AC30 build Redplating after 30min play time

[glassesnvests]

Hey, guys, I've come to troll the great wisdom of the EL34World.

I have an AC30 I built from scratch, and after some fun stuff to get it to work, I indeed did. And it sounds killer. Utterly fantastic.

Until about 15-30 minutes into being on, the signal fizzes out and the 2 EL84's closest to the rectifier (V6, V7), start to redplate. Voltages test great throughout the amp when idling, and I haven't gotten a read of them in the process of going red. Could this just be a biasing problem, or would a failing capacitor do this? Thanks!
Adam
 

[tubeswell]

Bias is failing as the amp heats up - could be a bad resistor in the bias circuit drifting with heat?

Could be bad tubes? Try and repeat the trick after swapping the pairs of output tubes around to see if its the same two tubes, or the same sockets that they are in that is causing the problem.

[glassesnvests]

I'll check the resistors. I don't see any physical signs of failure (bubbled solder, charred resistors, etc) but I'll test all the items in the region.

Tubes have been swapped (although I only have 4 on hand, they are consistently V6 and V7 that redplate; others stay cool.)
Adam

[labb]

The AC 30 is cathode biased with one biasing resistor for all four power tubes so if it is a bias resistor problem you should see the problem with all of the power tubes. I would check the cap that goes to the grids of  V6 and V7.

[plexi50]

Check bias grid capacitors with cap tester if you have one. Ive seen the EL84 Yugoslavian tubes do this red plate issue a lot. And yes having separate cathode resistors for each pair of tubes is recommended (stability). There can be a lot of variance in (4) tubes all depending on a single cathode bias resistor

[clyde]

Grid leak resistors measuring too high?