Am I burning up my EL84 tubes with a 439V anode?

[poweredbypoptarts]

A while back I started a thread about my Elk Custom Amp 30. It was really noisy and it made all sorts of strange sounds. I think I have eliminated most of the hiss. Only the tremolo hisses and there is a constant 60 Hz hum but I have a three more caps to check and one tube to replace.

I'm taking a digression from working on the hiss and hum to examining the power output section of the amp. The original design seems to call for 7189 tubes which, from what I can tell, are rugged military style EL84 tubes. Sometimes they are called EL84M or 6P14. The previous owner installed EL84 tubes (which are kind of old now) and he modified the circuit a bit. Two of the biggest changes were adding a 1.5k resistor to the screen grid lines and changing the original 220k resistors near the control grid to 100k resistors. I measured some Voltage values on the power supply, anodes, and screen grid. Here's what I got.

Rectified DC output : 441V
B1+ : 392V
B2+ : 304V
EL84 anodes : 439V
EL84 screen grids : 391V

Does that sound a little high?

I have a schematic that I am working with and I am modifying it as I make my way through the amp. The schematic was originally found on the web and it's old school style. It's dense and crowded. I'm attaching my modified version along with a change guide and the original. The Voltages are marked in the schematic. There are other markings to show the amp as it is now (i.e. "as found").

Can someone comment on these Voltages and tell me if I am burning up my power tubes?

Thanks!

-=- Boris
elk es30 schem with notes 2012 12 19.pdf

[soundguruman]

What you wanna do is adjust the C- voltage more negative, until the amp starts to distort before clipping.
Then, back it off until the amp stays clean all the way up until clipping...you should see the distortion just after the amp clips, not before.
If you adjust it right about there, crossover distortion just after the amp starts to clip, your output tubes should be OK.
If the C- is not negative enough to cause distortion before clipping, you will have to modify it a bit...get the control grids more negative.
If the C- is too negative, and you can't get rid of the distortion before clipping by adjusting the pot, then you will have to modify it a bit...get the grids more positive...
See, with the screens at 391, you should have no difficulty. There is plenty current limiting with 1.5K you already installed.

[poweredbypoptarts]

My C- Voltage is -21.6V.

The adjustment process you described seems to require an oscilloscope, which I don't have right now. Is there any way I could do this with a Voltage meter? If I did have an oscilloscope it seems like I would have to feed a sine wave signal and watch it on the output. Would I have to set my volume knob all the way up?

So my anode Voltage of 439V is OK? I thought it had to be like 250V for an EL84.

-=- Boris

[soundguruman]

That is for class A, you are running AB. And it depends on the tube also. NOS Phillips is like, 550V 46 ma plate limiting value per tube...or like, 65ma at cathode...
But for 439V your C- is not enough, and I think your output tubes may be turning red.
(are your output tubes turning red?) I would not doubt it...if they are glowing red or slightly red, you do not have enough C-
Turn off the lights, and look carefully in a dark room, for signs that the plates are looking red or slightly orange. This means that there is not enough C-
There should be enough C- to prevent the plates from overheating, turning red / orange with no sound, and then check with sound...
And then if the amp sounds muddy, there is too much C-

But, as you said, you can hook a current meter in series with the anode, or with the CT of the output transformer, or at the cathode. How much current at idle? how much current at full clipping? You will at least know if the cathode exceeds 65 ma at full audio clipping (?) That's a starter.

I am telling you this, assuming you are not going to electrocute yourself. Please don't do that.

And always have a speaker, or load resistor hooked up. Don't ever run the amp without a speaker output load connected.

And that's it. Too little bias C- and the tubes turn red...too hot...
Too much C- and the amp sounds all muddy and distorted. The tubes are too cold.
You are trying to find a place between too hot and too cold. That's what bias adjusting is...
If you have it in between, the tubes will get hot, they are supposed to get hot.
But not too hot. You see now?
You can also feel when the tubes are staying too cold. Too much C- will make them stay cold...and the amp will sound muddy.

[Enzo]

EL84 has a book rating of 300v, the 7189 has a book rating of 400v. But we do go over those book limits all the time. What kills tubes is dissipation. The current through the tube times the voltage across it will give you the idle dissipation in watts. Book values are designed to give long life and trouble free operation. Like dad's old table radio or TV set. Dad would not be happy if his tubes wore out every few months. But guitar amps expect to wear tubes out. It's like a car, your mom's car should last a long time, that's the book rating. If you took the same car and drag raced it every week, it would wear out a lot sooner, that's a guitar amp.

Keep in mind this is a guitar amp, not a hifi. Guitar amps ovedrdrive, so if you turn it all the way up, you are very likely to get a distorted output. SOnically we like that, but you'd never see the crossover distortion in that. I never use the crossover distortion for biasing a guitar amp, precisely because it is not a hifi amp. You won;t hear a little crossover notch anyway.

Most guys bias their amps by reading the tube current and setting the bias for 60-70% of rated tube power. For an EL84, maybe 10 watts is a target.

A common way to read current is to put a 1 ohm resistor in the cathode lead of the power tubes. Current through that resistor in milliamps will be equal, by Ohm's Law, to the voltage across it in millivolts.

ANother way is to measure the resistance of the output transformer primary from center tap to each plate lead. Note the resistance. Now power up and measure voltage from that center tap to each plate. Use Ohm's Law to calculate from that resistance and voltage the amount of current that is flowing. Then use that current and the plate voltage to get power in watts. Laboratory accurate? No, but close enough for our purposes.

There is also the shunt method, but I'd rather you stay away from that.

[kleuck]

Yes you are, 439 volts is just crazy.Dissipation is not the only concern. Drop the voltage a little with some zeners, and use 6P14P-EV only.

[poweredbypoptarts]

Yes, I think I am burning up my tubes.

I haven't taken any measurements yet but I flipped my amp upside down so I can look at the tubes. Right side up exposes the board so I can work on it. Here's a picture of what the rectifier (far left) and power tubes (middle and right) look like.



The glow looks purple in the pictures but that's the camera not pickup the color correctly. I think this is some sort of infrared emission picked up the CCD element. The glow is actually bright white.

I found some 6P14P-EV tubes on eBay. This is a bulk lot but the unit price per tube is the lowest.

6P14P EV 6BQ5 EL84M Tubes 8pcs | eBay

Measurements are forthcoming!

-=- Boris

[Enzo]

Yes, cameras usually see more than we do. Ever watch a cooking show when they flambe something? The flames from the pan look really orange and unnatural. That is because the camera reacts to infrared and puts it on the screen in visual range. I imagine yours is doing something similar.