freestompboxes.org

I'm feeling reassured that the Ammo Can load wasn't responsible for my amp blowing a tube. I stuck another set in there (EL34s this time), rebiased, checked all voltages, etc... Biased at 70% dissipation, I ran a sine wave directly into the phase inverter and got a very nice clean symmetrical since output without any obvious distortions. When pushed up to clipping, it would show a wave typical of the PI clipping out, so I will say on the bench running a clean sine thru the amp is no issue.

I played through it a bit, using only the Suhr load, and I noticed the tube flashing a bit again, and the tube in the same position as the one that blew was flashing noticeably more energetically than the other, and then I noticed a couple whisps of smoke come up from around the power tubes, so I stopped playing and shut it off. I cannot find what exactly the smoke was coming from - nothing appears burned inside. I had just carefully cleaned all on and around the sockets because I was afraid of arching, so it wasn't a matter of dust burning off. The only components that felt hot were the screen resistors. They are 1K 5watt and they got quite hot. I think this is an issue of something malfuncitioning in the amp, not the fault of the load. I've only owned this amp a couple years, but it has had some issue with screen resistors burning up before. It used to use a very large 10w 1K resistor shared for both screens, but it had been replaced at least once before I bought it, because the circuit board under that resistor was very badly burned and damaged - looking like the original resistor burned up and was replaced. I pulled the large shared 1K out and replaced it with two individual 1Ks. Anyway, that's a topic for a different thread, but I'm confident the issue is unrelated to the load, or at least not directly related.

Oh dear, As the Amp has already had Screen grid issues then the sockets may have hair line carbon traces (which you can't see) and once that happens it's likely the new tubes will arc again along the same path, in which case it's a Socket replacement.

Also worth note is that EL34's are notorious for Screen failure as it's the weakest design point. Add to that, a lot of EL34 Amps run the Screen voltage at or often above the Voltage limit of the Specifications.
Yes you can run EL34 Plates at insane HTV but screen is ALWAYS limited to ~240V. (from memory)
Yes some Brands claim higher Screen Voltages for their EL34's but a lot do not.

I recall that Tone Lizard guy mentions it on his pages.
I've had to lower the Voltages on a few EL34 disaster amps I've taken on.
A lot of Later Marshalls I've worked on seem to run Screens at dangerous levels.
They know that the Amp will at least last long enough to out last the warranty.
Phil.

phatt wrote: ↑16 Apr 2023, 23:23 Oh dear, As the Amp has already had Screen grid issues then the sockets may have hair line carbon traces (which you can't see) and once that happens it's likely the new tubes will arc again along the same path, in which case it's a Socket replacement.

Also worth note is that EL34's are notorious for Screen failure as it's the weakest design point. Add to that, a lot of EL34 Amps run the Screen voltage at or often above the Voltage limit of the Specifications.
Yes you can run EL34 Plates at insane HTV but screen is ALWAYS limited to ~240V. (from memory)
Yes some Brands claim higher Screen Voltages for their EL34's but a lot do not.

I recall that Tone Lizard guy mentions it on his pages.
I've had to lower the Voltages on a few EL34 disaster amps I've taken on.
A lot of Later Marshalls I've worked on seem to run Screens at dangerous levels.
They know that the Amp will at least last long enough to out last the warranty.
Phil.

What really miffs me is the that the tube that blew was a KT66.
With a pair of very used Ei brand EL34s, each dissipating around 17w, I measured 454v on the plates and 450v on the screens at idle. This is with 1K 5w screen resistors. The screens draw around 7ma at idle. When I run a signal thru, I got 30w into 8ohms just before clipping. As more power was pushed through it, the plate and screen voltage would both sag. Eventually, the plate voltage sagged to lower than the screen voltage. When pushed to severe clipping, I managed about 19,5VAC into an 8ohm load, nearly 50w. At this point, the plate voltage had dropped to around 398v-400v, and the screen voltage was around 422v. The screens were around 422v, and drawing around 35ma. The waveform was very squared with prominent flyback voltage "shoulder spikes."
The amp seems to have little trouble being pushed to its limits with steady sine waves - nothing started smoking or otherwise going wrong. But if I were to play through it and push it like that, it's a different story. The signal transients from playing cause havok with the voltage spikes.

Remember electrons flow to the Highest voltage potential.
Any power Valve would burn through the screen rapidly if the Plate voltage ever falls lower than Screen voltage.

As the amp is old the plate pin might be arcing across the socket somewhere which will pull plate voltage down. And I would assume not effect the screen voltage very much in such case the Screens would be dissipating way over there limit, hence the sparking.
As you have noted the amp has had screen issues before so the problem is likely still there but someone replaced a few parts , popped in a new set of valves and biased up, tested it at lower power and assumed to be ok,,,
But the original cause of failure is most likely still present.

I looked up a schematic and noted that it used EL34 but as you mentioned it's been modified so anyone's guess what has been done.
Phil.

Yes, the screen starts behaving as a plate, drawing lots of current, when its voltage goes above the plate voltage.
Please investigate why the plate voltage goes lower than the screen because this should not happen.

phatt wrote: ↑17 Apr 2023, 04:11 Remember electrons flow to the Highest voltage potential.
Any power Valve would burn through the screen rapidly if the Plate voltage ever falls lower than Screen voltage.

As the amp is old the plate pin might be arcing across the socket somewhere which will pull plate voltage down. And I would assume not effect the screen voltage very much in such case the Screens would be dissipating way over there limit, hence the sparking.
As you have noted the amp has had screen issues before so the problem is likely still there but someone replaced a few parts , popped in a new set of valves and biased up, tested it at lower power and assumed to be ok,,,
But the original cause of failure is most likely still present.

I looked up a schematic and noted that it used EL34 but as you mentioned it's been modified so anyone's guess what has been done.
Phil.

Yes, my main concern is that whatever caused its original screen resistor to burn up may still be an active problem, since it appears that the same part of the circuit is being affected.
The original damage must have been quite bad. The fibreboard has a hole burned completely thru where the original resistor had been, and one of the leads of the OT primary has been cut short with an extension spliced in place to connect to the socket - this leads me to believe the original wire must have been damaged also when the resistor burned up, because it likely would have run very near it. The various wires attached to the sockets have little paper numbers taped to them, presumably to keep track of where they had been connected, so they must have been desoldered at some point. Can't say the sockets look particularly new, though. Could be the person disconnected the wires to repair some damage, and then reattached them to the original sockets. I have a new pair of nice Belton octal sockets on-hand, so I shall replace them just to be certain I've eliminated that possibility. I also think it would be wise to install some "tube saver" flyback catch diodes.
I also don't feel good about the fibreboard - I don't like them generally, and seeing one with an area that has already been carbonized by fire makes me think it would also be wise to pull the whole thing out and rebuild it with a new board. I'd like to wait until I've settled on the final preamp design before replacing the whole board, though. Maybe I could just cut out the burned section.

Oh dear Well if there is that much damage then for sure you have carbon particles all around that area which could well be creating a pathway for the hi voltage to trace through.

Sounds like you will be forced to rebuild the Board if you want it to be reliable.

This is why you see flame proof resistors in some circuits as it's not just the fire risk it's also because when carbon resistors blow they blast carbon all over other parts. Those practicals are conductive and when the voltage gets high enough they join up and leak voltage potential back to whatever is close by.
Remember that AC signal voltage on the plate pin can be twice the DC plate voltage in PPull amps.

Re Diodes; IIRC, Diode strings across the plates is really just protection for open circuit on the secondary to clamp the primary voltage from going to insanely high levels causing breakdown in the primary winding.
My thought is that it would not protect from the issue you have with this amp.
Phil.

soulsonic wrote: ↑17 Apr 2023, 13:10I also don't feel good about the fibreboard - I don't like them generally, and seeing one with an area that has already been carbonized by fire makes me think it would also be wise to pull the whole thing out and rebuild it with a new board.

Sucks, but maybe an opportunity. I've got a YVM-1, and while pretty cool as is, that chassis would also be a great platform to build pretty much whatever.