Diamp
Please click
here for schematic
Diamp is an audio
amplifier that uses power supply-type rectifiers in
place of the usual transistors or vacuum tubes. It employs the
phenomenon of charge-storage in order to produce power gain. Stored
charge is normally characterized as a power supply rectifier defect;
given that it increases the reverse-biased recovery time. However,
Diamp converts this defect into a virtue! The first task involved in
building a Diamp is to select rectifiers having a sufficiently long
reverse-recovery time.
The waveforms shown on
the right were obtained from a test circuit
comprised of an AF signal generator driving the diode-under-test
through a 1k Ohm (current-to voltage converter) resistor. An
oscilloscope probe was placed across this resistor to obtain the plots
shown on the right. The 20kHz waveforms shown (top-to-bottom) were
produced by the following rectifiers: 1N4148, 1N4002, 1N4004, 1N4007
and 1N5401. The prominent negative charge recombination pulses produced
in all but the top trace illustrate why these devices do not
efficiently rectify AC frequencies much higher than several hundred
Hz.
Charge storage time is
the length of time required for the stored
minority charges to recombine when a forward-biased PN junction is
suddenly reverse-biased. I measured (in the above diode order): 0uS,
5uS, 5.5uS, 6.2uS and 11.2uS.
A transistor or vacuum
tube uses a DC collector or plate power supply.
In contrast, the Diamp uses an AC power supply, or carrier. The lowest
frequency required for the carrier is determined by both the signal
frequency and the rectifier storage time. The carrier frequency ought
to be at least ten times the signal frequency. A shorter storage time
requires a higher carrier frequency in order to produce an efficient
amplifier. In fact, the short carrier storage times associated with the
germanium diodes used in the original (circa 1954-58) prototypes
required a carrier frequency of 1MHz in order to produce a reasonable
gain.
Some History
The
idea of building an amplifier from a rectifier that exhibits
charge-storage appears to have first occured to Arthur Holt, an
engineer working at the National Bureau of Standards. In 1954,
he applied for a patent which was granted in 1959. Two General
Electric engineers,
Harold Abbott and Lawrence Wechsler, came up with an improved circuit
in 1958. They received three patents and presented their findings in
the literature of the day. To the best of my knowledge, this circuit
was never widely used. In fact, it appears to have been
all-but-forgotten. This stands to reason, given the increased
availability and performance of transistors.
I first came upon
this circuit in a booklet of the, "101 Ways to Use Semiconductors"
variety, some fifteen years ago. Unfortunately, I neglected to record
the details at the time and promptly forgot everything; save the basic
- albeit sketchy - idea.
Recently, I happened to
be reading an old issue of IEEE Transactions,
when
I noticed a footnote mentioning a diode amplifier. The text led me
to believe this wasn't tunnel diode or varactor-based parametric
amplifier circuit. It turns out this was indeed the same amplifier that
I'd happened upon a decade and a half earlier!
My Circuit
My circuit is based on
Abbott
and Wechsler's improved version. My first iteration used wide-band
transformer coupling. My latest version uses peaked LC resonators
having a modest loaded Q. A small DC bias is required at the
charge-storage diode cathode in order to place the amplifier in Class-A
operation. The optimum bias setting is fairly critical. I initially set
each stage for maximum gain using a variable resistor in place of the
grounded, bias-divider resistor.
To be continued...
This video clip
presented below shows a three-stage Diamp
working into a loudspeaker.
This
clip shows the
signal at the second-stage charge-storage
diode anode. This two-stage amplifier drives a pair of
headphones. It uses the same gain-less, RF front-end that was used in
the above video.
References
"Diode
Amplifier,"Electronic Design Magazine, October 1954, pp. 24-25
"Now
Diodes
Amplify,"Radio-Electronics Magazine, November 1954, pp.
94-95 (thanks to Jack Ward, curator of the Transistor Museum!); View Page 94, View Page 95
"Diode
Amplifer,"National Bureau of Standards Tech. News Bulletin,
Vol. 38, October 1956, pp. 145-147
Solid
State
Circuits
Conference,
Digest
of
Technical
Papers;
IEEE
International, February 1958, Vol. 1, pp. 57-59
"Semiconductor
Diode
Amplifiers
and
Pulse
Modulators,"IRE Trans. on Electron
Devices, Vol 6, Issue 3, July 1959, pp. 341-347
"Model
and
Experiments
on
Charge
Storage," IRE
Trans.
on
Electron
Devices, Vol. 8, Issue 2, March 1961, pp.
123-131
"Diode
Amplifier,"US
Patent
2,879,409, Arthur Holt (National Bureau of Standards)
"Semiconductor
Circuits,"US
Patent
2,981,881, H. Abbott and L. Wechsler, (General Electric
Co.)
"Semiconductor
Circuits
Utilizing
a
Storage
Diode,"US
Patent
2,976,429, Ibid
"Semiconductor
Diode
Amplifier,"US
Patent
2,997,659, Ibid
Homebrew QRP de AA1TJ
