Ampeg J-12T RI red plating

[hylaphone]

Hi folks, I am puzzled on this one!

Amp came to me red plating, screen node C21 had been bulging and had already been replaced. One EL84 socket showed signs of heat.
Whatever had gone on had taken out the 100R screen resistors which I replaced with 1K 3W flameproofs. New tubes, lightbulb test, etc.

Voltages are correct, and stable. Red plating still starts after a minute or so at idle.
Plates are at 331V
Screens at 328V
Cathodes at 11.5V

Again, the voltages are stable even as the plates begin glowing.
There is continuity at all the socket pins.
Problem persists with PI tube removed.
Cathode resistor checks out.

I noted an imbalance in the OT primaries - one side is ~100R, the other at ~155R. Seems suspicious but I can't wrap my head around how this could cause runaway plate current
Scorched socket is suspicious but there is no problem with the voltage readings.

Any ideas where to go from here?
Thanks!

[g1]

Disconnect C13 and see if it still has the problem.

[Enzo]

You don't report the grid voltages. Your plates, screens, and cathodes may be stable, but what if the coupling caps are leaky from the PI? Verify the grid voltages are staying at zero or not.

[The Dude]

I calculate around 48mV bias per tube. IMO, this is too high. Either you have a problem, or you might consider raising the value of the cathode resistor.

[hylaphone]

Sorry, I forgot to mention the grids are stable at 0 VDC.

Tried removing C13, no change.
Bumped up R21 to 150R - tubes are idling at 39mA
Still glowing!

[g1]

Any other symptoms? I'm wondering if you may have a high frequency oscillation that is inaudible but making the tubes run full tilt.

[hylaphone]

No other symptoms, but I will scope it and report back

[hylaphone]

Well, I scoped the grids and plates and found no evidence of HF oscillation.
Ruling that out - maybe I should swap out the sockets for good measure?

I cannot afford another pair of tubes on this goose chase, is there any safeguarding when troubleshooting something like this?

[The Dude]

There are different schools of thought on this, and some like to run the hell out of EL84's, so this is just my $.02.

I still think your bias is too hot. At 331V plate voltage and 39mA, you are still at around 110% dissipation. 25-29mA would be 70-80% and is what I would shoot for- and I would go to the low side of that. Again, just my opinion. There are no rules and if there are, rules are made to be broken. This may or may not be the cause of your red plating, but either way, it's going to eat tubes at 110%.

Edit: This assumes you were talking 39mA PER TUBE.

[hylaphone]

You are right - I just referenced the schematic, which has the tubes dissipating about 13W each.
I'll bump up the cathode resistor to 200R and see how we stand.

[Leo_Gnardo]

You are right - I just referenced the schematic, which has the tubes dissipating about 13W each.
I'll bump up the cathode resistor to 200R and see how we stand.

What the Dude said ^^^ , plus when you replace EL84 output tubes, order ones with a low emission reading. They tend to run just fine but not run away. I usually get Ruby tested & matched JJ's with Pc in the 18 to 20 range, they behave well in self bias amps. I'm sure you can get similar from TAD and Antique, maybe other sources too. OTOH about 20 years ago when EI tubes were available, just about all their EL84's ran sky high bias, you could count on 'em to red plate.

[hylaphone]

Ok the plot thickens. I bumped up that cathode resistor to 200R and the amp idled fine for 10 minutes, played fine for another 10. Turned up the tremolo intensity (no switch in this amp) and whammo - plates aglow.

I am not familiar with this style tremolo, but it is putting a wiggling DC voltage -up to 25V- at the Intensity wiper (and to the R17, R18 junction)
Shouldn't that be blocked by C30?

[The Dude]

If you follow the intensity control wiper, you'll see that it is fed to the grids of the output tubes. The tremolo signal "modulates" the bias voltage. The reason I said in my earlier post that I would shoot for the low side of bias was for this very reason. So, my best guess?...........your bias is still too hot. With the 200 ohm cathode resistor, what is your idle current?

[hylaphone]

Gotcha, it's wiggling the ground reference --
Current is now 31mA/tube.
Dude has had this amp for a while, presumably with no issues.

I'll take a closer look at the trem section in the morning, may wanna tame it anyhow as I recall higher settings sounding choppy/nasty..

[Jazz P Bass]

Quote: "recall higher settings sounding choppy/nasty.."

That type of Tremelo has a nasty habit of making the output section 'clip' if you are running close to the clip point, volume wise.

Just a heads up.