[sdiy] 910 pF caps (Buchla 292C/266-3)

[Grant Richter]

> What I'm wondering is, if I could substitute the 910 pF caps with 1000 pF in
> any of the projects I'm working on.
> 

Typical capacitor tolerance is 10%, so the original Buchla modules shipped
with caps that varied from (910 x 0.1 = 91pF, 910 - 91 = 819, 910 + 91 =
1001) So a 1000 pF cap is actually within original design tolerance.

> The first one is Bucha's Source of Uncertainty, in the Fluctuating Random
> Voltages module:
> 
> http://www.musicsynthesizer.com/Buchla/266-3.jpg
> 
> I'm not quite sure what C11A and 12A are doing (around IC 3 near module
> inputs).  The second couples IC 5 to 6 (C17A and 18A).  Seems like that one
> is less critical.

C11A swamps the op-amp so it does not go into oscillation. Very
non-critical. C12A differentiates the pulse for the sample and hold formed
by the 4066 and storage cap C12. It sets the sample aperture.

R36A, R37, R38, R38A, R39 and R40 don't actually do anything. Perhaps they
are there to bias or current limit bipolar op-amp input stages. But,
according to a SPICE model, they have no significant effect on frequency
response or time constant. Two polarized 10 uF in series form a
non-polarized 4 uF cap.

See

http://www.musicsynthesizer.com/WoggleBug/WoggleBug.html

for a block diagram of the circuit. The filter is the same as the Lopass
Gate with a much lower corner frequency (1 uF input, 4 uF tap vs. 1000 pF
input and 4700 pF tap).

Bear in mind that Mr. Buchla never designed a building block twice.

R32 ups the current when the 3.9 volt zener conducts and controls your top
LFO frequency. R34 and 35 keep the whole thing from shutting down when the
Vactrol goes to 100 Megohms in the dark.

> 
> The other project is an updated version of Buchla's LoPass Gate (292).
> 
> http://www.simple-answer.com/lopass-gate.jpg
> 
> C5 Appears to be doing some audio filtering, so my guess is that it's more
> critical  to have an accurate value there.  I suspect that if I change the
> 4.7Meg resistor, I can compensate for a different value.
> 

The 4.7 meg resistor is to supply bias current to the FET buffer when the
Vactrol resistance goes up to 100 megohms (dark resistance). It doesn't do
any real filtering.

The critical capacitor is the 0.0047 feedback resistor on the tap. That sets
the Q at unity gain feedback.