AGC circuit uses an analog multiplier
This automatic gain control works over the audio-frequency range.
Steve Lubs, Department of Defense, Washington, DC -- EDN, November 3, 2011
Originally published in the September 4, 1986, issue of EDN
In the AGC circuit of Fig 1, a 4-quadrant analog multiplier (IC1), an amplifier stage (IC2), an active, full-wave rectifier (D1, D2, R4-R7, and IC3), and an integrator (IC4) accomplish automatic gain control of VIN’s amplitude variations in the audio-frequency range.
Circuit analysis yields the frequency-response equation
This AGC circuit is suitable for
controlling the long-term variations
of amplitude within a limited range. It
doesn’t respond uniformly over a wide
dynamic range, however, because the
time response is inversely proportional
to input-signal amplitude.Talkback
-
AGC - here we go
www.onsemi.com SA571
dov.rossitzan@argoncorp.com
Dov Rossitzan - 2011-14-11 07:56:31 PST -
I think maybe it is good idea that EDN started periodically to re-publish classic circuitries, but it should be done very carefully.
Looks like the mistakes in original publications will live forever. I mean you can find some problems, post your comments and etc, but nobody will correct schematics and text.
So 25 years later this circuitry could be published with the same mistakes as an original as happen now.
Vladimir Doubovis - 2011-12-11 09:49:11 PST -
Despite the fact that many have detected problems with the schematic, I'd like to know the noise performance of this type of AGC. Multipliers, even these days, have notoriously high noise levels which limits the suitability of this type of circuit.
Steve Tomporowski - 2011-8-11 04:59:48 PST -
I agree with Vladimir in that the circuit does not look quite right. The area around U3 does not look like a precision rectifier, but more towards some clamp circuit. So there do appear to be a few errors. If that is a sample of the "best of" publication, then I will certainly save my money. I have been saving the design ideas almost constantly since 1974, so I may have the original filed someplace. Not having them organized does make them a bit less useful, though.
So please let us know what the truth is behind a published old idea with a few problems, OK?
William Ketel - 2011-7-11 18:20:07 PST -
Since it comes from a 1986 issue, this article is indeed quite old. The price of the multiplier is indeed quite high by current standards. The functionality is still a quite useful concept. The circuit could certainly be duplicated by more current parts, and probably would perform better as well.
So I do wonder about the motivation for republishing this circuit, and who it was discovered that it had been published some 25 years earlier. Would the editor of this feature please comment on that part.
What is demonstrated is that some applications do not go away as time passes.
One last comment, which is that Analog Devices now offers several products that deliver even better performance while comprising only one IC package and supporting passive components, and needing only a single supply voltage.
William Ketel 2 - 2011-7-11 14:53:27 PST
