[sdiy] 2164 Expo VCO tracking

[David G. Dixon]

...I should also mention that, for some reason, when I use the "proper"
reset timing component values (say, 15k and 22pF) and tune the VCO for
proper tracking below 440Hz, the octaves above 440Hz are too large (880 is
about 882, 1760 is about 1765, 3620 is about 3630), even with the HF trim
all the way off.  In other words, if I have 1V/octave below 440Hz, I have
slightly less above 440Hz.

Hence, I have made the RC components significantly larger than they need to
be (22k and 33pF) and added quite a large pseudo-Franco resistor (1k) and
then used the HF trim to pull up the HF values.  These large RC components
are partly to blame for the very poor tracking above 7kHz.  I can make the
tracking better with the HF trim, but then I lose accuracy in the 2-4kHz
range.  Also, the remaining overshoot at 880Hz reflects my inability to
eliminate this problem even with the larger RC components.

These are all artifacts of using the 2164 as a current source, and I'm still
at a bit of a loss as to how to fix it.  The things I've done (feeding the
2164 directly to the virtual ground of the integrator but separating this
point from the cap and JFET switch with a pseudo-Franco resistor, filtering
the JFET gate with a 100pF cap to ground, leaving off the real Franco
resistor) have all helped somewhat, but certainly not eliminated the
problem.

I'll keep on plugging away at it...