6550 shorted plate to filament -- possible causes?

[hylaphone]

Anyone seen this type of tube failure before? .. short between pins 2 and 3 on a JJ 6550.

Amp in question is a 70's Ampeg V4 - which has had a troubled past - and has been eating output tubes when pushed.
Tough one to troubleshoot, as the PCB is in poor shape and previous repairs have been crude at best.

So far this visit, I found the shorted tube and two open 100r faux CT resistors (5W!).
The original humdinger is long gone.

Second retube this year. Sockets are new and the failures are not exclusive to a single socket.

Other than possibly a bad batch of tubes, what might I look for here?
Intermittent short between PT windings?
Leaky coupling cap?

I'm used to seeing screen failures in these tubes, not sure what this one indicates.

[Leo_Gnardo]

Arcs from plate to nearby filament pin one of the most common failures. And when that hi voltage gets into the filament circuit, wave goodbye to humdinger pots or balancing resistors.

Some V4 and similar Ampegs had flyback quenching rectifiers factory installed. If yours doesn't maybe it's about time. Cheap insurance that may stop this from happening again. I learned about this technique before reading it on Ken Fisher's Trainwreck Pages, but I'll send you there to read up on how to do it. A half dozen 1N4007 rectifiers and 5 minutes might be all you need.

Jump to page 35:

http://www.blueguitar.org/new/schem/...reck_pages.pdf

MusicMan, Peavey, Fender have been using this for decades now. Well, OK MM's now outta biz, but when they were, OK? Still good tough amps.

[hylaphone]

This is an early V4 without the flyback diodes. I will add them.
However, I thought those were for unloaded OT protection?

Would really like to understand what caused this fault to begin with.
No continuity between those pins on the sockets, and no visible carbon..
Amp was biased at 60%.

[Leo_Gnardo]

General figure for breakdown of air as an insulator AKA arcing, is 8000V per inch, that's dry air at sea level pressure, derate for humid & lower pressure which is always. That's 1000V for an eighth inch, 500V for a sixteenth, there's not much space between tube socket electrodes so we're in that territory. And it's instantaneous voltage, not average. You already have over 500V on plate pins 3, just a little gap away, filament at near ground on pin 2.

In driving speakers, the cone attempts to follow the output waveform of the amp but is always just a little behind, because you're moving a mass and it takes a little time to get going, call that momentum. Also takes time to stop - and in that instant the voice coil is moving thru the speaker's magnet it's acting as a generator, sending a spike of current back thru the OT's secondary. That current spike is multiplied by the turns ratio and shows up as a rather large voltage spike on tube plates, most easily seen on a scope when driving the amp into clipping. Dial up the brightness on your scope & shift the main part of the image up or down, you'll see those spikes. This "flyback" voltage can go to 3000V no problem, now there is a voltage that can easily arc to nearby ground or near-ground. Clamping rectifiers in, no more spikes, not only does it help prevent arcing on the socket but tube life & transformer life can also be extended.

There are some who claim "but I can hear a difference!" though Ken Fisher says no. Well I do hear a difference, and it's for the better. One of the components of the overdriven sound with flyback spikes present is a static-y noise. With rectifiers in that noise is gone & the OD tone's a bit smoother. So if you must have that static-noise as part of your overdrive tone palette, well keep lots of tube sockets handy, have your soldering iron always hot & leave out the flyback rectifiers.

[Enzo]

If tube sockets were clean w could calculate the air resistance, but all it takes is some small debris or contamination, and arcs can strike. Just because you don't see a burn or measure a short, doesn't mean the socket isn't arcing.

[hylaphone]

I understand now, thank you fellas!

[frus]

There are some who claim "but I can hear a difference!" though Ken Fisher says no. Well I do hear a difference, and it's for the better.

I've recently made a 2xEL84, fixed bias
Adding the diodes made the amp much less buzzy when overdriven, but really a night and day difference. To the better of course.
I guess it depends on the amp/speakers/OT/whatever

[trobbins]

Courtroom example #1 ceramic valve socket.

[DrGonz78]

Courtroom example #2 from a 1970's V4. Probably just a bad tube in this case since it was the only socket that was effected. The hum pot was melted badly.

[trobbins]

Hmm. I would have initially thought anode-heater flashover. I wouldn't say that has to happen on both valves of a PP stage - it could have started on one valve base due to dirt/debris, and then carbon tracking between the pins allowed almost any voltage to track across. Being DC, the positive side (anode terminal) may well be consumed more than the negative terminal (heater).

[ThisLifeILead]

My 1971 V4 has flyback diodes. Almost all tube shorts ive seen are from pin 2 to 3 or from inserting a tube incorrectly (broken key). just my 2 cents. JJ's are suprisingly good tubes. Ive never encountered a bad tube from them since 2007 but i also order them from tube depot. (might make a difference). prob a dirty tube socket as mentioned earlier