VCA discussion

[Magnus Danielson]

>>>>> "g" == gstopp  <gstopp at fibermux.com> writes:

 >> 
 >> Does this mean that an opamp/transistor lin/exp converter will work 
 >> very nicely to control the circuit gain?  I hope so.
 >> 

 g> It appears that it's already exponential as it is. I hope that one of the 
 g> transistor theory gurus can verify this in theory at least (Joachim? Juergen? 
 g> Magnus?)

You rang!

Most VCA circuits that I have seen utilizes the exponential properties
of an transistor in the following sense:

The will convert the signal using an logarithm circuit, make an
addition of the controlsignal and convert it back thougth a
exponential circuit. In order to make an linear respons must the
controlsignal also pass througth an logarithmic converter before being
added to the signal. If you do not convert the signal, then will the
control signal have an logarithmic behaivior. This is very similar
with the input stage/current source part of the ASM-1 VCOs which I
would not be supprised if only a few people fully understood.

An Gilbert cell (like the MC1496) is a more complex creature but works
along similar lines, however, it will need some additional correction
circuits in order to do a linear multiplication correct for larger
signals. The Gilbert cell is in many ways among the best for
VCA/ring-mod. The lack of correction circuits will add distrosion
which may be apprechiated (but that is as allways a diffrent issue).

In an transconductive opamp will a transistor change the standing
current for a normally standard op-amp input pair and effectively
change the input gain. There is typhically an input transistor
connected to -Vcc which will be a current mirror or possibly just work
as a current amplifie The gain of such a stage is a (relatively) linear
function to the input gain. BTW does the Moog VCF work in similar ways.

That's what I have in my head rigth now... please let me know if ya
like more details, cause then I be more then happy to dig into my
books yet again. You learn as you explain... keep 'em comming!

Cheers,
Magnus