[sdiy] Vactrols subbed for dual-gang pot

[lanterma at ece.gatech.edu]

On Jul 11, 2010, at 10:33 PM, cheater cheater wrote:

> well, given that it's 'sort of' like a low-pass filter, and feedback
> around a low pass filter adds resonance..

Well, this depends on the filter topology. The negative feedback of a 4-pole cascade creates a resonance bump because of the 45 degree phase shift of each phase around the cutoff frequency of each 1-pole stage. Other topologies will give different effects.

In a state variable filter, the resonance amount is controlled by the amount of the first feedback loop, and for a SVF more feedback actually results in less Q; the feedback dampens what would otherwise be an oscillator.

> With normal LPF's if you have two of them in parallel, one with 0
> resonance and one with some good amount of resonance, the bump in the
> resonating one will take the content at the resonant frequency and
> give it, say, 1.5x or 2x gain. If you mix the outputs together then
> the pass-band will still be 1x gain but the corner frequency will
> still be passed through at nice enough gain.

Uhm, I'd hesitate to make such a statement *in general* without knowing more about the phase responses of the individual filters.

> So maybe that's what he's doing with two vactrols and just giving one feedback.

Uhm... not really. It's a linearization trick. Horowitz and Hill is a good reference on this.

- Aaron