[sdiy] New SSM2164-based Log/Linear VCA design

David G. Dixon dixon at interchange.ubc.ca
Fri Jun 12 19:08:32 CEST 2009


I've finished a (slightly) new design for a log/linear switchable VCA based
on the SSM2164.  (It's so nice not having to mess around with offset
trimmers, linearizing diodes, Darlington buffers, and the inevitable
pseudo-linear response of LM13700!)  I will be laying out a pcb and a
2U-wide pseudo-MOTM-format front panel for a dual version shortly.
Interested parties please email me offline for the schematic.  I haven't
breadboarded it, but the simulations are perfect, and they never fail me!

My design is a slight variation on the design of Gallo-after-Irwin.  I did
not like the way Mr. Gallo added the reference offset voltage to the CV
summing amp (a trimmer off the positive supply to the CV summer's
non-inverting input), so I followed some excellent advice I got from John
Blacet as a result of my recent queries in his direction and simply added
the negative reference offset voltage directly to the CV summing node.  This
is a much more straightforward approach to me.

I also worked up simulation models of two versions: a 5V-reference version
and a 10V-reference version.  My envelope generator (a customized dual MFOS)
generates 10V envelopes, and I also wanted to feed in LFO and oscillator
outputs which are 10Vp-p.  However, even though I'm feeding in 10V signals
to control the VCA, I found that the 5V-reference version gave me greater
flexibility -- a more useful range of CV attenuation, and better overdriving
capability.  Also, the way I'm achieving the 5V reference (at both the CV
summing node and the current input pin of the linearizing 2164 VCA) is very
simple: a tripod of 100k resistors, one to -15V, one to ground, and the
third to the destination.  The two 100k resistors in parallel to ground (one
real, one virtual) look like 50k to the reference; hence, a 1:2 divider.
Also, the initial gain control is simply a 100k pot between ground and a
200k resistor to +15V into a follower, thus giving exactly 0 to 5V offset
range.  The CV attenuator also feeds into a follower, so both pots give
precisely linear responses with respect to their settings.

I should have the thing laid out, built and tested by the end of the weekend
(assuming my Small Bear shipment arrives in time).  Next, I'll be sampling,
holding, and making joyous noise in beautiful pastel colours!




More information about the Synth-diy mailing list