Switched capacitor compressor idea... not working.

[earthtonesaudio]

This is supposed to be a compressor.  Ignoring the switched-cap portion, the audio comes in through a pair of inverting op-amps, and into a voltage-controlled PWM.  The output of the PWM is such that with a large positive (or negative) input signal, the frequency drops.  The flip-flop removes the duty cycle information and outputs a square wave based on the frequency of the PWM. 
This then controls a switched-capacitor-based resistor that determines the gain of the second op-amp.  For a large input signal, the frequency drops, which makes the resistor appear larger, therefore it should be compressing.

I have this on the breadboard and I'm getting audio out, but no compression.  I think maybe I don't have enough range of frequency coming out of the flip-flop to get the results I want (but this is difficult to measure with a scope because the frequency changes at the same speed as the audio input).



For the Schmitt trigger oscillator, I was using a CD4093 but I ran out of breadboard room so I used an AD744 op-amp wired as a square wave oscillator and put the signal in through a resistor to the (+) input, and the performance is about the same.

For the Switched capacitor resistor I'm using one section of a CD4053 SPDT switch.  The op-amps are from a RC4558 and the flip-flop is a 4013.

The thing that's puzzling me is that the "simulated resistor" should have a value determined by 1/(C*frequency), but even with varying the frequency, and swapping caps, I can't seem to get it to do what I want.

Oh yeah, I did try placing a cap (tried several values) across the feedback resistor in the second op-amp to keep the clock from coupling back in, but it didn't make any difference.

I'm sort of at a loss as to what to try next... any info appreciated. 

[cpm]

this is something completely new to me...

i would test the switched cap resistance on a voltage divider alone, feeding pwm with some known voltages, to see if its tracking...
from my ingenuity, thats where i would start

[earthtonesaudio]

Good idea, I'll give it a try.

[dschwartz]

i´m missing some kind of "detector" circuit feeding the oscillator, maybe the high frequency voltage changing confuses the PWM..also, with a detector you could add attack and decay controls...currently you have no attack or decay.....

[earthtonesaudio]

Hi Daniel,

The compression action is implemented similarly to Craig Anderton's compressor in that there is no provision for adjusting the attack/decay/sustain/release parameters.  Rather, the signal envelope almost immediately affects the gain control element, and the "detector" is part of the control element's feedback loop.

I also tried adding a capacitor (I tried various values) in the feedback loop of the second op-amp (to remove high-frequency switching transients) but this did not produce any audible changes.  The switching frequencies I've been using are quite high so parasitic breadboard capacitance presumably takes care of a lot of it automatically.

[dschwartz]

HMM. I don´t know then...

why use a switched cap and not a single 4066 switch? or a fet?

[earthtonesaudio]

For fun!