> I'll probably try the op-amp/jfet circuit. Interesting to see that the
>> key to the agc is to use the resistive region of the jfet. I was thinking
>> the variable gain would be done with a non-linear component.
>
> Has to be linear, else you'll end up with distorted audio.
Well yes ok. But if e.g. you have a square law response and the signal is very small it approximates to linear. Varying the gain is then just a matter of adding an offset to the input. Not saying it's the best way of doing it though.
Rob.
There are many other ways, for example Gilbert multipliers, dual-gate FETs
> and so on. The opamp/JFET combo is just one of the solutions and it works
> (I've used it).