'59 Bassman RI eating power tubes

[methodofcontrol]

the previous tubes were matched and one got eaten. the current tubes are matched and one is getting eaten. the problem is the w/ the socket position not the tube.
the 95 mA is during red plating. i'm not even playing through the amp when this happens. i just turn the amp on and measure, by the time the amp warms up enough for red-plating to occur, this is where its at. then i turn the amp off.

[Euthymia]

At this point, I would reflow the solder joints on the faulty socket. Heat them up enough for solder to flow, and maybe add a little fresh solder.

[g1]

Have you tried changing the coupling cap? Or swapping with the one feeding the good socket if you don't have a new one handy?
And as to your question "i don't understand why, given the same voltages, V4 would get eaten up and V5 would remain ok?". If the 470R screen resistor was open at the good socket (and you had a bias problem to both) it may not redplate like the other.
However, your bias voltage does not seem to be that far out of whack so it would seem the coupling cap or bad solder would be likely suspects.

[methodofcontrol]

which is the coupling cap? pardon my ignorance. thanks again!

[MWJB]

C12, 0.1uf. Not easy to change, you'll need to drop out the board. Have you checked the tube socket?

[jrfrond]

Essentially, there are only three reasons a power tube will redplate:

1) Lack of bias (you really do need to measure the bias while redplating and FYI- it's going to have to be WAY off for redplating to occur; check and retension the tube socket)

2) Shorted tube (but you already proved that wasn't the case)

3) Carbonized socket (can happen more often than you think). This included cooked contacts.

The weirdest problem I ever experienced was a 220K bias divider resistor that opened up ONLY under load (and ONLY transients at that), and I had one tube actually explode on me! Cooked a few before I found the problem. Just when the tube needed the bias most, it lost it and went nuclear. Whew!!!

[methodofcontrol]

i retensioned the socket. i also swapped C12 and C13 to see if the red plating would follow the cap. i also reflowed solder on the V4 socket. i still have the same issue. i don't think i have to worry about R30 or R31 because the negative voltage seems fine. any other thoughts?

[pdf64]

How about if the phase splitter V3 is removed?

[MWJB]

Sanity check - get a pair of known good, matched 6L6, that match up in another amp & retest.

Still odd readings? Then you need to ascertain that negative voltage is connecting to the tube & not just the tube socket. With a dry wooden chopstick/other non-conductive probe, poke the 220K bias feed resistor & ensure that it is firmly in place & voltage supply is not intermittent.

If all voltage supplies to the socket are good & constant, & tubes are known good...it looks like you need to change the socket.

[methodofcontrol]

to check the socket connectivity, i plugged in my bias probe sans tube. set my DMM to continuity and everything came up good. visual inspection of the socket also looks good. are you suspect of the 220k (R38) solder joints? according to my meter, there's the same negative voltage reading on V4 as V5, wouldn't that imply that the R38 joints are solid? otherwise the neg.voltage wouldn't even get that far.
i can change the socket if it seems like the next thing to do. i have a spare.

[MWJB]

Well your bias probe test looks good, but it's as well to be aware that I have seen tubes function perfectly in amps that would give no reading with certain bias probes...so I guess it's feasible that the reverse could be true?

Did the chopstick test reveal anything? Note John F's observations above.

These "matched" sets of tubes, how did you ascertain that they were matched in the first place, did you check them in other amps?

If the socket is good, if the negative voltage supply is good, then it can only be the tube. If the tube is good and the negative voltage supply is good, then it can only be the socket. If the socket is good & the tube is good...well, you can see wher I am going with this...the socket is the only thing that you have not changed.

[g1]

Like MWJB said, the socket is still suspect. You could have a broken pin that is making contact most of the time. Your measurement of bias voltage while tube was redplating was made at the socket, not at the tube.
However, since no one has mentioned it yet, oscillation may be another possibility.

[pdf64]

Oscillation - that's what I was thinking of when I suggested removing the phase splitter.
Also need to bear in mind that when a tube has red plated it becomes more and more prone to red plating runaway.
So until new tubes are tried, may not know if the original problem was fixed, and you carry on chasing your tail with tubes that won't stabilize.
The risk is that the new tubes red plate, due to the original fault still being unresolved, and get similarly damaged by the red plating. Pete.

[methodofcontrol]

PDF64- congrats, you win!
i was about to head out for the holidays, and before i left this thing sitting on my bench for another week i decided just for fun to swap V4 and V5. Lo and behold, red plating followed the tube. these known-good used tubes were pull outs from a mesa 400+ i fixed for a friend...suffice to say, i've got plenty of used matched 6L6's.
so i grabbed another 6L6 and popped it in place of the offending tube and now everything seems to be AOK. not exactly sure which of the steps i took cured the amp, but PDF64 was correct in mentioning that once a tube red-plates, it's more prone to red-plating in future.
so, basically... after my attempted fixes, i was testing the amp w/ a faulty tube and therefore convinced that the amp was still broken.
i left the amp on for a good 20 minutes and kept a close eye on it. even played it dimed for half of that time. i gave it back to my friend and told him to play the hell out of it but keep an eye on the tubes and if it gives him any more trouble, i'll be back in a few days.
if y'all don't hear me, that means case closed. happy holidays, guys. thanks again.

[stokes]

Glad you fixed it,but I did suggest switching the tubes about 10 replies ago.