[sdiy] Buchla 259 timbre circuit and op amp compensation

[Aaron Lanterman]

On May 1, 2008, at 4:54 PM, anthony wrote:

> Hey this has cleared up a sticking point for me! Just try finding  
> 3160's anywhere. What about 3140's? They aren't that much different  
> from 3130's besides having different supply voltage limitations,  
> right?

The interesting thing about the 3160 and 3130 is they have an output  
stage that can swing "rail-to-rail." This makes the Buchla 259 more  
predictable in that you can calculate what the deadband edges of his  
deadband circuits would be based on the +/- 6V supplied to the op  
amps. That circuit is based on the output hitting a rail and not being  
able go further, so the "golden op amp" rules get broken. (I don't  
think the 3140 has the same property but I haven't looked in detail.)

I think as long as you include the compensation caps, you can replace  
a 3160 with a 3130 no problem (or even mix and match).

One my student cleaned up his wiring a lot of the "goo" disappeared (I  
think there's some "waffle board" gremlins going on, as Harry Bissell  
calls them.)

His 259 nonlinearity section wasn't 100% working - the transfer  
function, instead of a set of teeth, looks more like a set of teeth  
with some higher than others and some knocked out. After many hours of  
debugging we can't find the problem... but we hooked it up to  
speakers, and listened to it, and thought it sounded really  
interesting so I told Alfredo to go ahead and build it in solder. ;)

- Aaron