EL34 Failure

[PeteM]

Anyone care to help why two sets of valves should fail inside 18 hours of total use.

The amplifier is a Yarland EL34 single ended ( yeah its Chinese made but read on) The valves that have failed one was an Electro Harmonix heater filament gone open circuit the amp was on at the time and one channel just faded away silently no pops or bang. The other failure was a JJ El34L lost vacuum over 2 days of being used I noticed the getter had turned milky before switching on.

Before we blame the amplifier It has run fine day in day out with the Chinese supplied valves, it has also run for 4 days non stop in the middle of the garage with just load resistors away from flammables and with an RCD fitted. I also have some unknown re branded EL34's that red plate easily yet these show no signs of stress without bias adjustment.

From the vendors point of view this smells fishy 2 different brands fail from the same customer within a short period, I have to agree it does look suspicious but I can not get the amp to destroy the cheap valves, furthermore the valves that failed were brought up slowly by a variac.

I can not see beyond the obvious any fault with the amplifier it runs fine even the power transformer is only warm after a full days play, I have had 2 multimeter's connected one giving a stable 6.2 Volts heater supply, the other sitting across the 10 Ohm cathode resistor giving a voltage of 0.64 Volts B+ is 279 Volts

I should add the vendor gave me a full refund but would like some help before I buy further replacements.

Regards

Pete

EDIT.. I have uploaded the schematic http://www.petemoore.pwp.blueyonder.co.uk/audio/FV34A.jpg and should add the failures were on opposite channels

 

[Alastair E]

Personally, I would say that you have just been VERY unlucky....

The two valves failed in different ways. I doubt that your equipment in any way 'killed' them. Maybe it was the shipping, stresses caused in transit, who knows....

 

[PeteM]

Personally, I would say that you have just been VERY unlucky....

Thanks for the reply Alastair.

I'm trying to see this from both points of view, the vendor thinks my bias was way off causing over heating.

If bias was way off would I have not seen plate glow in a dark room, I know for a fact the cheap Maplin Edicron re branded I purchased years back glowed in a World Audio design amp. At factory bias setting of the Yarland I have no glowing plates.

Correct me if I'm wrong bias on the Yarland is voltage across the cathode resistor. ie.. 0.64volts across 10 Ohm resistor is an idle current of 64mA * B+@ 279V = 18 Watts

Thought EL34's were good for 25W and the sample I purchased were suppose to be more bullet proof.

 

[Tweeker]

What the vendor really thinks and what the vendors says are not necessarily the same.

 

[PeteM]

What the vendor really thinks and what the vendors says are not necessarily the same.

True . I sent both sets back including the 2 good ones so he could see with his own eyes. Regarding the JJ's lost vacuum there was no sign of a cracked envelope it had to have leaked around the glass to metal seal, furthermore I have seen tubes lose vacuum while powered up this JJ was clean and no burnt filament.

The vendor had his reasons it would of been cheaper for him to replace 2 tubes but he insisted on a full refund.

Regards

Pete

 

[mississippi]

I read the story, and I gave a look at the circuit diagram.
I only have a small remark at the diagram, I don't think this could kill the tubes but some oscillations could appear.
The secondary winding (speaker side) is forgotten to give at a connection to the ground. The feetback is so not working!

 

[PeteM]

The secondary winding (speaker side) is forgotten to give at a connection to the ground. The feetback is so not working!

Not sure what you mean.

The feedback is screened cable and grounded one end.

 

[Frank Berry]

For the feedback to work, one end of the secondary must be grounded. According to the schematic, it's not.

 

[consort_ee_um]

The feedback connection is floating, the other end of the output transformer secondary winding must be grounded to have any effect.
Didn't the famed Marshall amps run without feedback to get a fantastic mid range presence?

 

[PeteM]

For the feedback to work, one end of the secondary must be grounded. According to the schematic, it's not

Thanks for the heads up

I think there is a discrepancy with the schematic, I will open up the amp and check. I know the power supply is much beefier than shown reason why I did not post that part.

I have tried again to blow up these bad EL34's I have, running them hot has not done anything.

Trouble the dealer planted a seed it must be my amp, I have taken his comments to heart.

Regards

Pete

 

[richwalters]

I accidentally s/c the o/p of my newly constructed AB2 amp using 6550B's. Bmax was only 360V but working late at night I misread the dvm current-wise. The amp is only rated a nom 20W output but I found that each tube was taking 200mA and "near meltdown". With a hot reset everything was back to correct settings and no tube suffered any adverse effects. << anyone beat that with a mosfet amp ?>>.In fact with the heat, reverse grid bias leakage reduced with better getter function. ((....one simply doesn't s/c a class AB2++. It will rebel and dump amps)).

On MI service once I blamed a poor batch of EL84's and found minute cracks in the glass where the jock strap or retaining chassis clip makes contact with the glass. Repeated exposure to draughts created thermal cycling vs. metal and glass over time eventually leading to stress cracks in glass.
I think with any mass produced item such as tubes, an element of risk always happens. Ask the vendor to run-em' in before shipment.



richj

 

[rcavictim]

Your problem is just crappy tubes. I have had nothing but trouble with Electro-HARM-onix. You may find that if you put a soldering iron to the filament pins of the one that went open filament that it might come back to haunt you with a different kind of failure in a few short hours.

 

[PeteM]

Thanks for all your advice.

I will look at the feedback and see all is well, I guess the ultimate out come was a refund but I wanted just a pair of replacements.

What is strange I have purchased a good number of tubes over the years from this guy, and never returned or questioned a single one of them. Yet he was so fixed on not selling anymore to me until I fixed my amp, I'm OK on the basics and can follow known schematics as seen here

An externally hosted image should be here but it was not working when we last tested it.

 

[Jim McShane]

Your problem is just crappy tubes. I have had nothing but trouble with Electro-HARM-onix. You may find that if you put a soldering iron to the filament pins of the one that went open filament that it might come back to haunt you with a different kind of failure in a few short hours.

I'm sorry you've had some trouble, but I want people to know that your experience is not universal nor particularly widepsread -at least in my view.

FWIW, I have the lowest failure rate on EH tubes of all the lines I handle. To be fair the Chinese tubes are even lower, but I only handle a small range of Chinese vs a lot of EH. JJ is the highest failure rate by far.

Again FWIW, I've fixed a number of "dud" NOS tubes by resoldering pins - it's not unique to current production stuff.

It sounds like the original poster simply had some bad luck with two tubes.

 

[Ragnar]

Grounding the second secondary tap of the output tranny is very important. I suppose that the gronded sec. tap is also some kind of phase courier, that say "this is the ground and the other is the live one, never to be reversed"
I believe, without grounding it, that kind of "feedback" would only generate inaudible oscillations that could damage (maybe oveload) the tube. Many people don`t realize, that frequencies higher than they can hear could be dangerous for the tube in a particular stage. I hope you grounded it. There are only two ways: fully working and safe feedback or no feedback at all (I personaly like this one ). system shown on the circuit is probably any good...imho