[sdiy] Secrets of Dan Brown's "The Buchla Code" Revealed

[mark verbos]

Grant Richter wrote:


> I spent a great deal of time simulating that circuit under Spice.
> 
> The only behavior I could match to the real world was as described.
> 
> If you consider the switch as a center off, it makes more sense.
> I know there was discussion of about a special version of the switch,  
> this may have been used in some version.
> Based on ohm meter readings of the switches in the actual modules  they 
> are just three position center off DPDT.
> 
> So center position has no terminals connected.
> 
> Amplitude side switches in resistive divider.
> Frequency side switches in the 4.7 nF, and a 3 dB attenuator at the  
> input stage for the earliest design.

The whole reason I got into this with Peter a while back is that in the 
292b (the one with the hardware switch) the resistor is switched in to 
attenuate the signal.  In the 292c the analog switch chip switches a 
resistor in to amplify the signal. This amplification of 3 dB only 
happens in the "Gate" position. So if only one position engages a 
resistor, that means that in the 292B the "both" position is 3dB hotter 
than in the 292C. Is it? That seems wierd to me. And the schematic on 
hyperreal.org (which is a reverse engineered one) shows the switch as an 
on-on-on version. That's not to say that Don didn't decide to change it 
part way through the product run......

> 
> See if you can find the 192 schematic, it is pretty much the same  thing 
> but used two VTL2C3 metal can Vactrols.

And it's all transistors and only the "lopass" mode. (no switches)

> 
> Also part of the filter for the Korg Polyphonic Ensemble keyboard was  
> lifted whole heartedly from the Buchla design (pretty sure the Korg  
> drawings have latter dates).

That I'll have to check out.

Mark