EL84 physical damage due to voltage spike?

[Diabolical Artificer]

I was over another forum earlier where the OP posted a pic of an EL84 with a crack in it & evidence of a short, see attached pic. The OP said this was caused by a voltage spike caused by back EMF on a choke or powering the amp with no load. The OP's first language isn't English so maybe that's incorrect.

Has anyone seen a valve physically damaged by an electrical fault? By that I mean envelope damage rather than an internal electrode. Also by a fault I mean the sort of fault you'd find in an audio amplifier not a 100kw transmitter.

Just curious, Andy

[Cruisin Marine]

I have personally never seen such damage before.

[high_vacuum_house]

I would say it was the valveholder that has badly tracked or the contact was very poor causing arcing between the pin and the socket causing a lot of heat being generated on that pin.
Christopher Capener

I think that is pin 3 which is the anode, so there cold be a high potential to other electrodes or chassis potential there.

[Boulevardier]

Originally Posted by high_vacuum_house

I would say it was the valveholder that has badly tracked or the contact was very poor causing arcing between the pin and the socket causing a lot of heat being generated on that pin.

That was my first thought as well!
Mike

[trobbins]

Pin 7 of EL84 is the anode.

Attached is photo of octal socket arcing from pins 3 to 2 (anode to heater). The pin 3 to pin 2 arcing mechanics are somewhat different to just an anode connection arcing I think, but with a bit of thought it may be reasonable.

The EL84 fault is quite plausible imho. The blackening looks like heat stress, which perhaps then caused the glass cracking nearby. The heat stress may have resulted from continued microarcing between valve pin and valve holder terminal, due to the relatively high dc voltage available. The forcing voltage to initiate an arc could be sufficient just from the anode's idle voltage at B+, or a normal signal swing to 2x B+, but can be higher by transformer action in a PP configuration given that leakage inductance can cause a form of flyback at the end of a period of over-drive conduction in a low frequency squarewave (eg. no speaker load condition but input signal overdrive, or even just a guitar overdrive signal with a speaker and a not-so-hifi output transformer). The no speaker load condition can cause relatively high levels of anode current flow, which I guess could cause sufficient energy per arc event to be dumped into the pin region, as the energy is related to the primary winding inductance, rather than a leakage inductance from a 2x B+ flyback arc.

[trobbins]

Other undisclosed faults that could lead to pin/terminal heating could relate to a leaky coupling cap causing anode to conduct no matter what, or similarly a gassy EL84, or even a shorted turn in the output transformer primary (dropping inductance and causing anode current to rise whenever input grid allowed it). Splayed prongs in a socket terminal are not uncommon due to age/use.

[Valvepower]

Hello,

I’ve seen guitar amplifiers with similar damage to the valves and their bases. This was usually because the speaker became unplugged, and the guitarist wound up the wick thinking it’ll clear the problem and that was it…

Back in the 70s a friend’s girlfriend accidentally poured some Patchouli oil in the top grille of his Marshall amplifier. This eventually found its way to the valve bases and hey presto they went intermittent and were arching over something rotten. Fitted a new set of McMurdo bases and EL34s (when you didn’t have to re-mortgage the house to buy Mullards) and all was OK – however, boy, that amp stunk for ages!

I was chatting to someone who recently restored an EL34 100Watt amplifier with a pair of EL34s running with 950V on the anode and 10K a-to-a load (yeah, that configuration ) and not surprisingly arcing had occurred on the valve bases as it had been stored in less than favourable conditions, plus the output transformer was groggy. From memory Mullard/Philips specified a special high voltage Octal valve base for this configuration.

Terry.

[kalee20]

It's clearly the anode pin which is the arcing culprit.

My thought is the valveholder broke down, tracking occurred, the phenolic material carbonised and became conductive, and the heat generated cracked the glass. Thus the EL84 was a consequential failure, not the prime failure.

Even the valveholder may not have been the initial failure - it could have been an open-circuit load. Particularly for single-ended output stages, the output transformer looks like an induction coil, and really high voltages can be generated if the valve is switched on and off by a high-amplitude signal on its grid.

[Diabolical Artificer]

Interesting: the consensus seems to be either a physical stress fracture due to an iffy valve holder or an overload of some sort which then caused the crack.

The only time I've seen similar cracks but with no evidence of a short is in Tektronix 500 series scopes that have been stored in cold conditions where the valve was seated in a socket - a temperature differential fault?

I've damaged some power valves by running them over spec but have never had an incident where an electrical fault has caused physical damage. I tortured an old EF80 once giving it over a kilovolt on the anode & was disappointed that it didn't go bang or do something interesting. I guess though if you dump a lot of volts or current into a valve in one go there's enough energy to cause physical damage.

Andy.

[Richardgr]

I suppose the proof of the pudding will be if the poster on the other forum posts again in a few weeks with another kaput EL84. Then it was the valve holder.