[sdiy] Buchla's funky 4 pole

[mark verbos]

Hi,

I recently had the opportunity to work with a Buchla 191 Sharp Cuttoff 
filter. This is a 4 pole lopass and Hipass in one module, with 
facilities to run the two as separate filters or as one bandpass. The 
Hipass section in particular sounds great, however I was not able to get 
it to self oscillate, even with TONS of external feedback. My guess is 
that it is a matching issue, but that is only semi-related to my question.

This configuration is unlike any other filter I've seen before. It is 
totally discrete, each of four cascaded integrator stages is made up of 
an NPN pair acting like diodes for the voltage control and 2 gain 
stages, the first a FET/NPN follower and the second a discrete OP AMP 
made from an NPN pair and a PNP (quite common in the Buchla 100).

I've drawn up one stage to check out here.
http://www.simple-answer.com/191/Buchla-191-lopass-stage.jpg

the  functional block diagram is here.
http://www.simple-answer.com/191/191-lopass-equiv.gif

the control ports (the top and bottom of the "diodes") are driven by 
differential voltages.

I'm thinking of cloning this thing, but I'm wondering if it is necessary 
to have all those trim pots. Particularly the ones to trim the 
transistor pair connected as diodes. I mean, that pot runs thru a 22M 
resistor! Were transistors MUCH more unmatched within a matched NPN pair 
in the 60s, so much that they would no longer need trimming these days? 
It is a TD101, which I have never heard of and can't find any info on. 
Should I try it with plain ol' diodes? Did anyone else use diodes like 
this? I know the Steiner filter uses diodes, but it's quite different. 
Also, the Korg MS-50, but that's a diode ring.


I think this thing is cool. but just because I have never seen something 
before, does not mean that it is innovative. However, I figured you guys 
would get a kick out of seeing this.

Mark