Modified big muff problem - strange interaction with other pedals

[rats_eyes]

Here is the way I've implemented my gain switch.

Caps seem correct in polarity and placement. I'm not getting any heat from anything either. Hmmmm. I'll keep trying things out, see if I can get there.

Next time someone asks me for a blend control,  I might just say no!

[GibsonGM]

  Ha, I don't blame you!  But I feel it's something simpler than that.  If it was a phase issue, it would be like throwing the 'phase' switch on a Strat wired for that...instant honk and squeak.   If it builds up over time, really suggest a cap problem.   

When the gain drop occurs, does it STAY that way?  Or, if you stop playing for a bit, does the gain go back to normal?   If it STAYS, perhaps you can get some transistor voltages before and during?   

Can you tap into "receive" and see if this happens using only the BMP?

What are you powering all this with - battery?

[rats_eyes]

I'm with you there. I don't think this is phase, as the signal isn't inverted in either path of the blend circuit.

My feeling is that it is the blend circuit. I'm quite comfortable with fuzz circuits generally and have never had a problem like this occur. Also, this exact circuit (this actual build) worked fine before I aded the blend. I'll admit, I don't know a great deal about buffers, so my ignorance in that area could be overlooking something really simple.

The drop in gain dissapears if I leave the unit alone for a while, then the unit is back to normal again. Power is from a 9v wallwart, no battery. I use the same wallwart to test all of my pedals at home.

I've resoldered the blend circuit, just in case something silly was happening there. I'll see if the problem continues. It's such a pain in the arse to test, as the problem occurs quite unpredictably in the first place. If I can get it to misbehave again, I'll get the probe on it again. Fingers crossed!

[karbomusic]

Can you test to see if by chance the current draw goes *up* when gain goes down? If it does, it could be an oscillation that you can't hear due to something like capacitive loading. If that's the case the oscillation will likely cause the current to rise even though the volume decreases. I had this happen once and the fix was to add a 50 or 100R inline before the cap on the output of the output opamp.

[Quackzed]

maybe the blend circuit isn't referenced to ground? it could be the muff output doesnt have a resistance to ground 'after' its output cap, and therefore the blend signal is somehow floating, could take a while till it floats up  enough to cause issues... i'd think that normally the muff would be referenced to ground at the output and input, as well as the blend circuit (is it active or passive? if passive maybe it isnt but your guitar would be :/ ) easy to try, just stick a 1M resistor from the blend signal to ground... if its an active powered blend CIRCUIT then a 1M from input signal to ground AND 1M from OUTPUT signal to ground... but it sounds like your describing a non ground referenced signal that slooowwwly floats up into the pita'osphere...

[GibsonGM]

^  Both good ideas.  Yes, ultrasonic oscillation that you can't hear may cause that drop in gain...no access to a scope?  You might be able to find it with that.  Or just put the R in like Karbo suggested!  And make sure it's referenced.

[rats_eyes]

I resoldered the (active) blend circuit and replaced some resistors which were not measuring correctly on my meter. Now the unit seems to be behaving properly, for now at least.

I am concerned that it will start to play up as soon as I give it back,  as the problem has been intermittent. I think I'll fit some 1m resistors as suggested. I assume that these go on the input and output connections to my switch, rather than around the send and receive connections to my fuzz circuit?

[GibsonGM]

Do you have a schematic of that thing, Rat?   Then it would be simple to see how it works.

For reference, you can put a 1M on all 4 of them if you want, and it should not cause any problems.  I'd do 'em all just for grins and to be sure...  a 1 meg "short" between the + and - of  a battery only allows 9 one-millionths of an ampere to flow (9V/1,000,000 ohms)...so it's very transparent - the circuit doesn't know it's there in terms of power flow, but for reference purposes (VOLTAGE rather than current), it does the job.    I feel that doing the send/receive also will make sure those are referenced, too...caveat - if there already IS a 1M there, adding another in parallel will make the value 500k, thereby 'worsening' the 'transparency'...so, looking closely is a good idea.   
Or - take your meter, read the input jack tip to ground for resistance...should be near 1M.
In fact, I am curious what you will read if you try that...gotta hold it steady til any capacitances charge (up to 1minute)...
If you trace the input to the board, you may well encounter a resistor to ground - take a look...

Guess you have to play that sucka for hours now, to be sure it's fixed, ha ha!