Weird problem on MXR CAE MC-401 (Booster/Line Driver)

[Luqqas]

Hello there!

I've built a DIY MC-401 (CAE Boost/Line Driver) following this schem
.

Ommitted parts are in red. Just ommitted this to use another type of LED switching because I couldn't find J117 FET's. Instead, use a DPDT with millenium bypass for On/Off LED.

Here is the thing: the effect works perfectly. But when it is turned off, a volume swell appears (going from low to high, like a fade in) .

This only happens when the potentiometer is from 8 o'clock to the end. The "fade in" increases as you turn the pot clockwise.

If the pot is set to 7 o'clock (we can call it unity gain or "no boost") there is no fade in. So, if the pot is in fully counterclockwise when the effect is turned off, the signal level remains constant as should be.

With the pot in any other position, a "fade in" shows up.

Any ideas for debugging this?

Greetings
Luqqas.

[Luqqas]

Anyone??

Completely stunned about this issue.

[ubersam]

I had a similar issue with a different circuit. It turned out that when I switched from 3pdt bypass to a relay bypass, I designed the layout wrong. I left out grounding the input when on bypass. Could that be your issue? I can't recall, but, does the millenium bypass scheme ground the input when on bypass?