18w EL84 Cathode Bias question

[pdf64]

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?

Reducing the HT voltage will indeed reduce all downstream voltages, it’s kinda inevitable.
At a guess, 5W should be ok. Given the application, carbon types would be unsuitable. Get something with a flame retardant coating.
For some reason, a rectifier valve’s limiting current is often specified per plate. Hence a resistor per plate looks to be a good way to ensure that.

[Greg_L]

Reducing the HT voltage will indeed reduce all downstream voltages, it’s kinda inevitable.
At a guess, 5W should be ok. Given the application, carbon types would be unsuitable. Get something with a flame retardant coating.
For some reason, a rectifier valve’s limiting current is often specified per plate. Hence a resistor per plate looks to be a good way to ensure that.

Awesome, thanks. This is certainly something that I'm willing to try and it's easy and reversible.

I wouldn't try a carbon resistor for something like this. I was thinking metal oxide or ceramic. If it needs to be 5w probably a ceramic. I can get those pretty easy locally.

Say I want to knock everything down 5-10% after the rectifier. I think that would be probably be enough to tame my voltages without dipping too low. What would be a good resistor value to start with at the rectifier plates? Or how would I calculate that?

[Dave H]

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?

I used a single resistor (because I had a bridge rectifier). It was 220R 7W. I measured resistor dissipation as 5W at 15W output and 4W at no output but I don't think you'll need to go as high as 220R. My B+ was 375V without the resistor.

[Greg_L]

I used a single resistor (because I had a bridge rectifier). It was 220R 7W. I measured resistor dissipation as 5W at 15W output and 4W at no output but I don't think you'll need to go as high as 220R. My B+ was 375V without the resistor.

Excellent, thank you. That's good to know. I was just guessing at trying something around 150R but had nothing to base that off of.

[Dave H]

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

With a 150R cathode resistor it doesn't show up until it's clipping quite hard.







Edit: Real measurements I did today, not a sim

[Greg_L]

With a 150R cathode resistor it doesn't show up until it's clipping quite hard.







Edit: Real measurements I did today, not a sim

That's actually similar to my results with the super hot bias 130R bias resistor.

[Greg_L]

So I've been reading up on this stuff, and I've seen that, like some of you have said, these amps just run hot. Okay, I can live with that. It sounds very good to me and the tubes are not redplating.

I still want to toy with dropping voltages, but now I'm also worried about screen voltage. I've read that EL84s can handle hot bias and high plate voltages but high voltage on the screens can really hurt tube life. My screens are hot. 340VDC. Most of what I've seen recommends 300 max on the screens.

Just as a test I installed a 10k/5w ceramic resistor off the screen supply filter node and that dropped my screen voltage down to right at 300VDC. 297, actually. Plate voltage is still up in the 365 range. Tubes are running about 96% plate dissipation.

Is this a responsible solution to lowering the screens? Is lowering the screen voltage something that will help tube life?

[Helmholtz]

Is lowering the screen voltage something that will help tube life?

I definitely think so, as this will considerably lower screen dissipation at high output.

But, acc. to tube data, the total screen current for both tubes rises to 22mA at full output, which will cause a voltage drop of 220V and a screen voltage below 80V.
Such low screen voltage would reduce available output and change sound and amp dynamics.

So I would not go above 5k here.

[Greg_L]

I definitely think so, as this will considerably lower screen dissipation at high output.

But, acc. to tube data, the total screen current for both tubes rises to 22mA at full output, which will cause a voltage drop of 220V and a screen voltage below 80V.
Such low screen voltage would reduce available output and change sound and amp dynamics.

So I would not go above 5k here.

Wow ok thanks. I hadn't thought of that. So much to learn. Thanks for the info.

[Mick Bailey]

Incidentally, I worked on an amp with 10k ohm screen resistors fitted by a previous owner and the sound was extremely compressed with the volume at home listening levels. Along with the compression came an unpleasant fizzy background distortion. I arrived at 2k2 ohm as a good tradeoff between screen protection and dynamics.

[Greg_L]

Incidentally, I worked on an amp with 10k ohm screen resistors fitted by a previous owner and the sound was extremely compressed with the volume at home listening levels. Along with the compression came an unpleasant fizzy background distortion. I arrived at 2k2 ohm as a good tradeoff between screen protection and dynamics.

Yeah this amp will spend most of it's life at low-to-moderate volumes so killing the tone down there is something I need to avoid. Thanks for the heads up.

[pdf64]

...Just as a test I installed a 10k/5w ceramic resistor off the screen supply filter node and that dropped my screen voltage down to right at 300VDC...

Do you mean that you swapped the 2k7 HT dropper between the OT CT and the screen grid supply nodes to 10k? Or that a 10k resistor was inserted between the stock screen grid HT node and the 2x100ohm individual screen grid resistors?

[Greg_L]

a 10k resistor was inserted between the stock screen grid HT node and the 2x100ohm individual screen grid resistors?

That ^^^^^

[Dave H]

a 10k resistor was inserted between the stock screen grid HT node and the 2x100ohm individual screen grid resistors

They probably intended to insert 1k but inserted 10k by mistake.

[Leo_Gnardo]

I was just guessing at trying something around 150R but had nothing to base that off of.

By coincidence I was working on a similar amp for a customer. Not a Mojo kit, B+ settles around 330V. Output tubes were rattly well-used EH EL84's, and biased about 10% beyond their suggested 12W plate power limit. 160 ohm substituted for the 125 ohm Rk, new pair of JJ EL84, Pp now 11.8W. Amp sounds great, crossover starts just at clipping. So folks, don't be afraid to sub the cathode resistor, run those outputs a bit cooler and within their suggested limit. FWIW on occasion I've used 180 ohms in a similar situation. With your higher B+ I wouldn't be afraid to go there, maybe even 200.

Good discussion, good graphs, to all correspondents.