MXR Style Envelope (Vowel) Filter

[amptramp]

It looks like the PIC would be the limitation in these circuits.  With a 74HC4066, you could get adequate switching response but you might need random logic circuitry to achieve the necessary speed.

[ElectricDruid]

It looks like the PIC would be the limitation in these circuits.  With a 74HC4066, you could get adequate switching response but you might need random logic circuitry to achieve the necessary speed.

Sorry, what would the limitation be? (or "which?"!) I'm not following you.

[amptramp]

The control input frequency limit on a 74HC4066 is 30 MHz.  I have never designed with a PIC before but I question whether it could drive the 74HC4066 at this speed.  If it can set up a 30 MHz PWM it would work but I have not seen a PIC used for switching at this kind of speed before.

[ElectricDruid]

The control input frequency limit on a 74HC4066 is 30 MHz.  I have never designed with a PIC before but I question whether it could drive the 74HC4066 at this speed.  If it can set up a 30 MHz PWM it would work but I have not seen a PIC used for switching at this kind of speed before.

Just because we *can* go at 30MHz with the switches doesn't mean we *need* to go at 30MHz!
As was mentioned earlier on, most PWM schemes for stuff like this run at 10's of KHz - not 100's and certainly not MHz. 30-50KHz is pretty typical.

Say we went for 62.5KHz. That's comfortably above the audio band, so we can filter the audio and remove any residual switching noise. A PIC can produce a 9-bit PWM output at such a frequency, which isn't too bad although you'd hear zippering if you spread those 512 steps out too far. Maybe an AVR can do better, I dunno. I'm not so familiar with those chips. Or you use a PIC with an NCO module and use a PDM output, in which case you could have 14-bit/2MHz output like the StompLFO. The point is only that the switch needs to be fast enough to handle that - and with a 30MHz top switching speed, the 74HC4066 would do fine, although the "old fashioned" version goes fast enough too.

[Mr. Lime]

An interesting project can be found here:
Aeronometron Project

It's a formant filter as well which uses 4066 switches as variable resistors and an ATTiny85 MCU as oscillator that directly drives the 4066.

What's even more interesting, is the LM2907N frequency to voltage converter which offers CV for the microcontroller and should be usable for the STOMPLFO too!

[ElectricDruid]

An interesting project can be found here:
Aeronometron Project

It's a formant filter as well which uses 4066 switches as variable resistors and an ATTiny85 MCU as oscillator that directly drives the 4066.

That answers a few of my queries about driving the 4066 run on 9V, for sure. The odd thing about that design is that one pair of switches are driven with a 0-5V control signal, while the other set are driven with a 4-9V signal (or as close as the op-amp gets)! This happens because the op-amp inverter is biased around 4.5V Vref, so a 0-5V signal is considerably off-centre.
Since it apparently works fine, it clearly doesn't matter too much. It looks like the 4066 is not fussy.

Using the F-to-V chip to derive another control signal from the input is a nice touch too, I agree.