blowing el84s

[jojokeo]

I've got a SE el84 amp which I seem to be able to blow el84 tubes with. Just wondering how easy it is for this tube to get too much signal and go caput? Is this "common" for this tube when driven too hard or sudden high signal hitting the grid?
There's an EF86 in V1 w/ a standard CF stage in V2 that has a 12at7 in it. The CF stage can be switched in or out of circuit. First time el84 blew I had the amp at pretty high volume when I switched the CF stage in while playing. The 2nd time was when I had several preamp/distortion pedals going when I switched them in. Both el84s were new JJ's (atleast they were not NOS types).

[Leevi]

I have built several EL84 based amps both SE and PP and have quite seldom had problems
with the power tubes. I would say JJ and Sovtek are more robust than the Chinese tubes. I normally
bias the tube to 48mA cathode current.
/Leevi

[jojokeo]

Both times a tube blew was when I hit it hard all at once via a switch at high signal. I've cranked the amp many times while playing normally full volumes but it seems that when the tube is hit at once w/ the high signal that this has happened now a second time.

Maybe the reason could be compared to an analogy, like the difference between hitting the accelerator of a car while in gear versus hitting the gas while in neutral and throwing it into gear???

It makes me nervous if something like this was to happen at a gig or show? Maybe another reason to not use el84s if they're so sensitive to being signal-slammed?

[The_Gaz]

I've never heard of this happening, plus think about all the pedals people are running into poor little el84 amps. I think the signal being so high must be part of the equation since it happened twice, but I'd look elsewhere in the circuit than the tubes themselves. What's the screen voltage and value of the screen stoppers?

[tubenit]

What is the plate and screen voltage, please? 

I had a JJ EL84 go bad before in short order & put in Sovtek EL84's and had no further issues.

With respect, Tubenit

[Fresh_Start]

How are the EL84s biased, fixed or cathode, and how hot?

My Blues Jr. ate El84s until I added adjustable bias and dialed it a bit cooler.  No problems since then.  Just thinking maybe if the bias point is dancing on the edge and you give it a sudden slap that might be the problem.  Also, how big are your screen resistors?  Bigger might be better here.

When you say "cathode follower stage", do you mean 2 triodes with one as a common cathode gain stage and the other as the cathode follower or just adding a cathode follower after a gain stage that's in the circuit all the time?

HTH

Chip

[jojokeo]

plate voltage = 289
screen voltage = 261 then goes through 100r/1w screen resistor
cathode resistor = 150r/5w

CF stage is typical two stage w/ 12at7 w/ 100k load resistors w/ 2.2k & 4.7uF cathode & 212v feeding this.

[EL34]

This offers a btter picture of your amps voltages
http://www.el34world.com/charts/valve/ValveData.php

[jojokeo]

T/Y - I only made node voltages on this one back in time. Once I take it apart I'll have to get better/more thorough ones.

[stratele52]

Blowing power tubes is alway a bias too hot or a too low resistor  ( or no resistor ) on EL84 pin 9 grid or output transformer short.

In your case it looks like a bias too hot or low ( no ) resistor on pin 9 or both.

[DummyLoad]

I have built several EL84 based amps both SE and PP and have quite seldom had problems
with the power tubes. I would say JJ and Sovtek are more robust than the Chinese tubes. I normally
bias the tube to 48mA cathode current.
/Leevi

@ what B+?  idle dissipation or current draw @ a specified B+ would be much more informative.

[jojokeo]

It seems people are getting wrapped up in bias conditions but this isn't the problem or related to the issue at hand IMHO. You can see the plate is < 300V w/ a 150r bias resistor as many many amps are or even w/ 130r resistors. There's the typical 8k2 grid resistor and 100r screen resistor too. The tube is "happy".

What has happened several times (when I've "blown them") - I'm hitting the grid too hard and too suddenly w/ too high of a signal voltage. Isn't there a limit on something like this? Has anyone else hit their power tube really hard, at high volume, and had the same thing happen before? (especially w/ an el84 since they have a higher input sensitivity and don't need to be driven as hard as 6v6/6l6 type tubes)

[phsyconoodler]

An oscillation will make the tube go nuts.If your preamp or lead dress is causing it to oscillate then all bets are off with the EL84 cathode resistor or plate voltage.Something is driving the snot out of that EL84 and the preamp is a good place to start.Seems congruent with your use of input signal. What amp circuit are you using?

[John]

I feel weird actually offering advice instead of asking for it, but in TUT3 the man says that EL84 and similar tubes should have individual screen resistors of at least 2.2K/1W. His reason is that most of the amps that use these tubes tend to be smaller and therefore driven hard, which is hard on the screens, causing tubes to fail quicker.

Probably nothing you don't already know, but figured I'd throw it out there anyway, respectfully.

[jojokeo]

Lead dress is good, no oscillations, the amp plays/sounds and sustains excellent as it should for over two years now. There is nothing wrong or out of the oridinary, it kicks butt and more than holds it's own against any SE el84 amp out there and I'm being honest not subjectively biased. It's a simple circuit that was rebuilt and revised many times over in a way that even tubenit would understand and appreciate. I haven't done anything to it in a long time to change anything since - maybe a year and a half? I'm about as anal w/ lead dress, layout, grounding, etc. as any here and assuredly these are not issues.

This "phenomenon" has only happened twice and both times were when the amp was being modestly pushed: I have a pedal board w/ multiple overdrive/distortion boxes, compressor, several boosts, and the other usual stuff etc., I was stomping away hitting the overdrives and boosts in various combinations getting things ready for an upcoming festival and then whack! I thought it was a preamp tube but it ended up being the el84. The first time this kind of thing happened was months prior when the amp was between half to full power and I hit the switch for the CF stage to go from being bypassed, so it slammed the el84 too hard again. Afterwards both times, the el84 sort of wants to try to work but there's like static, instability, much lowered and intermittent volume and operation, etc...it happens instantly too where you go, "oops, I shouldn't have done that?" I then replace the tube and all is normal again and remains that way no matter what I do unless I repeat either of the first two situations. But, I haven't done either again on purpose just to ruin another tube for the sake of repeating it. I think both times I was using a JJ tube too.

When the CF stage is on and even the tone stack is bypassed I can turn everything up full and play as long, hard, and loud as I want. I have a hunch I could do the same w/ the pedal board situation too - but w/ everything being already on? My gut feeling is that the el84's grid or something in the tube can't take being slammed too hard all at once?! This is the only common denominator in both instances to cause the el84 to fail.

[phsyconoodler]

You certainly don't need 2.2k screen resistors.Kevin O'Connor is a very opinionated dude.But that being said,it may be that the screens in the amp are what's being pushed too hard.Maybe try upping the screen resistors and try again.

[jojokeo]

Phsyconoodler, sorry I didn't answer your question regarding the circuit. It's an ef86 going into a switchable CF stage using a 12at7 w/ the usual 100k load resistors & 1k cath. r w/ switchable bypass cap > tone stack (also bypassable via switch) > el84. Three tubes, that's it.

So do you think that upping the el84 screen resistor will allow the tube to take more abuse the way I've described things? Would a Sovtek be more robust, aren't JJ's supposed to be pretty stout tubes also?

[phsyconoodler]

That could be it but I'm not totally sure what's failing in those tubes.It's easy enough to try and flog her for a while.

[G._Hoffman]

Phsyconoodler, sorry I didn't answer your question regarding the circuit. It's an ef86 going into a switchable CF stage using a 12at7 w/ the usual 100k load resistors & 1k cath. r w/ switchable bypass cap > tone stack (also bypassable via switch) > el84. Three tubes, that's it.

So do you think that upping the el84 screen resistor will allow the tube to take more abuse the way I've described things? Would a Sovtek be more robust, aren't JJ's supposed to be pretty stout tubes also?

Merlin also seems to like larger screen resistors, so it seems to me that bumping them up to 1k and giving it a try wouldn't be a bad idea.  If it effects the sound too much, than maybe drop down to 470r.  I'm am FAR from being an expert, but both KOC and Merlin describe failures on the screens to be transient issues, and if it is happening when you switch something on that seems a transient issue to me. 

On the other hand, someone more knowledgeable than me could come in here and tell me I'm an idiot, and I wouldn't argue at all.


Gabriel

[HotBluePlates]

Exactly how snug are the tubes in the sockets?

Now that I read this thread, a thought occurs to me... Some JJ tubes a few years back had quality issues, where the diameter of the pins were smaller than typical. When used, they wouldn't fit into the socket tightly, and as a result, the tubes had intermittent contact with the grid and lost bias. Tubes redplated, people were pissed, yada, yada.

Either you got a couple of defective tubes, there's a QC issue as above, or there's an issue with the details of the use in the amp. I highly doubt it's simply too much signal to the output tube.

[Ritchie200]

Hey Joe,
I noticed on your posted schemo that you show all the impedance taps on the tranny.  Does this mean you have them on a switch?  If so, are there any issues there?

Jim

[jojokeo]

No Jim, they go to output jacks and one has since been converted to a line out.