Log/exponential calculations re. amp gain?

[Scott Kelley]

I have an amplifier circuit with AGC. I am told that the AGC circuit
attenuation of this amplifier can be determined from the voltage across a
diode in the AGC circuit. As I understand it, the voltage across a diode is
logarithmically related to the current, with the current, in this case
directly related to the attenuation.

Problem is, I don't understand what I think I know about logarithmic &
exponential functions.

What I ultimately want to do is:
* Measure the signal level out of the amplifier.
* Measure the voltage across the diode, & calculate the attenuation.
* From this, calculate the input signal level (I am actually only concerned
with relative signal levels, not the absolute level of the input).

Here's what I measure:

Input Voltage Diode Voltage
< 0.02 .4241 Quiescent voltage of diode with
1M bias resistor to 5V+
0.02 .4242 The AGC is just starting to
kick in at this point.
0.05 .4279
0.10 .4334
0.20 .4424
0.5 .4620
1.0 .4827
2.0 .5085
5.0 .5483
10.0 .5827
20.0 .6235

Can anybody help me to understand where I need to go from here?

Thanks
Scott Kelley

 

[PeteS]

There are a number of unknowns, the primary one being we can't see the
circuit. Without that, it's hard to help.

In general, though, AGC feedback is proportional (indirectly, as you
deduce) to input signal level, but that specific relationship depends
on it's implementation. Another thing; are your input voltages ac? One
would assume so, but it's nice to clarify.

Cheers

PeteS

 

[Scott Kelley]

There are a number of unknowns, the primary one being we can't see the

circuit. Without that, it's hard to help.

In general, though, AGC feedback is proportional (indirectly, as you
deduce) to input signal level, but that specific relationship depends
on it's implementation. Another thing; are your input voltages ac? One
would assume so, but it's nice to clarify.

Yes, the input listed is AC, and is in the middle of the passband.

I have been assured by the circuit designer that the attenuation can be
determined by the voltage across this diode, and that it is a logarithmic
function. Unfortunately, he is no longer around to ask for further details.
Although I know that I have it, I have not been able to find the schematic.

I would think that the numbers shown would at least give a clue as to the
relationship between the diode voltage and attenuation.

For the sake of current discussion, the output, within the AGC range of
voltages listed is constant.

Scott