EL84 Arcing Failure

[scstill]

Not much experience with tubes hard failures....

Son was playing the Pro Junior, said it made a loud noise and stopped working.
The Pro Junior had a blown fuse. When replaced one of the el84 was arcing like crazy.
Replaced the el84 tube pair with a known good pair and all seems fine.

Could the tube have failed on its own? Or should I look elsewhere in the circuit?
Could the tube arcing have caused the fuse to blow?

[Latole]

1- Could the tube have failed on its own?
 Yes

2-  Or should I look elsewhere in the circuit ?

Yes. This amp need Output Power tubes bias check and adjustment.
Never put news tubes without check bias; In the worst case, they may not last long time. In the best case sound may not be at is best.


3- Could the tube arcing have caused the fuse to blow?

Sure it is . Replace by the same fuse, if not you may destroy power transformer ; 2 Amp fast blow ( or regular, no slow blow or time delay)

How to bias Pro Junior;

[scstill]

The EL84 plate voltage is 316v
The plate current is 31ma on one and 34ma on the other (measured across the OT primary)
In the video, the fellow was trying to get 25ma on 330v. Why 25? Would I be ok leaving it at 30+ or too hot?

Seems like a crappy design if you need to replace the resistor whenever you put new tubes in.
How about replacing R29 with a 10k resistor in series with a 10k trim pot (like Hoffman redo)?
Leave everything else the same. This gives 10k to 20k bias adjustment. Or would greater range be better?

[sluckey]

Your tubes are idling at 89%. IMO, that's too hot for fixed bias amps. Most people like to run at 60% or 70%. Tubes will last longer.

The YouTube guy knows from experience that these amps run too hot.  He has decided that 25ma is his happy place (notice he did not end up with his happy place, but only close). I suggest you lower your dissipation like the YouTube guy or follow Hoffman's lead and add a bias pot.

[Latole]

I run between 50 and 60%. Never at 70%, running amp at high volume may ( will ) arm tubes.

Test sound with guitar, you won't hear any difference between 50/60% and 70%

[pdf64]

I think that these percentages being bandied about need a context.

For a given valve type, 70% of its design max anode dissipation limit is 15-20% hotter than 70% of its design centre anode dissipation limit.

Given the widespread conflation of design max and design centre limits, no wonder there’s misunderstanding and disagreement about whether 70% or whatever is too hot.

EL84 have a 12W design centre limit, so fixed bias idling up to 70% of that, 8.4W should be ok.

If using a design max limit, eg 14W of GE 6V6GTA, idling up to 60% of that may be a better policy.
Oh look, that works out to 8.4W

https://tubedata.altanatubes.com.br/sheets/010/e/EL84.pdf

https://tubedata.altanatubes.com.br/sheets/135/6/6V6GTA.pdf

[shooter]

running amp at high volume

this is also overlooked, if you're doing a slow ballad the tubes probably won't swing much voltage
if you're doing 120BPM the voltage swings are huge, and the tubes are hot


most datasheets show max drive signal, most builders don't measure it
when voltage swings by 2X your idle volts and you only bias for idle instead of playing style, tubes suffer the oversight.

[scstill]

most datasheets show max drive signal, most builders don't measure it
when voltage swings by 2X your idle volts and you only bias for idle instead of playing style, tubes suffer the oversight.

So if you are a loud rocker, you should bias even lower than previously suggested?

[acheld]

So if you are a loud rocker, you should bias even lower than previously suggested?

It depends . . .   If you're a rockstar with a dedicated amp tech, and you like the sound of your amps biased hot, well then, go for it.  You'll keep your tech and Tung-Sol happy.

If you want your tubes to last long, bias them as low as you can whilst keeping your tone.  As shooter rightly points out, playing style matters. 

But I'm a fan Latole's idea -- bias lower, then check your tone, rinse and repeat.   Personally, I can usually hear the difference when biased in the range of 50% or lower (fixed bias of course, that's what we're talkin' about), but every amp is different.  I generally bias my amps at roughly 60%, but I do check for a lower limit when I build them.

Everyone has a different thought on this topic, it seems.

[PRR]

> if you are a loud rocker, you should bias even lower

Plate dissipation can fall if you can play LOUD NON-STOP. See graph.

[scstill]

Installed a 20k trim with a series 10k fixed in the R29 leg
Adjusted to get 7.1watts (322v 22ma) on blue and 7.7 watts (322v 24ma) on brown.
So idle should be 60-64% of 12w max.
Compared to another stock PJ, I cannot tell much difference in sound, but glad the tubes will like it.

Since it was opened, I also added a NFB on/off switch.

Years ago I added a 8ohm 50w L-Pad so I can run at breakup settings and dial back the loud.
The L-Pad mounts to the back panel with hidden screws so it can be removed without tolex damage.
Panel laying on the cab bottom in pic.