[sdiy] Vactrols subbed for dual-gang pot

[Mike Pepper]

> On Jul 11, 2010, at 4:46 PM, cheater cheater wrote:
>
> > Yeah, I know all about non-linearities - but you can usually get some
> > sort of general idea anyways. I guessed low-pass-ish, but I wonder if
> > it's more 12dB/Oct or 24dB/Oct or ... :)
>
> Well, you need to be careful to specify exactly what it is you're talking
about modeling. If you're talking about the audio path of a typical VCF,
that's one thing, and makes it easy to talk about 12db/Oct or 24db/Oct or
whatever. If you're talking about the *control* mechanism of such a beast -
say a Buchla LPG - that would be trickier to characterize.
>
> You can, it seems, do tricks to effectively speed up the vactrol response
(roughly speaking), if you use a couple of vactrols; the amplitude
modulation circuit of the Buchla 259 and the ring mod circuit of the Music
Easel both use this trick. I can sort of follow the 259 circuit but the
Easel circuit makes my head spin. Grant Richter probably has the best
understanding of how the Easel circuit works. It's all based on the sort of
datasheet trick that Harry Bissell pointed out a few days ago - the idea is
to use one vactrol in a feedback loop do basically "undo" the non-ideal
behaviors of another vactrol.
>
> - Aaron
>

The non linearity of the vactrol looks rather like slew limiting to me,
though the scale is a bit square law - ish. there is a hidden problem with
CdS photocells in that their attack/decay characteristics often don't match:
a sudden pulse of light will drop the resistance fast, but it will take
longer to return to the high resistance state. There may be another effect
at play too, in that there may be a sensitivity to the length of time the
cell has been in the low resistance state having an effect on the time to
return to the high resistance state. A bit of experimenting with a
photocell/LED  or a Vactrol, a pulse generator and an oscilloscope might be
interesting

At the end of the app notes here:

http://www.perkinelmer.com/CMSResources/Images/44-3429APP_AnalogOpticalIsolatorsAudioApps.pdf

the two element lineariser is shown, it relies on the fact that the elements
will match within 10%, and the linearity within 5%. The same trick can be
used with photodiode and FET isolators: I used a photodiode one in a
heartrate monitor years ago, to isolate the battery powered front end from
the logging/scope system. A second Isolator, reversed, provided linearity
feedback. After a few tests, it seemed that the matching between components
was good enough without selection.

||\/||ike