Voltage Rating(DC) of 9-pin tube socket

[Charlie Wilson]

Hello,
Can anybody recommend which 9-pin tube socket would be best for an amp that uses EL84s and has 550v on the plates. I know that sounds crazy but I have a Seymour Duncan amp with a socket that arched to the chassis and it has voltages that high on the plates(screens half of that).
CW

[martin manning]

It'd be nice if the socket manufacturers would include voltage ratings in their spec sheets, but I can't find any. Micalex seems to be good for 10kV/mm, so the popular Belton brand should be fine.

[gui_tarzan]

550v on the plates of a nine pin tube?

[Leo_Gnardo]

I'd recommend porcelain tube sockets.

[R.G.]

The voltage rating of a tube socket is much more dependent on the surface contamination than on the materials.

Worse yet, if it ever flashes over, the arc's high temp, which may be as much as 10,000 degrees in the arc itself changes the contaminants to oxides of the stuff that used to be contaminants, and that is usually more conductive than the organic gook that used to be there. So it arcs over even lower.

The limit is the breakover voltage of air; it will always be less than that.

[Charlie Wilson]

Thanks Guys, yeah this Seymour Duncan 84-50 has 550v or so on the plates with a quad of El84s. It has much lower screen voltages. The plate pin of one of the power tubes flashed through the socket to the chassis even though there is a slight gap between the socket and chassis. I agree with R.G. that it is now a pathway for blown fuses or worse. I will try the Belton socket.
CW

[Stevem]

I would go the Ciramic route and add the 3 diode set up on each side of the OT.
Note also that there is way more than 550 volts on that socket with the signal swing added on top of it!

[Leo_Gnardo]

The plate pin of one of the power tubes flashed through the socket to the chassis even though there is a slight gap between the socket and chassis.

Breakdown voltage of air is generally taken as 8000V per inch. And that's for dry air at sea level pressure. Humidity and loss of pressure with altitude derate that figure. If your slight gap is less than 1/8 inch you can see it's an invitation for an arc, even more so if the area in question has any accumulation of dust.

Steve's suggestion of adding flyback quench diodes/rectifiers is a good one. Been standard issue on MusicMan amps run with 700V on the plates ever since the 70's, and you find 'em on a lot of standard issue Peaveys and Fenders over the last 25 years. You can seek individual diodes with 3000V or more breakdown rating, or link 3 common 1N4007 in series as suggested by Ken Fischer in his Trainwreck Pages. Voltage on the plate can spike well in excess of supply voltage when the amp is driven into clipping. The flyback quench circuit keeps plates from traveling beyond B+ and tempting an arc. It's not an average 8000V in that breakdown spec, it's instantaneous, and just like lightning a lot of current can follow a "leader" arc so all it takes is one to get the game going.

Might be a good idea to also put a layer of fish paper or cambric insulator between the chassis and plate tube socket connections. In college hi voltage lab we used GE RTV cement for spot insulation & it held up to at least 25,000 V. Same stuff you find in the hardware store. Given the location of your arc it doesn't so much matter what the tube sockets are made from, but I'd opt for ceramic/porcelain anyway.

[Charlie Wilson]

Thanks again, I guess part of my job repairing the amp is to make it as safe as possible to use(even if the manufacturer didn't), so I am going to replace all of the tube sockets and add the diodes.
CW

[TUBEDUDE]

How the blazes can an EL84 survive a Vp of 550? Is Rk at 20k or more?

[Charlie Wilson]

How the blazes can an EL84 survive a Vp of 550? Is Rk at 20k or more?

It is grid biased but has pretty low screen voltage.
CW

[Stevem]

I have never heard one of these amps, but I would have to say that with the needed low screen voltage to have them live that they likely do not sound much like 84s anymore and in my mind at that point why no use another tube?

[TUBEDUDE]

Why? Did those Koreans find a deal on transformers? I guess it's just me, but I like the lower (reasonable) voltage/higher current approach. This seems like a like a water hose with 80 PSI on the line and a small diameter sprayer to contol flow. That hose is gonna split eventually.

[pdf64]

My finding is that plate-to-ground reverse biased OT protection diodes mess up the overdriven tone. On checking, the waveform change using a real speaker load is clear on a scope.
Try putting a DPST switch in their ground return and seeing if you can hear their effect.
I'd use them if the application requires such mitigation, eg crazy high VB+, but for more sensible designs I'd stick to a high value resistor permanently wired across the secondary, so as to eliminate the possibility of an open circuit secondary, and resultant excessive flyback.

[Leo_Gnardo]

My finding is that plate-to-ground reverse biased OT protection diodes mess up the overdriven tone. On checking, the waveform change using a real speaker load is clear on a scope.

We can't fly one past you can we, you got your ears on mate! Shaving off those flyback spikes does change the overdriven tone, no doubt of that. What I find, and this is subjective of course, my opinion using my ears, is that having the flyback diodes installed removes a small but nasty & spiky aspect of the overdrive tone and thereby sweetens it. Now of course being a fair minded person I have to say "that may not be to everyone's liking." I have demonstrated the difference to some of my golden-ear customers, and surprisingly some can't hear it, maybe they've been listening too loud, too long, and besides losing their hi frequency hearing due to age (presbycusis). The ones who do hear the difference 100% opt for the tone with the flyback diodes installed.

There's something to be said for the flyback diodes extending the life of output tubes & transformers too, a definitely positive effect, if your ears can stand the tradeoff.

Of course you have to use a real speaker load to generate those flyback spikes, as they are a result of the speaker voice coils moving thru the intense magnetic field in the vc gap. Speakers can act as generators, not only as motors. The flyback spikes don't occur when driving resistors.

EL84 rated up to 800V IIRC. (!) Of course not much current can pass in that situation, not much amenable to guitar amp use. I think they were used in some Euro made TV's back in the day, in a function where they only had to pass a brief voltage spike, IOW low duty cycle. Forgive me for not being totally conversant with television receiver circuits, I pretty much ignore that part of electronics.

Noting Tubedude & stevem's recent comments, yes they really should have used another tube, and yes the tone goes to hell when you panel beat a circuit like this into working, sort of, with EL84. At the time these amps were made, some guitarists were aware of the sweet tone available with '84s and others were being made aware thru magazine articles. Naturally some amp builders had to push EL84's to and beyond their taste limits. Boogie was also guilty with their .22 and .50 amp series using EL84's, the (Laney?) Velocette as well, not to forget some power amps issued by Peavey and Mesa. Trying to satisfy the need for that sweet '84 tone AND make an amp loud enough to cut it with a drummer plus the rest of the band, I don't see where it was ever really done successfully. I see the "good ol' 275-325V B+" EL84 amps vs. those with over 400V B+ as a parallel between street and racing cars. Sure you can alter your sweet ol' street car with racing kit, big engine, hi gear transmission, all the rest, but then it's fit for track use & won't be suitable any more for everyday driving.