Annyone seen this problem?

[ingvar ahlberg]

i´ve got a russian amp, Natural A7, that behaves strange. Starting up it works well for about 30 minutes then one output tube, el34, starts to glow intensely and the amp starts to distort heavily. It´s a triode conected pp with 12ax7 input and ecc88 phase splitter. The problem is not tube related but position. All tubes have the same bias -28v, seems a bit wrong in triode connection? Both channels distort but the only common part is the 12ax7 and that´s ok, and psu. I can´t find anything defektive among the passives and everything seems to measure identacal between the tubes. Gratefull for any tips on this.
Regards

[DoomPixie]

did you check all the voltages after the amp had warmed up and the fault had occurd? mabey theres a component that is failing to short when hot or similar..
Owen

[kevinkr]

Sounds like a bad coupling capacitor to me.. The grid bias may be on the low side as well depending on the supply voltage - if it is under 400V it is probably ok..

[ingvar ahlberg]

Thank´s for input, turned the amp uppsidedown and monitored, what happens is: bias of brightly glowing tube drops, backwards so to speak to about -14V while the rest remains at about -25V.
Cathode of bright tube goes to 2V, cathode resistor is 10 ohms so cathode is just some mV above gnd. Will replace cathode resistor (measures 10ohm off power) but still can´t understand why both channels go into Neil Young distortion teritory (good as it is) when just one tube seems to fail.

[dhaen]

still can´t understand why both channels go into Neil Young distortion teritory (good as it is) when just one tube seems to fail.

If Kevinkr is right about the coupling cap failure (and I think he is) then the leakage might be affecting the bias supply.

[ingvar ahlberg]

Replaced cathode resistor, did´nt make any difference, so i´ll
try replacing coupling caps. It´s a "pure" triode connection, no resistor between plate and grid, could the problem be related to excessive grid current? As i´m planning to make it switchable to pentode mode maybe it shoulb be better to introduce some 50 ohms between plate and grid for triode mode?

[dhaen]

Can you link or post the schematic?

[ingvar ahlberg]

Schematic Natural A-7:

[dhaen]

maybe it shoulb be better to introduce some 50 ohms between plate and grid for triode mode?

According to the schematic, the designer did not find it necessary with the present layout. However if you plan to extend the screen wiring for switching, it might be a sensible precaution against possible parasitic oscillations.

[ingvar ahlberg]

Will do that, after i get it to work properly as is, then start the modifications. It´s nicely built, hardwired and with pretty good or very decent parts. Before problem sounded good but a little bit low on power, very good bottom end so opt´s are probably good.