[sdiy] CV mixer trouble

Magnus Danielson magnus at rubidium.dyndns.org
Wed May 16 23:59:30 CEST 2012


On 05/16/2012 06:11 PM, David G Dixon wrote:
>> I'd say using the power supply rails might be OK if you ~own~
>> the rails (like you say you do, Tom...) Then you always know
>> what is connected and how stable and clean they are.
>>
>> The MOOG ROGUE uses its +/- 12V for VCO references...
>>
>> Other than that, its a really bad idea. I don't ever use the
>> rails of my own system, even to drive
>> the pots on the front panel :^)   Precision references are
>> not very expensive these days.
>
> I recently changed my thinking on this issue, opting for proper references,
> on VCOs at least.  I find it interesting that Paul Schreiber uses the rails
> as references in his MOTM VCOs (or, at least, he did), and he prides himself
> on his careful and accurate designs.  Of course, this is predicated on
> having a decent power supply, and not overtaxing it with too many modules.

You have to be careful since not only your CV source needs to be derived 
from it, but also your expo-converter and reset-comparator limit. As 
long as they scale from the same reference, all is fine, but if you let 
say only the expo sits of a reference level, but not your CV pot, then 
you are in trouble as they relative variations between them will become 
apparent. If they where from the same source, then it would first degree 
cancel, and only variations on that voltage would cause issues as the 
variation isn't completely cancelled as the integrator action 
phase-shifts the variations.

VCOs is a bit tricky in that sense. :)

Cheers,
Magnus



More information about the Synth-diy mailing list