(zipzit) wrote in message . com...
Anybody out there work on Audio circuits?

I'm trying to design & build a motorcycle intercom, similar to the
Starcom1? see http://www.starcom1.com/

snipped

Is there anyway to use microphone sound density input to generate a
isolated variable resistance output, to work as an automatic volume
control for varying ambient noise levels? Are there other ways to do
this?

Thanks in advance for any design hints?
LB
Detroit, Michigan

Cross posted to:
sci.electronics.design
rec.audio.tech
sci.electronics.basics

------------------------

LB,

As John Woodgate (I think it was) said, you could use an FET as a
voltage-controlled resistance, to control the gain of an opamp, or a
similarly-controlled amplifier.

I did something that's somewhat similar to what you are wanting to do:

To sense and convert the signal (in your case the mic output) that
should affect the gain, I used an opamp circuit similar to a "peak
detector" or an "envelope follower". There are many examples of both
of those, on the web, especially in some of the opamp application
notes at IC manufacturers' websites. I can email the schematic of the
one that I used, to you, if you like. That circuit should produce a
quasi-DC voltage that is proportional to the mic output's amplitude.
You can then put that through another opamp circuit, to buffer, and
possibly change the level of, and/or the sign of, that voltage. You
could also do other stuff to it, as needed, before using the voltage,
or a current produced with it, to drive some variable-resistance
device that controls an amplifier's gain.

In your case, you may need to do a little more conditioning of the
signal, at some stage, or find some way for the circuit to tell the
difference between the times when there is only ambient noise and the
times when there is a voice or other input, and some way for it to
"hold" the gain-setting voltage that's derived from the ambient noise
level, when there is a voice or other desired stronger input present.

Instead of using an FET as the controllable gain-setting resistance, I
used a simple analog optical isolator, a VTL5C2 "Vactrol", which is a
current-controlled-resistance device and is actually just an LED
encapsulated with a photocell. So there are four leads: two for the
LED (and your control current) and two for the photocell, across which
a "pure" resistance is presented (at low frequencies, at least; never
looked at higher than audio).

In the case of the VTL5C2 model, the resistance across the photocell's
two leads varies from about 2 megohms, with 0 mA through the LED, down
to about 200 Ohms, with 40 mA through the LED. (The VTL5C2 is
available from bgmicro.com, for $0.50 for qty 1.) Check out the intro
and appnotes links for them, at:

http://optoelectronics.perkinelmer.c...e=VTL5C+Series

In my circuit, the final control voltage output from an opamp circuit
was connected to the vactrol's LED's positive lead through a resistor
of a few hundred ohms (with the LED's neg lead connected to ground),
in order to produce the proper current range for my application.

Note, also, that the vactrol's resistance does not vary linearly with
current. In my case, that didn't matter too much, since it was inside
of a feedback control loop and just needed to get to the correct
resistance within about a second or so. But it probably won't matter
for your AGC application, either.

Good luck. I hope this helps.

Regards,

Tom Gootee

http://www.fullnet.com/u/tomg

-----------------------------