICS MAGAZINE

think I have ever completely recreated that design. I heard later that some
variation of it (with a bipolar transistor instead of the MOSFET, for example) eventually went to production,
but I did not see that and do not know
for sure.
My first real production design
started after I graduated. One week
after graduation, I went to work in the
Small Systems and Terminals power
supply group at the Digital Equipment
Corporation (DEC) in Maynard, Massachusetts. The facility in Maynard,
known as "the Mill," was a complex of
New England-style mill buildings
clustered around a millpond. Some of
the buildings dated back to the early
1800s. These buildings, which are
still there, have thick brick walls,
well-worn wooden floors held up by
thick wooden columns, and massive
wooden beams that could not be
duplicated today. Every building has
lots of tall windows that let in the
maximum amount of natural light. It
was a unique place to work, with a
"heart and soul and character" that
could never be matched by a modern
office building.
I was assigned to design the power
supply for the Professional 350, DEC's
first desktop computer, which was
using a processor chipset from the
PDP-11 computer family. This was a
very high-priority and high-profile
project. The project was codenamed
the "KO" project after Ken Olsen, the
founder and president of DEC, who

z June 2020

was taking a very personal interest in
the project.
Although the system concept and
power supply specifications were still
being developed when I joined DEC,
it was known that there would be a
roughly 200-W power supply with
output voltages of +5, +12, and -12 V. I
was given the schematic of the VT103
power supply to study as a starting
place. The VT103 power supply had
been designed by Vatche Vorperian.
Yes, that Vatche Vorperian. After
working at DEC, Vatche went to
Caltech for his Ph.D. degree and then
later created the average pulsewidthmodulated switch model.
The VT103 power supply was a
fairly standard design for DEC's small
systems. The power supply input
was configured with a switch that
changed the input rectifier and filter
from a voltage doubler for operation
from 120 V to a full wave rectifier for
operation from 240 V. The main dc-dc
converter was a two-transistor forward converter operating at roughly
30 kHz. The control was closed
around the +5-V output, and the control IC was a Silicon General SG3524.
The +12-V output was regulated with
an auxiliary switch transistor in
series with the +12-V transformer secondary winding. This switch was controlled by its own SG3524. The controls were referenced to the secondary and powered by an auxiliary line
(continued on page 100)

IEEE Power Electronics Magazine - June 2020

Table of Contents for the Digital Edition of IEEE Power Electronics Magazine - June 2020