"Uncanny Valley" - analog bitwise wave manglor

[Skruffyhound]

Sorry to bother, but I need to subscribe to this thread to get updates.

[earthtonesaudio]

Bleh.  Please disregard my last post.  It turns out with the proper resistor values this schem DOES work.  

On the breadboard I have (verified and working):
R1=10k
R2=220k
R3=56k
R4=1M
Also R5=1M and C1=0.1u

These values give a full-scale output for ~60mV p-p input.


I also had the chance to use the ABC outputs to control a CD4051, which in turn processed the same source audio.  I put the guitar into an inverting op-amp with a fixed input resistor, and several feedback resistors controlled by the 4051.  I used the op-amp output to drive the comparator inputs.  Definitely not a subtle effect; the output is a sort of stepped waveform with lots of harmonics:


On the other hand, by taking the ADC out of the feedback loop, a different sort of response is obtained:

For this configuration, the output looks something like this:

[Taylor]

Beautiful waveform.

This is awesome. At this point I can take no credit for what you're coming up with, but I'm glad I get to watch!

[Hides-His-Eyes]

Holy smoly that's a clever use of a 4051.

[digi2t]

I would really love to hear a sound clip of this. Looks like a real audio mangler! I like!

[Taylor]

Hey Alex, is the waveform there a simulation or scoped from a real hardware unit?

It seems like, basically, if everything was perfect, you'd have a regular wave right there, right? So the aliasing going on there would seem to be a result of the mismatch of the resistors due to tolerances if I'm understanding things correctly. In any case, the next thing I'd like to try would be pots in place of the Rf resistors. Should give some pretty nice control over the distortion, and a filter on the end would probably be good, so we can remove the aliasing due to the clock and get something more like an analog lookup table distortion - basically what I was aiming for in the first place. 

[earthtonesaudio]

The waveform drawing is not a scope shot, just an "artist's rendering" in MS Paint of what the scope looked like when I ran guitar through the circuit in the second schematic.  I made sure none of the resistors were matched, and you'll note that the bottom half of the waveform is distorted differently from the top half; that's due to using one set of values for Rf0-Rf3 and a different set for Rf4-Rf7.

I don't know when I'll get a chance to do a recording, but the sound is not as extreme as you might think, after seeing the waveform.  It sounds sort of like a cross between plain old harmonic distortion and the way an FM radio station sounds with poor reception.

[Taylor]

ETA, I assume this was drawn as a bipolar circuit for simplicity. In a single-sided 9v supply setup, should the AGNDs in the ADC go to 4.5v, or ground?

[earthtonesaudio]

AGND should be 1/2 the total supply voltage.  So for a 9V supply yes, it would be 4.5V.