Building my own Wah-Wah Pedal

[Mellotron Man]

I'm trying to put together a good quality Wah-Wah pedal, based on a Voltage Controlled Filter design out of a synthesiser. I've thought of using a traditional wah-wah design, but they all seem to use a 500mH inductor, and transistors. I prefer using op-amps, and I'm not keen on inductors. There used to be a dedicated I.C. called a CEM3320 (made by Curtis Electronics), but that's obsolete. There appears to be a replacement VCF, based on a CEM3320, made by someone in Latvia.
One thing I've noticed with synthesiser VCFs is the same terminology for different aspects of the resonance 'spike' at the cut-off frequency. I'm talking about Q. Some people say that Q is the bandwidth of the resonant peak, others say it refers to its amplitude. Who's right?
The design I'm after will be a Low-Pass Filter where the cut-off frequency is controlled by a foot pedal, (i.e. a pot), the amplitude and bandwidth of the resonance are also each controlled by pots, and the roll-off can be switched between 6-, 12- or 24dB/Octave. Am I being over-ambitious?

[eschertron]

One of my favorite foot-operated filters is the Maestro (semi?)parametric filter, made by Moog in the long-bygones. So over-ambitious? I don't think so.

IIRC correctly the Q is the relationship between bandwidth and amplitude, so a 'spikier' bandpass equals a higher Q, regardless of the value of either parameter.

[Helmholtz]

The half-level bandwith and the resonance peak height are related. Higher Q means both, higher peak and lower bandwidth.

[Mark Hammer]

1) Coolaudio now makes a CEM3320 replacement for all the "classic" analog synths that Behringer is now producing. Small Bear carries them.

2) I've also made my own inductor-based wah. A local surplus place was selling nice big ferrite-core inductors for a dollar so I bought a few. The come in at 900mh with a tap at around 800. Expecting it to be only suitable for bass, I went ahead and built the circuit, using a Crybaby design (and the 800mh tap). Much to my surprise, I got no sweep at all until I adjusted the emitter resistance and gain of Q1 and it sprung to life with a sweep range entirely suitable for guitar. Sounds like early Jeff Beck. The moral? 500-600mh IS optimal for the stock circuit, but the stock circuit can be tweaked to work well with other inductor values.

3) Sweepable state-variable filters can make very nice wahs. Indeed, it was an old studio trick to add wah in post-production by running a guitar track through a parametric EQ, cranking the boost, tweaking the Q, and sweeping the resonant frequency back and forth. One of the nice things about SVFs is that increasing resonance around the corner frequency of a lowpass filter provides the "focus" of a bandpass, without losing the "body" of the guitar as you sweep up and down. A true bandpass would get thinner-sounding as you sweep up.

4) SVFs can be done with op-amps and a dual-ganged pot. However, not all wah shells accommodate dual-ganged pots, and few, if any, dual-ganged pots arrive with a desirable taper, optimized for foot-control. In that respect, using voltage control of an OTA-based filter, like the CEM3320, can permit use of an optimal and robust foot-controlled pot.

5) Note that OTA-based filters are easily configurable for allpass, merely by rerouting one end of the caps. Four-pole resonant lowpass is a wonderful sound, but foot control of phasing is also a nice effect.