Filament transformer for 8xkt88

[Enzo]

Devacuumizing? You mean a 1 ohm resistor will cause the tube vacuum to fail and let air in? No way on earth. A 1 ohm resistor between cathode and ground will not act or appear any different from a direct cathode to ground connection. All it does is let you measure a tiny voltage drop across it and compute current.

Let us say you have 40mv across the 1 ohm. Your B+ supply to the tube might be 400v. I leave it to you to figure out what percentage of 400 volts 40mv might be. Hint, a VERY TINY amount.

[Mr_bibbles]

Devacuumizing? You mean a 1 ohm resistor will cause the tube vacuum to fail and let air in? No way on earth. A 1 ohm resistor between cathode and ground will not act or appear any different from a direct cathode to ground connection. All it does is let you measure a tiny voltage drop across it and compute current.

Let us say you have 40mv across the 1 ohm. Your B+ supply to the tube might be 400v. I leave it to you to figure out what percentage of 400 volts 40mv might be. Hint, a VERY TINY amount.

That was sort of my thought too, but Catalin is correct in saying that it doesn't match what should technically be on the pinout... so why did the engineers settle on the pinout?

G1 had stated

Your earlier drawing showed pins1&8 tied together, I think that is bad with these and 6550 tubes.

, so there seems to be a consensus this is not the best way to do it, but still no one has been able to state why...


Enzo I agree with you that the 1 ohm resistor to ground (when in good condition) would not cause a tube to "white-top", but perhaps the base and cathode are interacting in a way no-one has mentioned? I think it's worth exploring the idea especially with the evidence we see (arcing on the tube side).

[Helmholtz]

When tubes draw air (which oxidizes the getter metal layer) the reason is most likely excessive electrodes or bulb temperature (i.e. apart from production problems or mechanical effects).
See e.g. R.B.Tomer's book "Getting the most out of Vacuum Tubes".

So check idle dissipation and watch out for redplating at medium to high output.
Another reason could be thermal runaway caused by high temperature operation with too high Rg1-k (grid leak resistance).
Recommended max. Rg1-k is 100k per tube. So a shared grid leak resistor should be 25k per side with 8 tubes.

Try to draw a schematic of the power stage.

Also pictures of the amp could be helpful.

[The Dude]

I may have missed it, but what brand tubes are in the amp? They aren't new production Mullards by chance, are they?

[g1]

There are a few different things being discussed. One is the safety factor due to the metal base being sometimes connected to pin1.
If you tie pin1 & 8, and ground through the cathode 1ohm sense resistor, that resistor can fail. If so, the metal base can rise to a high voltage and pose a hazard to the user or tech. Of all the 6550 & KT88's, we don't know which brands have the base metal connected to pin1. Sometimes their own datasheets don't show it. So to be safe we assume they all do.

Another issue is if pin1 on the socket is used as a tie point for the grid resistor. Now you have a situation where a metal tube retainer (spring or clamp type) can short the bias to ground. This could possibly lead to excess temperatures maybe in the range that Helmholtz is talking about. Could this be enough to compromise the vacuum? I said before that 'it's not impossible' but maybe that is more a philosophical stance that 'nothing is impossible'.
I have seen bulbs completely melt down and deform before they finally lose their vacuum.
But looking at your drawing of the actual build, pin1 is not used as a tie point so I think we can rule this out.

Your second quote in post #24 was from Mick Bailey . He mentioned instability but again, it was when socket pin1 is used as a tie point, so I don't think it applies.

In any case, I think the idea that any thing electronic could be responsible for vacuum loss, is something we are all very skeptical about. It's much more likely due to some physical issue, but we are just spit-balling other stuff.

[catalin gramada]

When tubes draw air (which oxidizes the getter metal layer) the reason is most likely excessive electrodes or bulb temperature (i.e. apart from production problems).
See e.g. R.B.Tomer's book "Getting the most out of Vacuum Tubes".

So check idle dissipation and watch out for redplating at medium to high output.
Another reason could be thermal runaway caused by high temperature operation with too high Rg1-k (grid leak resistance).
Recommended max. Rg1-k is 100k per tube. So a shared grid leak resistor should be 25k per side with 8 tubes.

Try to draw a schematic of the power stage.

Also pictures of the amp could be helpful.

Think is a kt88 talking...
This could be true in respect with dissipation limit some brands stated for fixed bias operation: -35w: 220k / +35w: 100k

[Helmholtz]

Tube retainers with sharp edges?

[catalin gramada]

Tube retainers with sharp edges?

Didn't mentioned.

[Mr_bibbles]

Ok I'll try and addess each of these posts.... First and foremost these are JJ KT88's. The Tubes in beforehand were Genalex.

When tubes draw air (which oxidizes the getter metal layer) the reason is most likely excessive electrodes or bulb temperature (i.e. apart from production problems or mechanical effects).
See e.g. R.B.Tomer's book "Getting the most out of Vacuum Tubes".

So check idle dissipation and watch out for redplating at medium to high output.
Another reason could be thermal runaway caused by high temperature operation with too high Rg1-k (grid leak resistance).
Recommended max. Rg1-k is 100k per tube. So a shared grid leak resistor should be 25k per side with 8 tubes.

I will check idle dissipation, Grid leaks are 2.2k on each individual tube (mounted directly at socket)

Another issue is if pin1 on the socket is used as a tie point for the grid resistor. Now you have a situation where a metal tube retainer (spring or clamp type) can short the bias to ground. This could possibly lead to excess temperatures maybe in the range that Helmholtz is talking about. Could this be enough to compromise the vacuum? I said before that 'it's not impossible' but maybe that is more a philosophical stance that

Your second quote in post #24 was from Mick Bailey . He mentioned instability but again, it was when socket pin1 is used as a tie point, so I don't think it applies.

In any case, I think the idea that any thing electronic could be responsible for vacuum loss, is something we are all very skeptical about. It's much more likely due to some physical issue, but we are just spit-balling other stuff.

I have also never seen an electronic issue white top a tube, but this amplifier has done it to three tubes in different sockets, and I assume it is UNDER PLAYING CONDITIONS AT THE MOMENT IT FAILS. The customer reports a great sounding amp for 10-15 minutes of play before it pops another tube, so there has to be some explanation for the pattern.... mechanical force seems unlikely...

Tube retainers with sharp edges?

There are also no spring clamps on the unit.

One of the items that keeps getting brought up is heat; Heat seems to be an reasonable explanation for a tube losing vacuum... with 8 KT88's, I could see this reaching a point where it affects the tube, especially since the failure has happened in multiple positions....

I don't think we covered it, but if the OT is under-rated at 250W, what could be some of the reprecussions? Even with 6 KT88's running, that's 225W of output, but with 8 we're 50 watts over. There's no chance flyback voltage is using pin 1 for an easy path to chassis?

[Helmholtz]

Grid leaks are 2.2k on each individual tube (mounted directly at socket)

Those should be the grid stoppers.
Grid leaks are the resistors which connect from grids to the bias supply.

[g1]

A fan if it doesn't have one, and some kind of shock-mount for the chassis or headshell (if it sits on speaker cab) would be things to consider.
If they help the problem, great, if not, they can only aid reliability.

[Helmholtz]

A fan if it doesn't have one..

See post #16.