Fender Champ 12 red-knob channel switching problem

[Mr Johnny Birchwood]

Hey guys, I'm working on a Champ 12 with a confusing issue. The channel 2 gain pot is acting all weird. Owner of the amp had previously installed a torres pot mod with a 500k in the gain pot and a 100k in the ch2 volume pot. Both are stock 1m tapered pots. The installation was pretty messy and the gain knob didn't work right so I installed new 1m pots. The volume pot works just as it should on the gain channel but the gain pot doesn't do much of anything. The fender part number on the schematic is the same so I know it's the right pot(because the volume pot works.) The gain pot makes absolutely no difference through the first 90% of it's travel, but when it approaches max(where gain should be highest) it cuts the signal significantly. It is basically either on or off. I have tested all of the LDR's in the channel switching and they all function as they should. Pulled C10 and tested it, it's good. Tested with a known good cap in C10 and no change. I've been staring at the schematic for an hour now and I can't make any sense of it. The circuit board is pretty fragile, so I am reluctant to just start pulling parts and testing them at random. My other thought was a broken trace somewhere in the circuit but I can't make any sense of why the pot would cut the signal at the top end of it's range. Anybody have an idea where I should be looking or what I should be looking for? Thanx, schematic attached

[52 Bill]

All of the controls here work by shunting signal to ground and are turned on and off by the optos. Is it possible that the gain control is wired in backwards?

[DrGonz78]

The gain pot makes absolutely no difference through the first 90% of it's travel, but when it approaches max(where gain should be highest) it cuts the signal significantly.

To me this describes a no-load pot where they scraped away the graphite at the 9 to 10 position on the pot. Is this the new pot??

[J M Fahey]

You wired the volume pot right and the gain pot backwards or by mistake you put a linear pot there.
Or you got the no load type.
But it's easy to solve
The 1M log originals work reasonably smooth.
Not as smooth as if they were regular "voltage divider" connected (the classic way), but the log/audio curve helps.
They do this to simplify design, it's not a real 2 channel preamp but a single (high gain) one, with switchable attenuation.
But it works quite well, I like it.
It deserves a much better speaker than the light stock one, not for power but for better response.
A customer of mine put a classic Celestion Greenback there with killer results.
Stands only 20w RMS but there it's fine.

[g1]

As far as I know these pots are board mounted so there is no way to wire a pot backwards.
I'm not sure how you tested the LDR's but I still suspect them. They can measure properly out of circuit but fail when under voltage or after on awhile. Suggest you try swapping the two LDR's around and see if the symptoms change.

[Mr Johnny Birchwood]

I tracked the problem down to a broken trace that was shorting the wiper to the high side of the pot. It was likely caused by previous installation of a Torres mod. The overdrive channel is barely usable as designed. I can see why the 500k and lower pots are popular in these spots. This PCB is so fragile that I'm not gonna remove the pots and replace with smaller ones, just run some resistance in parallel with em.

Customer supplied a Hammond 125ESE output tranny and a Marshall labeled Greenback G12C. Channel 1 sounds pretty damn good now. I Installed a 5751 in V1 and re-biased the preamp tubes(similar to the Torres mod but with my own values subbed in.) Deleted C6 to tame some of the ultra-bright harshness of this thing. Pretty happy with channel 1 and now the task is to make a decent overdrive channel out of this mess...

[Mr Johnny Birchwood]

I should add that I was pretty wary of the voltage doubler B+ supply, but new 47uf F+T PS caps really punched it up. With the new caps, speaker, and output tranny this amp has a LOT more clean drive at all frequencies, but particularly in the low end. Still thinking about how to attack the gain channel. I'm thinking of dropping the pot resistance of both pots to 500k with a 1m resistor in parallel and maybe deleting C14. C10 could go up to .22uf too. I'm also looking at deleting R20 but I'm not sure if this would make the amp unstable. The nicest thing I could say about the channel switching scheme with the LDRs is that it is 'unconventional.'

Considering rewiring the power tube socket to run an EL34. The rats nest under the cathode of the power tube makes me a little wary of the bias of the different tubes. I guess I could move R31 off the PCB and make a bias switch with different resistors. I'm a little hesitant to attempt this due to all the junk that runs off the current from the cathode. As it is I have 25vdc on pin 8 of the 6L6GC and 12.1vdc at the junction R31-R32etc.. As long as I end up with 12vdc at the junction I wouldn't worry too much I guess, but this seems like it might be problematic. I've heard of some guys just rewiring the socket(p1 has to be grounded and is currently a tie point for a 1.5k g1 stopper) and slapping in an EL34 but that seems kind of sketchy to me. I've already replaced R101 with a 1k 2w. The original part was a 1K 1/2w, yikes! EL34 g2 would draw quite a bit more current than 6L6GC at 500vp. Probably 20ma+ I would guess. Thoughts?