Carvin 3200 conversion to JTM45 problem

I'm doing a conversion for a friend and have hit a snag with the bright channel I hope someone can point me in the right direction. The normal channel seems to be functioning properly, the bright channel farts out with increasing vol (>2 or 3) and is worse with higher treble or middle, the bias reading jumps at the same time.
I followed the triode electronics layout and component values. Voltages appear to be within 10% of specs. Tried various tubes and re dressed wiring. I'm using the original PT and OT (8ohm out with 2x12 8ohm speakers from the Carvin) Also I re configured the pot order to fit the Carvin graphics and have longer wire runs than I would like, maybe I shouldn't have done that. I'm ok with changing anything to improve the situation it doesn't need to be authentic JTM. Was thinking of redoing the pots first.
Changes - addition of 5v fil tranny for rectifier, blocking diodes, pot order, Carvin iron, no standby, bias measuring jacks, artificial center tap. Any tips appreciated.

thanks Gordon

 

To clarify the bias jumps up and vol cuts out with both middle and treble increase, Vol not so much. I just changed the couple cap 0.022uF checked the old one, ok. Changed cathode bypass cap 330uF at 50v, removed one also checks ok.
All sockets are new and I see no sign of a problem there. I also move the lead dress to the pot and it seems to have improved things, cuts out now at vol of 5 instead of 2! Thanks for helping me out.

g

 

I will re-order the pots and try it again I thought about this during the build but I guess I made the wrong call. I'll post a photo after re wire since Im going to swap them about. The Carvin chassis isn't ideal for board alignment to pots and sockets since I reused the pot and socket holes.

thx for the help g

 

Rebuilt the tone stack and re-ordered the pots to match the jtm layout as best I could, still have the problem but its not as bad. I think my next step is to recheck the voltages. Not much room to make the final connections but I checked
them all with a magnifying glass and they seem fine.
Pot and socket layout is governed by the re-purposed chassis layout this is about the best I could do to have the most direct runs.
Any observations appreciated Im about over this build! where's the whisky.

thx g

 

she's sorted, its was the OT wiring, long runs because of the original transformer position. I re directed the primaries away from the board and everything is stable again. Sounds brilliant on all inputs and quiet. Thanks for the help.

g