Supro S6497 Red Plating Mystery (to me)

[ServiceBob]

6497 Back end.png

Hi folks!
I'm trying to restore a neat old dumpster find. Its condition was not too bad, though it had no tubes and all of the filter caps were invisible to my capacitance meter. I have replaced all of the filter caps with F&T's and all of the other electrolytic capacitors. I've cleaned all of the jacks, pots, switches, and tube sockets I've put new matched Russian 6L6GCs (that's what the seller said they were - they've no markings at all). The amp powers up and the amp sounds solid if a bit too clean, and maybe not loud enough, but after a few (5-10 minutes) I noted that one of the tubes was beginning to red-plate so I shut it down. I swapped the tubes and retried. In a few minutes, the tube in the same socket started to red-plate. Next, I put a bias adjusting adapter on the bad socket and watched the current start at about 60 mA after it had warmed up but slowly head to almost 80 over about 5 minutes before I shut it down. The increase got faster as it went along.
Since the power amp/power supply is in a separate chassis than the input/control panel/preamp/Phase-inverter section there is a two-wire cable that carries the signal from the PI outputs to the power amp inputs. I swapped the two wires where they attach to terminal strips in the preamp chassis, and sure enough, the red plate switched sockets. I think that rules out the power tube sockets, the OT, and the cathode resistor.
Next, I checked for DC voltage on the control grid and sure enough, it starts out at about 0.003 VDC and slowly but steadily creeps upward to almost 2 volts before I shut it down. (The control grids were both at about -28VDC relative to their cathodes when I checked them at the beginning of this process). The increase picked up speed once it got over 1.2 VDC I noticed that the "good" side barely moved from 0.003 - 0.005 until the "bad" side crossed over 1 volt. Then it began to rise also. But it never got near the "bad" side. (I figure that this is because they share a cathode resistor). I replaced the 0.02 uF coupling cap (a ceramic disc) with a new polypropylene film capacitor. There was no change. I also changed the 470K grid-leak resistor on that side as it had drifted to 560K and I replaced both grid-stoppers as they had both drifted pretty far north. Still no difference. I changed the 200-ohm 7W bias resistor to a 260-ohm 10W resistor just to rule out that element and to lower the dissipation a bit. It had been 21W before the runaway began and is about 18W with the new resistor. But it still rises on the "bad" side.
I haven't been able to find a schematic for this particular amp on the web, so I did a quick one showing the pertinent part of the circuit. All of the component values are correct as marked, but to be honest I have not tested all of the resistors. Both caps tested fine. I replaced the 0.02 uF coupling cap anyway.
The question is where is the DC voltage coming from on the "Bad" side? Why just that side?
Thanks in advance.

[pompeiisneaks]

If C2 from your schematic is leaking DC then you'll get the behavior you're mentioning. It's easy to test. Desolder that lead that goes towards the power tube section and see if it still sees DC voltage there. If so, then it's adding bias to the grids of the power tube and causing the red plating. If so, replace that cap with a new one and that should solve the problem.

~Phil

[Stevem]

Is that tube getting power to its plate from the OT winding, if your seeing a negative voltage on the plate then it’s not.

[ServiceBob]

Thanks for your input!
I did consider C2, and as I mentioned in the original post I replaced it even though it tests good. Nothing changed.
I have +391V at pin 3 on both 6L6 tubes.

I did lift C2 just as a test to see if it made any difference. Surprisingly, no difference. So the voltage seems to be coming from the tube (?). Is it electrons building up on the control grid? I don't get it.

UPDATE:

I decided to let it idle to see how high the DC voltage on the control grid of the bad side (V4) would go. It rose slowly for about 20 minutes and then stabilized at +1.7 - +1.8 VDC. The "good" side is about half that. The cathode voltage stayed steady at about 31 volts the whole time.

No red plate. It sounded fine. But still too clean. No breakup at all. But that's a different issue.

[statorvane]

That poweramp is virtually identical to the 1688TN, except the PI coupling caps are 0.047 uF, and the PI cathode resistors/bypass cap are swapped. I built one a couple of years ago, but added 470K / 5W screen resistors and upped the poweramp cathode resistor to 250 R. No redplating. I have about the same plate voltages as you (388 VDC). I did not notice any dc voltage, but I did not check.

The only time I have seen that sort of behavior is when I forgot to solder a grid leak resistor at a power tube grid in a SE amp build. Might want to check that.

[ServiceBob]

statorvane,
Did you mean 470R / 5W screen resistors? 470K seems awfully high.

[Lauri]

If the coupling caps aren't leaking then your tubes are probably gassy and pulling too much control grid current. You're getting higher voltage on the other side because there the grid to ground resistance is over 500k when on the other side it's about 330k. Try it with a new set of power tubes and see if it changes anything.

Can you post a picture of the Russian 6L6GC tubes you have? Sometimes old 6P3S tubes are sold as 6L6GC equivalents but they have maximum plate dissipation of 20W and can't handle high voltages very well.

[ServiceBob]

Here's a photo of the tubes.
Right now the four "Matched" tubes I bought are the only 6L6GC tubes I have. Actually, I have one used Ruby Tube 6L6GC that looks nothing like these. I do have a new set of Chinese 5881A tubes. Would they be worth trying?

[Lauri]

I'm not sure but those probably are 6P3S or 6P3S-E tubes. Try it with 6L6GC power tubes and measure the voltages again.

[ServiceBob]

I have ordered a pair of new JJ 6L6GC tubes. I will let you know if that makes a difference.

[statorvane]

Yes, 470R / 5W screen resistors.

My power tubes are Tung Sol 6L6GC-STR. I have also tried some Sovtek 5881s. Rectifier is a Sylvania 5U4GB.

[Stevem]

Those Russian tubes are not 6l6 type tubes and are not even good for as high a plate voltage as the China made 5881 tubes you have!

In those Russian tubes have all the connection points run in a very arc prone way right up thru the center of the tube .
Take a close look at what I am talking about.
I have 6 of these type of Russian tubes and even though they are labeled as 6l6 tubes I have only used them in my Fender Champ at times .
I would not trust them in any amp running over 360 volts on the plate.

There terrible construction reminds me of a 194 blade type lamp bulb for Automotive usage!