CMOS inductorless wah-Rough draft

[earthtonesaudio]

Hey all, I've been experimenting with various notch filters for a wah pedal, including Twin-T, Bridged-T, and the "bridged differentiator."  Here's my rough idea for a wah based on either the 4049 or 4069 unbuffered inverter chip.  I'd like to know what you think of this:



Some notes:
-IC1b contains the "bridged differentiator" filter.  This could be a twin- or bridged-T just as easily, but I drew up this one to be unique 
-IC1a and IC1d are there to bring the signal back to unity or just a little hotter.  Got the idea from the ROG Mr. EQ input section.
-Capacitors C1, C2, and C3 can all be equal, but mismatching C2 and C3 gives the most variation in frequency range.
-SW1 allows comparing either standard or bootstrapped-for-higher-Q versions.  High Q often leads to oscillations with this setup, but there's a sweet spot in there somewhere...

Any input on helping to make this circuit better would be greatly appreciated!

[R.G.]

It's an interesting design. The only issue that strikes me is that the Q of the filter  changes as you change the tuning resistors. It gets higher Q at higher frequency, and I think that may be where your oscillations come from.

I think you also want to use a coupling capacitor to break the DC connection to ground on your Q pot, both to the output amplifier and the Q buffer.

[earthtonesaudio]

Thanks RG, you are right on the money about the feedback occurring at higher frequencies of the sweep.  Does this look better?

I added C8 and C9:


I'll be breadboarding the changes tonight, if I don't get sidetracked by the forum's "search" function 

[brett]

Very cool.   
I suggest rolling off with fc around 10kHz to stop oscillations.

[earthtonesaudio]

Thanks!  Now I guess it's time to start plugging in caps and resistors to see what happens. 

[ACP]

Don't you need to bias those gain stages?

[earthtonesaudio]

Don't you need to bias those gain stages?

I did unbiased gain stages for simplicity, but biased gain stages would work also.  See Runoffgroove's Mr. EQ, Double D, and others for examples of unbiased (I think) inverter-based gain stages.  On the other hand, circuits like Tim Escobedo's CMOS Boost do use biasing.  I'm not sure what the differences are.  Seems like biasing allows for more gain, but I could be mistaken.

Thanks for the feedback. 

[gez]

All the above inverters, except 1C1c, are biased already.  There's resistance from output to input, which is all that's required.  IC1c needs said resitor (and an input resistor perhaps?)

The circuit will have lowish input impedance as shown, so you might find things a little muddy.