Safe Plate Voltage For Unbiased 6L6GC?

[tbryanh]

To prevent the power tubes from being harmed during idle when the bias supply fails, what is the highest plate supply voltage that can be used in this amp (100W Silverface Twin Reverb)?

http://www.ampwares.com/schematics/1...F_Twin_Rev.pdf

Also, what is the maximum drive signal that can be applied to the PA without harming the tubes during this condition?

[Chuck H]

With no bias voltage, won't the tubes try to draw more and more current until they fail regardless of plate votage?

A more cumbersome but reasonably fail safe bias method would be to use a cathode bias but with a zener across the resistor to "fix" the bias. The amp would behave as if you lowered the plate voltage by the amount across the cathode bias circuits though. But you won't get any cathode bias "sag" with the zener. I would use two cathode bias circuits. One for each "pair" of tubes.

Chuck

[Merlinb]

To prevent the power tubes from being harmed during idle when the bias supply fails,

Why would the bias supply fail? The standy switch doesn't affect that...

[Enzo]

Bias supply failing is way down my list of likely failures. What about when a rectifier shorts and sends AC to the filters? What about when a filter cap shorts off the B+ supply? What about when tube sockets arc?

In the absence of bias at the grid, the power tubes turn into diodes and will conduct as hard as they can. Plate supply voltage doesn;t much matter, the tubes will draw as much current as the B+ supply can create. Your tubes will red plate regardless of signal.

[jrfrond]

Fixed bias supplies usually don't fail catastrophically, though I have seen bias diodes open up and series resistors fail. Yes, shit does happen. Most of the time, it's open caps, and the result is hum caused by ripple on the bias supply.

Cathode bias can circumvent the catastrophic failure of the bias supply itself, but NOT the opening of series resistors.

A possibility would be to construct a circuit that monitors plate current draw and bias supply, and when plate current surpasses the maximum threshold or bias disappears, the circuit switches a relay that toggles to cathode bias as a safeguard. If the bias disappears at the tube grid, but not the bias supply, it can toggle a relay that opens up the power tube cathodes and shuts down the offending tube, with a resulting sonic degradation. Now, I just came up with this off the top of my head, and while it IS an eminently possible design (it would require a logic circuit), to me, it's just a better mousetrap.

There is no "safe" plate voltage for an unbiased tube. No matter what the voltage, it will try to hog all of the current it can. You CAN fuse the cathodes though. However, given the reliability of fuses, I wouldn't want that in my amp, personally.

[Merlinb]

A possibility would be to construct a circuit that monitors plate current draw and bias supply, and when plate current surpasses the maximum threshold or bias disappears, the circuit switches a relay that toggles to cathode bias as a safeguard.

Or how about a fuse!

[loudthud]

Kevin O'Connor presents several circuits for "electronic fuses". Basically a large MosFet in the cathode with an op-amp that monitors current that disables the tube if current goes too high. Sorry, I don't remember which book

[jrfrond]

Or how about a fuse!

If you are talking about the power supply fuse, it might be too late by the time it blows.

If a tube redplates because it's faulty, the tube needs to be replaced anyway, but what I am getting here is a way to shut down the tube itself if it loses bias.

[Merlinb]

If you are talking about the power supply fuse, it might be too late by the time it blows.
.

I can see your concern, but valves are pretty hardy. They can usually survive redplating for many seconds- long enough to blow a properly rated fuse.
(As we all know, solid state components, on the other hand, are there to protect the fuse!)

[Chuck H]

I'm picturing the circuit, and just to be clear for anyone not picturing it... It's just a cathode resistor bypassed by a fuse. The fuse keeps the cathode at ground until too much current blows it and then the cathode resistor restores bias to the tube.

It wouldn't help with an open grid or reference resistor though. But it would blunt the blow a bit.

Is there a fuse/breaker type of thing that will "switch" as opposed to just going open? I think that would be very useful.

Chuck

[Enzo]

It still seems like we are going to a lot of effort to protect against just this ONE posssible circuit failure.

I'd say tube failures are a LOT more common than bias supply failures. The fuse/resistor deal would probably answer the original question, but in the case of a failed tube, why would we want it to continue in operation with the new resistor in its cathode? We'd want it gone completely. Better three tubes out of four running then three plus a broken one with resistor.

And what fuse would you use?

[Chuck H]

Actually Enzo, I asked the fuse/breaker question for my own benefit since we were on the subject. Not realated to any elaborate curcuit for this application.

And I agree with you about there being no point in keeping a dying tube in the circuit. So how about just taking it out. But not just the one tube. Both tubes or two of four tubes. Someone mentioned that given the poor quality of fuses they wouldn't want them in the cathode circuit. But I actually think it could be a nifty idea. If you had a four tube amp and fitted each pair (one tube from each side that is) with a fuse, and a failure occured, the fuse would blow and leave the other pair operating in push/pull harmony. Albiet at the wrong impedance. But you would probably get through the show. On a two tube amp it would of course render it useless. But it could prevent a more expensive failure.

Not trying to be too elaborate, just thinking out loud.

Chuck

[Merlinb]

Someone mentioned that given the poor quality of fuses they wouldn't want them in the cathode circuit. But I actually think it could be a nifty idea.

Actually I do that in my amps! One fuse, shared between the two output valvev' cathodes. The only difference is, I put a large resistor+LED in parallel with the fuse, as a failure indicator. But I guess if you were afraid of the bias failing during a gig, then your idea of letting the valves run on cathode bias instead, would be useful.

[Chuck H]

I put a large resistor+LED in parallel with the fuse, as a failure indicator. But I guess if you were afraid of the bias failing during a gig, then your idea of letting the valves run on cathode bias instead, would be useful.

Well, as Enzo said, why keep a failed tube in the circuit? Your circuit with the indicator is a terrific idea. You could set up with the cathode resistor to act as bias AND use the LED so the failure would be recognized. Just run the LED off the cathode voltage as your doing it now. I'm not familiar enough with the possible failures as none of my amps seem to fail for any reason other than builder error So I can't say which would be best.

Chuck

[Enzo]

Thinking back over repairs, I don;t recall the last time a bias supply up and failed, not altogether. What I do see time to time is a bad filter cap in a bias supply. That results in hum, and an effective reduction in bias voltage. The tubes run hot of course, but there is no immediate likelihood of tube failure. I am not so sure our efforts here would respond effectively to this far more likely threat.