PP EL84 Build Questions (Bevois Valley)

[cgrums]

My previous thread helped to educate me on the design of Morgan Jones' 'Bevois Valley' PP EL84 amplifier. I've moved on to the build and am prepared for a number of build related questions.

I've wired the power supply, input stages, phase splitters and output tubes. I have yet to do the chassis work to wire the input/pot and the output/NFB. My power supply voltages stabilize close to their design values of 320V and 285V and all tubes glow. So far so good.

First problem: I've burnt through two 47R resistors connecting the screen of one output tube to its OPT connection. Each burnt resistor was accompanied with a brief light show in the tube (probably not good). I moved the tube to a different socket and that problem has abated.

Second problem: the left channels output tubes have cathode voltages of only 7V-9V while the right channel has one tube at 14.5V and the other at 25V. The design value is 11V. I've checked the resistance between all cathodes and ground and each is at 270 ohms as it should be.

Bad output tubes? They're a matched quad of new JJ's. What should I be looking at next? TIA

[kevinkr]

Wiring errors and the coupling capacitors..

[gingertube]

Kevin has it right -
The screen resistors going up in smoke is indicaitive of the output tubes trying to conduct too much current. That is also indicated by at least one of the cathode voltages you showed ( +25V).

The Bevois Valley runs a concertina splitter so one side has approx +90v and the other approx +180V at the splitter outputs. When I see the output tube on one push pull output side trying to conduct much more than the other I suspect the coupling capacitors from the splitter to the output tube grids. Are you using at least 400V DC rated capacitors, if not change them immediately. Are they new and not something pulled from some old gear?

As a check - with the output tubes removed measure the DC voltage between grid 1 (pin 2) and cathode (pin 3) on the tube sockets. You should read 0V. If its more than 0.1 or 0.2 volts then you have a problem, either a wiring error or leaky coupling capacitors. Don't plug the EL84s back in to you find and fix it.

Cheers,
Ian

[cgrums]

Thanks very much for the replies. I'm using Wima MKP 630V rated caps and they're new. I'm going to double check all of the wiring and measure the grid 1 to cathode voltage in the output tubes with the tubes removed tonight.

[tubelab.com]

While you are making measurements check the plate voltage on the EL84's at the socket, without the tubes. A tube operated with screen voltage and no plate voltage will act in a manner that you describe. It won't live long either. OPT wiring errors would be the most likely cause if there is no plate voltage.

[cgrums]

Originally posted by tubelab.com
While you are making measurements check the plate voltage on the EL84's at the socket, without the tubes. A tube operated with screen voltage and no plate voltage will act in a manner that you describe. It won't live long either. OPT wiring errors would be the most likely cause if there is no plate voltage.

Sometimes it hurts to have a brain like a sieve. I remember now that my first inspection did turn up a mistake in the wiring of the tube that was burning through grid resistors. I accidentally connected B+ to pin 6, not pin 7 (anode). Fixing that did indeed stop the resistor burning issue. Unfortunately the disparate cathode voltages were measured with the wiring error fixed. I'll check the other voltages soon and report back.

[cgrums]

Measurements showed the valve with the 25V cathode reading to have 300 mV between grid and cathode. Re-soldered each and all is well. Cathode voltages are now 11V + 6V and 8V + 8V (plus represents single channel).

My 285V line has dropped to 255 and the 320V rail has dropped to 305V. If I'm doing the math right that means I need to add ~882 ohms to the 285V rail and ~ 90 ohms to the 320V rail. Do these values seem sane?

[kevinkr]

Originally posted by cgrums
Measurements showed the valve with the 25V cathode reading to have 300 mV between grid and cathode. Re-soldered each and all is well. Cathode voltages are now 11V + 6V and 8V + 8V (plus represents single channel).

My 285V line has dropped to 255 and the 320V rail has dropped to 305V. If I'm doing the math right that means I need to add ~882 ohms to the 285V rail and ~ 90 ohms to the 320V rail. Do these values seem sane?

No, first I suspect you might need to replace one of the EL84 that was apparently previously damaged. (The one with the 11V cathode reading.) For best results replace both with another matched pair. The mismatch between cathode voltages on the first pair is pretty major at over 90%. Once you do that then you can make the required adjustments to the supply voltages.

Your description of the cathode voltages is a bit unclear so it is possible I have misinterpreted them. (what does this mean : plus represents single channel??)

[cgrums]

Sorry about the confusion. I meant to indicate that one channel has output tube cathode voltages of 11v and 6v and the other channel is 8v and 8v.

[kevinkr]

Originally posted by cgrums
Sorry about the confusion. I meant to indicate that one channel has output tube cathode voltages of 11v and 6v and the other channel is 8v and 8v.

Ok, so I did interpret that correctly.. I assume that tube was the one that was originally in the miswired socket. I'd get a replacement pair for that channel.

FWIW it is always a good idea to buy a few extra. Accidents are common as you have found out, and early life failures are not very uncommon with modern production tubes either.