IEEE Power Electronics Magazine - September 2022 - 93

White Hot (continued from page 96)
cam caused the contacts to be closed.
The timing of the contact opening
was adjusted by physically rotating
the distributor. Although automobile
manufacturers started experimenting
with solid-state ignition system in the
early 1960s, the electromechanical
ignition system was still in wide use
up until about 1980. I still have in my
garage a timing light and " tach and
dwell meter " I used years ago to perform
engine maintenance on my cars.
Another form of nonsemiconductor
switch-mode power invented in
early 1900s used saturable magnetic
devices, often called " magnetic amplifiers "
or " mag amps. " These devices
typically feature two windings on a
common core. One winding carried
the power to be switched and was
made with the fewest windings possible
that gave enough inductance that
the impedance was large enough to
essentially block the flow of an ac
current. The other winding, called the
control winding, was made with a
larger number of turns of smaller
gauge wire. When the control winding
was not energized the core of the
magnetic amplifier was not saturated
and the power winding presented a
large inductance (and impedance)
that blocked the flow of ac power. By
energizing the control winding with a
relatively small current through a
large number of turns the core of the
magnetic amplifier could be saturated.
This greatly reduced the inductance
and impedance of the power
winding and ac power could then
flow. To turn off the magnetic amplifier,
the current in the control winding
was removed, returning the core to
the unsaturated state. One of the
main original uses of magnetic amplifiers
was for lighting controls including
dimming using the same phase
cut method later used in triac based
dimmers. The use of magnetic amplifiers
diminished after controllable
mercury arc rectifiers came into use.
Mag amps did have a resurgence in
the 1980s in switch-mode power supplies.
Mag amps were used as switches
on auxiliary outputs of the forward
and bridge converters that were commonly
used in computer power supplies.
The power supply would typically
regulate its main +5 V output by
varying the duty cycle of the main
switches. A mag amp in series with
the secondary of a second output, typically
+12 V, could be used to regulate
that output independently of the main
output. This was a simple and relatively
inexpensive way to precisely
regulate an additional output. The
late Chuck Mullet was so well known
for his expertise with this device and
circuit to the point he earned the
nickname " Mr. Mag Amp. "
The first true power electronic
devices were vacuum tube diodes
developed in the very early 1900s.
These remained in wide use, even
with their relatively low efficiency
and need for heater power until
replaced by the first solid power electronics
device, the selenium rectifier,
in the mid to late 1940s.
The first controlled power electronic
switches were mercury arc rectifiers.
In a mercury arc rectifier, the
internal arc will extinguish, stopping
conduction, when the current falls
below a critical threshold. To restart
conduction, an ignitor was used to
create a small arc that vaporized
some of the mercury to allow conduction
to resume. While this controlled
conduction method was available,
mercury arc rectifiers were generally
used more as uncontrolled rectifiers
for high voltage and high current
applications. They remained in use
until the 1970s when solid-state rectifiers
and thyristors became main
stream in high-power applications.
So far in the discussion, power " electronics "
is not really a field. Electromechanical
switch flybacks, magnetic
amplifiers, and vacuum tube diodes
don't make for much of a power electronics
technology. The only real topologies
in use are the flyback converter
(automotive ignition, television receivers)
and the phase-controlled rectifier.
So what comes next that really
launches power electronics as an
important technology? Is it device
or topologies? The answer is very
clear, the development and commercialization
of solid-state controllable
power switches.
We all know that the transistor
was first developed at Bell Laboratories
in 1947. The small signal transistor
developed steadily through the
1950s. By about 1960, the portable
transistor radio, typically powered by
a 9 V battery, was common. I still
have one that my grandmother gave
me in about 1962. Transistors suitable
for switching and controlling
power were investigated but the first
really successful power transistor
was the 2N3055, introduced in the
early 1960s (and sti ll avai lable
today!). The other key invention of
that era was the silicon-controlled
rectifier (SCR) in 1957. With these
devices, the modern era of solid-state
power conversion was launched.
The 1960s saw rapid early development
of the first switch-mode power
converters. The 1966 paper " Basic
considerations for DC to DC conversion
networks " by Moore and Wilson
provide some of the early theoretical
explanation for the fundamentals of
switch-mode conversion and circuits.
Dan Wolaver's 1969 MIT doctoral thesis,
" Fundamental study of dc to dc
conversion systems, " provided rigorous
proof that for voltage or current
dc-dc conversion at least one nonlinear
device (switch) was necessary.
The need for light and reliable
power converters for the emerging
aerospace industry drove developments
of circuits, topologies, and controls
of dc-dc converters. The SCR
made solid-state motor drives and
inverters possible. Power electronics
engineers quickly took advantage. The
classic textbook, " Principles of Inverter
Circuits " by Bedford and Hoft, was
published in 1964. There was work on
various resonant type circuits that featured
zero current switching that
caused the SCR to turn off with natural
commutation. Forced commutation
circuits, which foreshadowed the auxiliary
resonant pole converters
September 2022 z IEEE POWER ELECTRONICS MAGAZINE 93
IEEE Power Electronics Magazine - September 2022

Table of Contents for the Digital Edition of IEEE Power Electronics Magazine - September 2022
