18w EL84 Cathode Bias question

[Greg_L]

Hello guys...and maybe gals? Any ladies here?

I'm working on a Mojo kit 18w TMB "Marshall".

The schematic calls for a 130 ohm cathode resistor bypassed with a 25uf/25v cap.
With this arrangement my plate voltage is 355vdc.
Plate-to-cathode is 343.5 vdc.
Voltage drop across cathode resistor is 11.5vdc
About 40ma net current.
All of this conspires to have the tubes running about 13.75 watts per tube. Too high IMO. Sounds good, no red plating, but that's well over 100% for EL84s.

I'd like this amp to be trouble free and have respectable tube life so I put a 180ohm resistor for the cathode bias and that dropped my bias right down to about 33ma; 12 watts on the money. But my plate voltage is now up around 360vdc.

I know it's fairly common for EL84 amps to push the plate voltage over the EL84s specified limits, but I'm wondering if you think this plate voltage is okay as long as my cathode-bias current is safe for the tubes.

[The Dude]

What is your filament voltage? If that's in reasonable range, I'd up the cathode resistor value and call it good. If not, you may need to do something on the power supply side.

[pdf64]

I’d put it back onto Mojo; why can’t they specify their PTs better?

[Greg_L]

What is your filament voltage? If that's in reasonable range, I'd up the cathode resistor value and call it good. If not, you may need to do something on the power supply side.

6.7vac. Seems reasonable to me.

[mozz]

Most of the voltages are at 345v on the 18 watt website. That's how you run them, hot. If you have a scope check for crossover distortion, kind of hard to get rid of on el84's.

[Greg_L]

Most of the voltages are at 345v on the 18 watt website. That's how you run them, hot. If you have a scope check for crossover distortion, kind of hard to get rid of on el84's.

Yeah I do have a scope. With the smaller cathode resistor and therefore way hotter bias (127% dissipation!) there was crossover distortion, but I didn't worry about it.


I'm under the impression that a tube can withstand high plate voltage better than too high bias current. Is this true? If so, I'll live with the slightly high plate voltage as long as the tube is happy with the bias current.

[g1]

I'm under the impression that a tube can withstand high plate voltage better than too high bias current. Is this true? If so, I'll live with the slightly high plate voltage as long as the tube is happy with the bias current.

Traynor YGM3 ran them around 400V. They reissued them around 10yrs. ago and dropped that all the way down to around 370V.

[Greg_L]

Traynor YGM3 ran them around 400V. They reissued them around 10yrs. ago and dropped that all the way down to around 370V.

Haha point taken.

[Dave H]

Yeah I do have a scope. With the smaller cathode resistor and therefore way hotter bias (127% dissipation!) there was crossover distortion, but I didn't worry about it.

I'd expect to see more crossover distortion with the bigger cathode resistor. To reduce crossover distortion I think it's better to lower the B+ voltage rather than increase the cathode resistor value. The traditional way to do this was to put a resistor in series with each plate of the rectifier tube. Try a 150R cathode resistor with the B+ reduced to 345V by suitable choice of plate series resistors. EL84 Pd should then be down to about 12W.

[Mick Bailey]

The Marshall 2020 EL84 amp has a b+ of 400v and tube dissipation is 14w. Tube life isn't a problem, just like with the Traynors and Mesas where you expect the high voltage along with high-ish dissipation to kill the tubes but in reality they last quite well. I'm in agreement with Dave H that increasing the cathode resistor can lead to more crossover distortion, but if the amp sounds fine then I'd leave it at that. Tubes can stand a lot more voltage than their spec sheets suggests. I have a 807 (similar spec to a 6L6 with lower screen voltage) amp that runs happily with 1Kv plate voltage. Originally it was 1.2Kv and it appeared to be OK even at that.

[Helmholtz]

I'd expect to see more crossover distortion with the bigger cathode resistor.

I don't have much experience with crossover distortion measurement in cathode biased amps and I wonder where it would be seen.

As even with a cathode resistor of 180R, idle currents are higher than with typical fixed bias circuits, I guess crossover distortion will not start before the tubes leave the class A region and cathode voltage increases at some medium output. Right?

[pdf64]

6.7vac. Seems reasonable to me.

It's over 6% too high; what's your mains voltage, ie at the wall outlet?
If your HT was 6% lower, maybe the EL84 would be at a much happier place with the stock arrangement?

I think it's pretty reprehensible that Mojo and other kit suppliers are instructing builders to fuse the neutral https://923962.app.netsuite.com/core/media/media.nl

I don't have much experience with crossover distortion measurement in cathode biased amps and I wonder where it would be seen.

As even with a cathode resistor of 180R, idle currents are higher than with typical fixed bias circuits, I guess crossover distortion will not start before the tubes leave the class A region and cathode voltage increases at some medium output. Right?

Correct IME, it seems to mainly affect high signal levels; as the bias voltage at the cathode increases (in the class B area of operation I suppose) the fidelity of the output waveform tends to become increasingly affected.
It may be due to the ratio of Ia peak to Ia idle is somewhat lower with typical cathode bias than fixed bias designs, the crossover kink occurs at higher levels, rather than closer to the zero crossing type that's tends to be seen with fixed bias.
To see for yourself, consider increasing the value of the cathode resistor in your AC30, eg to 82 ohms. But maybe that would be a sacrilege

[Helmholtz]

Correct IME, it seems to mainly affect high signal levels; as the bias voltage at the cathode increases (in the class B area of operation I suppose) the fidelity of the output waveform tends to become increasingly affected.

Yeah that's what I thought. At medium to high output (depending on cathode R), where the cathode voltage increases and the negative going grid signal drives the tubes into cutoff.

[Greg_L]

I'd expect to see more crossover distortion with the bigger cathode resistor. To reduce crossover distortion I think it's better to lower the B+ voltage rather than increase the cathode resistor value. The traditional way to do this was to put a resistor in series with each plate of the rectifier tube. Try a 150R cathode resistor with the B+ reduced to 345V by suitable choice of plate series resistors. EL84 Pd should then be down to about 12W.

Interesting. Few questions...
Wouldn't that lower the voltages everywhere?
What value/wattage resistors should I try if I wanna go this route?
And why would I put one on each plate and not just one off the rectifier output?

[Greg_L]

It's over 6% too high; what's your mains voltage, ie at the wall outlet?
If your HT was 6% lower, maybe the EL84 would be at a much happier place with the stock arrangement?
:

Maybe. My wall voltage is stout. 125 VAC some days. I'll try the amp with a variac and drop the voltage a little and see what happens.