Uneven plate voltages? - now a Marshall troubleshooting thread

[pdf64]

Another possibility is a dodgy socket connection to pin 5.

That’s another pair of EL34 probably wrecked, I suggest to fit cathode fusing of some sort, even just 1ohm 1/2W fusible cathode resistors. Why are you relying on the probe, rather than having done that already?

[Yoda]

So, forgive the newbie question but can I test the PI output coupling caps for DC leakage by simply lifting the output end of the cap off the PC board and testing for a certain DC threshold, with the amplifier fully on? Thanks.

[pompeiisneaks]

Yes basically desolder them so you don't get some kind of bad reading coming from the other side. With the amp on, you connect your DC voltmeter/DMM and read the DC voltage, there may be some small amount of millivolts there I think all caps have some tiny leakage, but if you see a few volts or more, then it's probably leaking and needs replacement. You could also do a current test, but it's more dangerous as using the current side of your DMM and not knowing how much current there is can lead to dead DMM's or dead amps... If you dead short the wrong thing etc. Fusing in the DMM is supposed to protect you and it, and they most often work, but still, it's dangerous to read current. Most of the time people just assume if the lifted leg is showing volts of DC there's enough leakage to be a problem.

I've not done this test myself, I've read about it a ton here, etc.

Someone that's more versed might be able to give a bit more confirmed details on the exact thresholds that become problematic.

~Phil

[Yoda]

Thank you. I’ll keep that in mind if I do test those coupling caps but for now another interesting thing came up. So I started testing the amp with only 2 output tubes installed instead of 4, and I was able to mix and match the tubes looking for the closest matched pair and while on the current limiter one output tube in V6 would draw a steady 10.1 mA while the other in V5 climbed to 40mA before I turned it off. I swapped the two tubes around and repeated this and the same exact thing happened but this time the runaway bias followed the tube! And it was on the other side (V6 + V7) of the power amp than what I believed to be the affected side ( V4 + V5). So now I believe I have one bad output tube out of my new set of 4, because I checked the third tube that was not part of this test and it also stabilized around 10-11 mA during with the current limiter. This was difficult for me to see before with all four output tubes installed but only 2 bias probes available to take readings.

[Yoda]

I received a replacement tube matching the 3 good ones and got the whole quad of output tubes up and running again and after some time has passed I’m happy to report this JCM 900 isn’t blowing fuses or red plating anymore. Nor do I need to run it on a current limiter for any reason. I’ve got it biased up nicely at 38mA and it seems to be stable for the time being.