Ampeg VT-40 Red Plating

[rf7]

One power tube socket runs the tube hotter than the other socket. When the bias is set to eliminate the crossover notch, then the tube red plates in 5 minutes of guitar playing. When I put the amp back to stock, biased cold, then the tube slightly red plates when it is played for 10 minutes. All of the voltages measure OK, and the resistors measure ok also. I put a 1ohm resistor on the cathodes and I can see that it is drawing more current than the other socket. I actually connected another OT, and had the same issue.

Any ideas?

This amp gets dirty really fast, in fact the power tubes start clipping before the preamp does. My old V4 was clean almost all the way up.

[nsubulysses]

I'm certainly not a tech authority like others on this board are but if all resistors in the phase inverter and output section check fine, I would check the coupling caps in the phase inverter for leakage, which would cause bias drift, usually making one side of the push/pull hog current draw.

Would be nice if you could scope the phase inverter and see if the imbalance is occuring there or in the actual output section.

Although it is normal for this amp to run the output tubes somewhat unevenly like this, another thing could be that once a tube redplates, it usually will repeatedly. A clever guy told me this, "The main reason is that by the time the plate structure is hot enough to be glowing red, the elements inside are even hotter, and are typically deformed/melted. This is particularly true of the fragile grid and screen wires. Once the spatial alignment of the electrodes changes, so do the tube's electrical properties... typically in the direction of repeat thermal runaway. "

Also switch the power tube and see if the redplating follows the tube or the socket.

hope this helps a little. gotta run

This amp has 6L6s??

[The Dude]

...... switch the power tube and see if the redplating follows the tube or the socket.

[Enzo]

What does all the voltages are OK mean? After the thing has been on long enough to red plate, is there still proper bias voltage on pin 5 of that socket? And what voltage might that be?

[Diablo]

While you're checking the bias voltage on pin 5, you could measure the voltage with the amp on standby, and then with the amp fully powered up. The bias voltage might look fine on standby, but if you have a leaking coupling cap on the PI, you should see a change in the bias voltage when the high voltage from the PI plate leaks through the cap.

[rf7]

I swapped the tubes already and found that it happens only in the one socket initially, that's why I said it was the socket in my first post. I guess I should have made that clearer. I've put in a new set of power tubes and the same thing happens. The amp has JJ 7027's in it.

The Bias voltage sits at -65V in standby, then goes to -68V fully powered up. At clipping, the bias to both tubes sits at -69V. The tube that red-plates is seeing a little more signal from the phase splitter. Just before clipping, the tube that is red-plating is seeing 2Vac more signal than the other tube. Still puzzled.

Since I returned the bias circuit back to stock values, it barely red-plates when it is being played hard. And it stops red-plating when you stop playing. With the bias set hotter, then it red-plates a lot worse, of course. Is slightly red-plating ok? The way the amp came to me, it was cooking the one tube quite a bit.

[Enzo]

What is the bias set at? I mean what current and plate voltage are the poweer tubes seeing? What dissipation at idle? How about backing it off some?

[g1]

What are the screen voltages on the power tubes? The same at both tubes?

Schematic:

[rf7]

The Plate voltage is 575V, the screen is at 574V on tube V9, but on tube V8 they seem to be reversed. The tubes have different current draws at idle, one is 13.8mA & the other is 19.1mA. The tube that looks burnt (and was red-plating when the amp came in) has the lower current draw. So dissipation at idle is around 10.9W for one and 7.9W for the other, which is pretty cold.

At clipping, the hotter running tube pulls 135.8mA in socket V9, while it pulls 116.5mA in socket V8. The colder running tube pulls 116.1mA in socket V9 & pulls 110mA in socket V8.

I think that the PI output is unbalanced and the hotter running tube accentuates that at clipping.

When the amp came in, the tubes were dissipating 25W and 29W respectively, which was too hot. Putting it back to stock makes it run cold. I was just surprised to see the V9 tube slightly red plate when it is being played. Maybe I should just leave it as is, with the colder running tube in socket V9.

[g1]

So you've tried other tubes, and also tried another OT. I think it would be worth a try to change the coupling cap to that socket, or swap the two coupling caps and see if the problem stays with the cap or with the socket.
Especially biased that cold, I don't think a couple volts difference between the PI outputs should make the one tube redplate.

[nsubulysses]

I think that the PI output is unbalanced and the hotter running tube accentuates that at clipping.

One power tube socket runs the tube hotter than the other socket. When the bias is set to eliminate the crossover notch, then the tube red plates in 5 minutes of guitar playing. When I put the amp back to stock, biased cold, then the tube slightly red plates when it is played for 10 minutes.

The phase inverter is imbalanced and it IS accentuated at clipping. If you push a V4, VT22, VT40 with a signal and watch the scope screen you will see a crossover notch develop at clipping, then move lower down the screen the farther into clipping you go, as one side of the push pull is driven harder than the other.

I was perplexed by this as well but have accepted it as normal as every V-whatever I have seen does this. This is because of the floating paraphrase phase inverter. I feel a little strange quoting this clever guy again, but I think he is right and it really applies here. It helped me, so maybe you too -- "The whole output section is enclosed a negative feedback loop, which linearizes the power stage. Once something (anything) clips, the NFB loop breaks down and performance becomes unpredictable. The PI tries to compensate for what's happening on the output, and very quickly the first PI triode is distorting. Now at this point, the 2nd PI triode is being fed a different signal than the 1st triode, because the 1st triode is seeing a clean wave plus the distorted NFB, the 2nd triode is seeing whatever weird mashed up thing the 1st triode is putting out in response to that. This is different from an LTP where the two sides are seeing roughly the same thing regardless."

The clue may be in the original post which is a problem of time-dependent red plating, to me indicating bias drift (thermal runaway?). The two tubes were dissipating 25W and 29W respectively as you said, when the amp came in. Makes me wonder if they both dissipated ~29W when they were new, but the tube in the drifted socket has been beaten up. When a valve ages it's anode resistance increases and it's transconductance decreases. Maybe this is the result of a long life in the "wrong" socket of this floating paraphrase or maybe a coupling cap is leaking.

Also 135mA isn't that much. Shouldn't be red plating, right??

Do new tubes do the same? Could this old tube be perpetually redplating because it has before??

[rf7]

I swapped the coupling caps and the problem stays with that socket. Brand new tubes do the same thing - the one on the end red-plates. It should not be red-plating with the bias set that cold.

I replaced the original bias caps today, just for kicks. That did not change anything, not that I expected it to.

[nsubulysses]

Wow, that would lead one like me into even greater levels of philosophy/mysticism, which may or may not yield results.

Someone with more experience please chime in. What is the potential resistors are failing under load (screen grid and control grid resistors for power tubes, plate and control grid resistors for PI, bias and phase inverter "balance" resistors or whatever they're called {R25, 27, 29, 30, 32, 33} )???? Like 5% chance?

If coupling caps, phase inverter (did you try a new phase inverter AU7, just in case??), OT, and power tubes have been accounted for...........

Or the speaker jack is making contact with ground somehow and dumping all that current. This is an anomoly but I saw it on a HR Deville a few months ago and I was mystified.........for a while

thanks for coming back and reporting these results. I wish I had greater advice but I may lead you further down the rabbit hole. Don't really know what else to suspect though, unless the tubes you're using literally cannot handle those near 600V plate voltages. Last VT-40 I saw I biased to 30mA at 571Vp with no red plating. They were JJ 6L6GC.

The tube socket is shorting or leaking to ground as it warms?? The speaker load is shorted?? D7-10 are shorted? Wow this philosophy could go on forever but I will stop for now. might be time to hook up like 5 multimeters and see what happens as the amp drifts into red plate territory. This is kind of a joke but not really

[g1]

If you want to test your theory about the extra drive from PI causing the issue, swap the leads from PI to output tube grids.
You will also need to disable the NFB loop so you don't go into oscillation.

[rf7]

If you want to test your theory about the extra drive from PI causing the issue, swap the leads from PI to output tube grids.
You will also need to disable the NFB loop so you don't go into oscillation.

I did this and the problem moved to the other power tube socket.