Now with a day or two under it's (read: my) belt, it seems I was
full of it on my initial commentary on the A149 on a couple of
fronts:

It seems that the top and bottom half are running off the same
random source. When you trigger them with the same signal
they do track each other. Not that they give the same voltage
result mind you due to the basic conventions of the module, but
they definately are NOT going in their own directions.

Secondly, this module triggers on both rising and falling edge of
a trigger signal, not just the falling edge only as I reported
before.

I have come to the conclusion that this is a bit of a pain. Here's
why:

If you want to sync envolopes to the 149 you need to trigger them
from the same source. If you trigger them both off a square, you
will get two voltages out of the 149 per envelope.

The only workaround I've found for this is to use a trigger source
that has both square AND inverted (ramping downward)
sawtooth outputs. You use the square for the EG and the
inverted ramp for the 149. I guess it needs a hard elbow to
trigger from the falling edge.

The BIG problem with this is depending on your VCO, the square
and saw outs aren't always in phase aligned* The workaround
for this is to offset the riding edge of the square using the PWM
pot. Once it's set up, it all work perfectly, but it takes a bit of
juggling to get it there.

* - By phase aligned, I mean the saws and squares of some
VCOs are by design 180 degrees out of phase to one another.

It depends on two things: the manner of waveshape conversion
incorporated, and the type of waveshape generated at the
devices' core. Most being sawtooth core, some manufacturers
use the reset pulse in which to generate the square, and this is
OK as the reset causes the saw to charg up again so they are
cool. However, in triangle core VCOs (Buchla and Aries being
two of them) , instead of a retrigger PULSE to the flyback switch,
it generates a square wave who's trailing edge is exactly at the
peak of the triangle. Add to the mix, the sawtooth waveform from
triangle core VCOs usually use both the triangle and square
wave signals through an anlog switch for sawtooth conversion
and without going into boring details, by the nature these saws
are almost always rising only and 180 out of phase with the
square wave.

As an unrelated yet possibly interesting sidebar, this is why Don
Buchla used his somewhat unorthodox method of square
generation via an adaptation of the sine shaper found in
Electronotes volume 1 (even though his use preceeded that
publishing). He doesn't use the retrigger signal as the square,
he generates another one. In doing so, his VC waveshaping
(square to sine) is in fact put back into phase so that that
function doesn't go all wack in the middle (passing a half sine/
haf square waveshape).

Man, this letter went totally OFF topic, but I thought some of you
might find it interesting.

- P