[sdiy] Two-OTA VCAs (a la Oakley)

[David G. Dixon]

Well, my random walk along the path of building synth modules has finally
led me to the venerable VCA.  Upon looking around for nice circuits and
panel layouts, I ended up (once again) at the Oakley site, where I got
embroiled in the Triple VCA User's Manual write-up.  These VCAs involve two
OTAs in series, and are based on a circuit published by a guy named Mike
Sims in the Design Ideas section of the January 19, 1995 edition of EDN
magazine.  Here's the link:

 

http://www.edn.com/archives/1995/011995/02di1.htm

 

The signal comes in via a current-limiting resistor into the inverting input
terminal of OTA #1 whose Iabc input is hard-wired (via a resistor) to ground
and whose output is fed directly back to the inverting terminal (no
resistor).  This gives a much attenuated output current at very low
potential which is fed to the inverting input of OTA #2, whose Iabc is
controlled in the usual way.  This configuration is an alternative to either
a) using the linearizing diodes (always a bit dicey, in my experience) or b)
severely attenuating the input signal with a divider, to eliminate
distortion.

 

Well, I've simulated Mr. Sims' circuit, and it works just like a hot damn.
Now, I'm wondering: what is the specific advantage of doing it this way, as
opposed to simply reducing the signal down to just a few mV with a divider
and using a single OTA?  If anyone has any specific insight into this issue,
I'd be very interested to hear it before I commit to a specific design.

 

Something else I'd like to hear people's opinions about:  Just how useful is
logarithmic control of a VCA?  I notice that the Oakley VCAs are strictly
linear, and I'm wondering whether it would be worth enhancing the basic Sims
circuit for log/linear control.