I read that entire thread. Interesting. I originally thought a lot of the same things about the noise coming from the 644. It is all in the 3sk30. I scoped it and the static alone with no input signal was like 1 volt peak to peak. Then I shorted the output to that first stage and measured the noise and it was negligable.
Also, these effects hail from 1974. Apparently a large cache of these and other Walco effects were found in a Warehouse in New Jersey. They weren't stored in the best conditions. I had an opportunity to buy around 2500 Walco units, but the price wasn't right.
Modifying the time constant would be simple if there was room on the board for additional parts or room in the box for a pot. I also thought about using a FET to instead of the 644 rather than the bipolar which seems overly sensitive. And due to the poor recovery, if you hit the thing hard enough, you lose all output for 2 seconds or so.
I wonder if this effect would be more useable if the volume on the guitar were turned way down. Most guitars in the 70's didn't have the same kind of output that modern guitars have now. On top of that, most Japanese guitars have very low outputs. Think Teisco Del Ray.
Anyway, replacing the fet does wonders for the noise, but doesn't make the effect any more practical. One thing I did notice though, is that perhaps if you replace the fet and turn the volume way down, you might be able to get something and if the signal loading overly darkens your sound, there is a drilled hole on the board of one of the gain stages to fit a small value capacitor as an emitter bypass which could be used to compensate for the high end loss.
I think it is quite obvious that I've spent a bit too much time with these units.
-Zach Omega
My take on the chord & note sustainer:
http://www.diystompboxes.com/smfforum/index.php?topic=48977.msg367410#msg367410
I breadboarded the sound go round. Nice effect but the LFO was too noisy. I'm going to try the vox repeat percussion instead.