PAIA Hex VCA circuit

[Mark Smart]

Hi everyone.

After forgetting about it for a year and a half, I have dived back into my
GR-300 guitar synth hacking project. I have been messing with this circuit
invented by PAIA which uses a 4049 CMOS chip to create six VCA's which mix
into one output  (the output comes out the ground pin!). What I'm using
this for is to mix together pulse and square waves derived from the
GR-300's internal sawtooth waveform. My saw-to-pulse converter and octave
dividers create three pulse waves which oscillate between 0 and +15 volts.

This circuit mostly works execpt that there is noticeable leakage between
channels which is especially bad at low volumes. I have three
signals, the pulse wave, the /2 square wave, and the /4 square wave. When
you turn on the control voltage for the /4 square wave, it comes on, but
with a noticeable amount of the /2 and pulse signals leaking in. I would
appreciate the advice of anyone who is familiar with this PAIA hex VCA
circuit. I could also post a schematic at some point maybe. I tried
lowering the 4.7 Meg resistors to 2.2 Meg, and this helped some, but did
not eliminate the problem. My biggest problem in debugging the circuit is
that I really don't understand how it works very well. The description
which came with it does not go into that much detail. It does not specify a
limit on CV or audio voltage ranges. The basic idea of the circuit is that
the FET's in the CMOS inverter gate are used as voltage-controlled
resistors, and somehow the audio inputs mix inside the chip and come out
the ground pin.

I would appreciate any advice. Also thanks to the people on the list who
have helped me in the past in my sporadic forays into synth-building. 


************************************************
*     Mark Smart                               *
*     Electronics Engineer                     *
*     And Musician                             *
*     smart at medusa.nn.com                      *
************************************************