EL34 amp blowing up, anode glowing red hot

[JensDensen]

so I have this multitone stereo 2*50 watts amp and suddenly there's a humming noise in one of the channels, the whole thing shuts down and one tube has an anode glowing red hot (number two) and that was it. I bring it to somebody to repair, puts two new tubes in it, and there it is, the same thing again, same tube, I switch it of, do some measuments, seems, ok, still working, and there goes number four, anode glowing red hot, switch of the high-voltage, cool down, switch on again, no problem, still working. What's going on here? G1 measures -40 volts on all tubes, anode 470 volts, g2 460 volts. The amp is a standard design using 2 el34 per channel stereo.

[john65b]

Bad tube pin connection? Bias runaway? Check your cathode resistors.

I had similar problem on ST-70 with EL34. Changed all tubes, including rectifier tube, and all was well again.

[TheGimp]

The Traynor YVM-1 (PA Amp with EL-34s in fixed bias) I'm working on has done the same thing.

In my case it was traced to a poor ground on the speaker connection. The bad ground would occasionally remove the load from the output which in turn radically changed the impedance seen by the output tubes. In my case I still don't understand why one tube seemed to take the brunt of the overload.

[SoNic_real_one]

Ultrasonic oscillations?

[DigitalJunkie]

Loose pin connections/solder joints,or leaky coupling caps would be my first guesses.

[JensDensen]

Wow quick replies. I think I'll pick the 'bad socket' option. I've reseated the tubes and now waiting to see what happens. The coupling caps are polyster 400v types, I think they are ok. The bias is set with cheap pot's maybe they are a problem or their feed voltage circuitry which uses electrolytes. I guess I just have to monitor the situation. Perhaps hook up a multimeter permanently to ?? one of the g1's? The sockets look a bit 'cheap' but not 'oxidised' or anything. The whole thing is handwired and pretty solid. There was a crackling sound on one 'occasion' but for the rest both channels continued working. It's quite a thing to see the whole anode glowing orange. I'm suprised the tubes are still working after that ordeal. I guess they're made for that stuff huh?

[Ty_Bower]

Originally Posted by JensDensen

The bias is set with cheap pot's maybe they are a problem...?

If you have low quality bias pots (or even if they are good quality) it is prudent to connect a 100K ohm resistor from the top of the pot (the lug closest to the bias supply) to the wiper. That way if the wiper should accidentally lift, the bias voltage to the tube will go towards the maximum available bias voltage instead of going to ground. It will cause the tube to cut off instead of letting it try to melt down.

Here's an example schematic from the bias supply for a Dynaco MkIII. Look at the 100K resistor at the top of the bias pot.