Bad red plating problem!

[Silvio55]

I installed a crossline master volume on my 18 watt build, I used a 500k no load pot (I scratched the end of the carbon strip in the pot), the amp works but Iīm having a strange problem, with the volume full up one of the power tubes red plates in a bad way, and with the volume full down the other power tube red plates! I donīt have a clue why this is happening, does someone know what could be happening here? Thanks!

[pdf64]

Need a schematic of exactly what you've got, and a detailed photo of two of the implementation.
Suggest you remove all tubes apart from power and rectifier, monitor the grid and cathode voltages whilst adjusting the crossline.

[Silvio55]

Here's a layout and schematic, with the mod in red:


I don't have a camera right now Just two wires coming from the PI coupling caps, the layout is exactly how I installed it, pretty simple.
I will measure the voltages and report

[pdf64]

Ensure that the wires to the crossline are well away from any input wiring.
Twist the wires to the crossline.

[Silvio55]

Yes pdf64, the wires are firmly twisted and away from everything else.

[Silvio55]

Measured the voltages:
Grids (pin 2) one tube measures 0.75V with volume down and 1.25V with volume up.
The other measures 0.45V with volume down and 0.04V with volume up.
Cathodes measures 11.7V with volume down and 11.8V with volume up.

[eschertron]

Are you sure that the pot's third terminal isn't shorted to the case or any other point? With the volume 'up', especially with the no-load arrangement, the amp should work just as before your mod. But it doesn't. Something in the mod (or in the amp as result of surgery) is not as you expect it to be. I'd check for blobs of solder or some other unwanted artifact of the mod. But that's because I'm likely to do something clumsy like that

[Silvio55]

eschertron, I checked the pot and the conections in the board and found no shorts or anything bad.

[Jazz P Bass]

Which end of the pot taper is scratched off?
I would imagine the top.

Not too sure where this 'Crossline' thing is going.

Control up all the way (scratched off taper) it does nothing.
All the way down it ties the grids together?

[eschertron]

Which end of the pot taper is scratched off?
I would imagine the top.

Not too sure where this 'Crossline' thing is going.

Control up all the way (scratched off taper) it does nothing.
All the way down it ties the grids together?

The basis for my point. If at the 'up' position, the pot is physically disconnected, then the amp acts the same as it did before the mod. So logically either 1) something in the mod or the amp is not as we think it is and should be investigated, or 2) the amp always red-plated one of the tubes when run full out. Which is it?

[DRH1958]

Have you tried to replace the El84s? It seems as though this might be worth a try. Do you know both tubes to be good? Disconnect the master first to see what happens, then replace the EL84s and see what happens. If there is no redplating, reconnect the master to see what happens.
Also, any chance of DC leakage in the .01uF PI to power amp coupling caps?

[Silvio55]

Jazz, the top portion is scratched off, so when the pot is full up it desconects.
eschertron,I installed the MV some time ago, it may be possible that the amp redplated before, it redplates with no, or very low signal, playing hard (like I've been using it most of the time) it doesnīt red plate, but it's hard to me to belive it did it before, never had a problem with it before, but most of the time I played with the master volume around the middle position (it doesn't red plate then), I played with the amp very little before modding it.
DRH1958, They're brand new JJ's, tried other tubes I had laying around and the same thing happened. Leakage dosn't seem to be the problem, since I tried removing the pre amp and PI tube and it red plated anyway.

[DRH1958]

The only other variables I can think of is to take plate and screen voltage measurements with various settings of the MV. Try quiescent, then low, mid and high vol settings. I'm wondering about the screen grid resistor, R25 100 ohms. Maybe take a resistance reading before powering back up. Do they give you a voltage chart for static conditions? Make sure the screen grid is a few volts lower than the plate. Do you have a dummy load resistor?
You were on the ball and took cathode voltage readings, but with full volume, the reading should have gone up more than .1(11.7 to 11.8) volt at full volume since it is cathode bias. Current will(should) increase more than that and the voltage should go up by a few volts I would think. I suppose you should be ready so as to not damage the tubes when redplating.

[Dave H]

Leakage dosn't seem to be the problem, since I tried removing the pre amp and PI tube and it red plated anyway.

Removing the PI tube will make any leakage worse (higher voltage). The voltage readings in post #6 are far too high. I think one of the coupling capacitors must be leaking. Remove the PI tube and both power tubes and measure the voltage on the EL84 grids. It should be no more than a few mV.

[J M Fahey]

I would try an unmodded pot, don't trust an opened, scratched and reassembled pot at all.
500k plate to plate is easy to drive for a PI so no real need for a no load pot, in fact it was always done that way for decades.
Just think about the VOX "cut" control, the Matchless crossplate Master control, etc.
500k loaded or not, PI plate signal is way too much to fully drive power tube grids anyway, so no problem if you lose a little.