Grey smoke from a KEL84 - Help!

[misterc6]

This morning I decided to try a different set of 4 EL84s in my KEL84. Lovely music for about 1 minute then a crack and the dreaded grey smoke.

I have looked inside and it's clear that R29 and R31 (100R 0.5W) are badly burnt, C15 (100u/35v) is bulging at the top, and I suspect that C13 has also suffered some damage. I now think that R27, and possibly R25 (270R 1W) are also damaged.

Questions:
Is that likely to be the extent of the damage?
Will the Output transformer be damaged and how can I test this?
When I've replaced the components what should I check before refitting a known good set of EL84s?

I have replacements for R29 and R31, but not R27 and R25, nor the capacitors. Is anyone able to help me with some replacements? I will of course pay for these and the postage.

TIA,
Malcolm

[John Caswell]

Hi Malcolm
From your description the EL84s you put in were faulty, as that have taken out the screen grid g2 oscillation stoppers and also the cathode k bias Rs and Cs, the valves were probably Screen grid g2 to control grid g1 short circuit
I doubt very much the OPTx is damaged as the 100Rs would probably have protected it and also you switched off pretty quick.
PM with address and I will see what I have in stock.

John

[misterc6]

Originally Posted by John Caswell

Hi Malcolm
From your description the EL84s you put in were faulty, as that have taken out the screen grid g2 oscillation stoppers and also the cathode k bias Rs and Cs, the valves were probably Screen grid g2 to control grid g1 short circuit
I doubt very much the OPTx is damaged as the 100Rs would probably have protected it and also you switched off pretty quick.
PM with address and I will see what I have in stock.

John

Many thanks, John.

PM sent

[Richard]

Hi both, I've got a set of 4 x 100uF 35V cathode caps if you want them just drop me a pm

[John Caswell]

Hi Malcolm
Various resistors in the post Tuesday - you can be as pedantic as you like over them.

John

[misterc6]

Thanks, Richard, that's very kind. I should be OK with the one spare Rubycon I have here which is the same as the remaining three. I'm trying to minimise the number of components I have to change. Please reserve the four you have for me and I'll let you know if I need them.

Thanks John, I look forward to receiving the package.

Regards,
Malcolm

[bikerhifinut]

Ah the results of plugging in unknown "toobs"!
I take it only one channel went down?
The KEL84 runs the EL84 quite hard so if they're on the way out it could be interesting.
Now I fettled up a KEL84 a good few years ago, I remember amongst other things having to sort out a previous "improvement" where the heater rectifiers had been subbed for Schottky diodes with a low Vf which meant I had to work out the resistor value and make a simple pi filter to get the HT from something over 7V down to 6.3 on the ECF80, along with as I remember it replacing the screen grid stoppers (your 100R) that looked as if they'd got a bit warm at some point.
Now I also replaced the 270R cathode resistors for beefier 2 watt jobs as I feel 1W is just a little too close to the actual dissipation of around half a watt. And stood them well clear of the PCB. Which you'd do anyway.
I'm not one for swapping out parts just for the hell of it or even on a time basis for electrolytics, modern good quality caps arevery robust but in your case I'd replace all the bypass caps just in case.
Regards
Andy

[Richard]

Hi Andy, all,


Yes it runs them at 12W plate which is max for Mullard. I don't know if modern valves are better or not but I had a Leak St20 which ran them the same and they always looked very hot to me. If I'd kept it I'd probably have tried biasing it down, perhaps trying 330R cathode resistors. Anyone done this and checked the resulting voltages and dissipation?

[bikerhifinut]

EL84 always look a bit blackened on the inside, Mullards were the worst for it but never seemed to affect their general robustness which is very good if rununder their maximum ratings.
Richard I think you'll find that leak revised the St20 to runa lower HT voltage as at 330V you're really pushing your luck along with any other parameters.
I wonder if thoseof us with High Mains voltages are getting slightly too high HT Volts, that could be sorted with an RC filter after the second HT cap downstream of the choke? That'd do more good than fiddling with the classic 270ohm cathode bias.
As I remember it the KEL84 only runs 10% UL taps which allow it to produce nigh on 15W at clip? that's really hammering the poor old EL84 which really is best at about 10w PP. Only my opinion of course, and the 10% taps mean they're pretty much running as pentodes anyway. Leak ran as I remember 20% on early amps and 43% on later ones but I'll check my data with my "Leak Bible"

andy

[Richard]

They were similar Andy, mostly following the Mullard data sheet values. ST20 ran 318V to the OPT CT (with 20% UL iirc) and Kel84 runs 310V with 12.5% UL. Both use 270R at the cathode. That gives 12W plate dissipation so when Mullard say 12W is max plate dis for EL84 there’s not much room for poor valves or high mains is there. It leaves me thinking they were showing what the valve is capable of not what you might want to do with it.

I’ve changed bias on KT88/6550, EL34, and 300B amps to save the valves and noticed no loss of sound quality but haven’t tried EL84.

Change the resistor value a little more than might be expected, it’s easier done in practice than trying to work it out before hand. The voltages see-saw, and the less current drawn unloads the mains TX a little and HT goes up a few volts, so check them again when they’re all done. An extra 20% on the cathode resistor value is worth a first shot then check voltages and see if it puts the valves in a safer area. If Mullard say 12W max dis I’d like to see 10W but maybe that’s just me