MXR EF Debug: bottom of envelope drops out

[Dji]

I just got my MXR Envelope Filter (with range & reverse mods) working... mostly. When the sweep direction is downward, and I hit and hold a note, I get a dramatic wah, as expected, but at the bottom of the sweep I lose some or all sound. Usually I end up with a wet-sounding series of very low BOP noises until the strings stop sounding.

When the sweep direction is up, I get a decent set of sounds most of the time, but at some settings (high on the threshold, I think) I get high-pitched popping noises.

I'm using the Fairchild chips, and I've gone over all of my solder points. Some were weak during my first debugging pass, so I had to work them over a bit to get to where I am. I mis-ordered and got .005 caps instead of the .05s, so I had to go with the .047s I could get locally. However, I had one .05 cap, so the values are not consistent. (It's C3 on the GEOFEX schema.)

I'll continue my debug, but I welcome any suggestions. Thanks in advance.

[Mark Hammer]

Check some of the builder comments at Tonepad.  The problem may be solveable by tinkering with the values of R16/R14 and the reference voltage they set.  The filter is swept by varying the on-vs-off time (duty-cycle, mark/space ratio) of the CMOS switches, and by consequence, the effective resistance at the input to the invertor/filters.  It might be possible that the duty cycle is a little wonky and some portion of the filter's sweep time is spent with those switches spending too much time either on or off.  The fact that it occurs at the uppermost part of the upward sweep and the lowermost part of the downward swep is conspicuous.