Reissue AC15 power tube consumption

[Wes]

I keep getting these amps in with tubes just eaten alive.

Every one of them has, like, 330-340V (K to A- not ground to A) on the plates with the screens not far behind. They have 120R
for a cathode resistor and by my calculations these jokers are pulling about 15 watts each.

Am I calculating that right? This tube wear agrees. If that is the case- why are they making these amps like this?

[Chuck H]

Well... EL84's are funny tubes in some ways. Mechanically marginal for combos in general and people like to run them hot. Also, the screens dissipate a higher ratio WRT the plates IME. This skews the perceived plate dissipation. Also, because of their lower plate dissipation of 11W max it's important to factor the actual cathode to plate voltage. For example, you figure that a tube is dissipating 15W at 335V uncompensated. Now if there's 10V on the cathode that's a working voltage of 325V. That same tube now figures at 14.6W. Not a big difference, but something to consider since EL84 designs often push the limits. Still, at over 300Vp and under 350Vp the max "whole tube" dissipation I've had acceptable results with is 14W per tube. And that's about max plate dissipation. When cathode biased the amp will still edge into AB operation when clipping due to bias shift.

I might install a 150R and see how it goes.

Something that confuses me is that I've built a half dozen of one design that has 355Vp on a pair of EL84's and uses a 130R shared resistor. These amps bias right at 13W. So how you're getting 15W with a 120R with 20 less volts I don't get. You mention your calculations, the plate voltage and the cathode resistor value but not the cathode voltage. Of course, different tubes bias up differently too. So a brand difference here might be where my confusion is.

[EFK]

I think the simple answer to your question as to "why" is simply because, that's the way Vox has always done it. It's what people expect. And EL84 are cheap.

[Mick Bailey]

Quite a few amps run EL84s well over the limit - the Marshall 20/20, most Traynors, Dr. Z Maz 18, Matchless and lots more. The Sovtek EL84M will stand up to this punishment a lot better. If you do a search everyone will tell you they're too 'Hi-Fi' for guitar amps. I've found them to be pretty good with no complaints at all. Matchless fitted them as stock.

You have to know the voltage across both the cathode resistor and screen resistor to get to the true plate dissipation. Calculate the total draw then subtract the screen current.

[okcrum]

12 watts plate and 2 watts screen dissipation is what RC30 (RCA) said in the day as limits. Screen overdissipation is a tube killer, so definitely check that. Changing the screen resistor will change the nature of the power amp breakup, but raising it from 100R to 1k is one thing to try if the screens are overloaded.

[Enzo]

The RCA book talks about DESIGN limits. The book gives specs for designers to use. There were no hot overdriven guitar amps then. Designs were for TVs, radios, hifis, etc. Those design limits were for circuits to run clean and reliable for long term. No one wants to put new tubes in a table radio twice a year. But we expect to use two sets of EL84s in a year. We exceed the book limits because we are not running polite little mom and dad radios.

[g1]

Near as I can tell, the schematic shows them at around 17W so you're a little on the cool side .

[okcrum]

The RCA book talks about DESIGN limits. The book gives specs for designers to use. There were no hot overdriven guitar amps then. Designs were for TVs, radios, hifis, etc. Those design limits were for circuits to run clean and reliable for long term. No one wants to put new tubes in a table radio twice a year. But we expect to use two sets of EL84s in a year. We exceed the book limits because we are not running polite little mom and dad radios.

What makes you think they are anything other than a data point for design?

Or you can take modern day a la cart specs, per tube, per manufacturer, or trace the curves yourself.

I'm not sure what your beef is here, but there were sure as heck overdriven guitar amps in the RC30 days, and there were even 6CA7s that were pentodes.

[Enzo]

What makes you think they are anything other than a data point for design?

Huh?? That is exactly what I DID say.


My beef? Got no beef. But a lot of people worry that some amp runs tubes at way higher voltages or dissipations than the RCA boom recommends as though the RCA numbers are some kind of absolute maximum. The numbers in the RCA books is from an older era, and the last thing on RCA's mind when those specs were made was a screaming Marshall stack. Did they make Marshalls in the 1960s? Yes, of course, Jimi Hendrix comes to mind. But that is not what RCA cared about. They wanted GE or someone to design a piece of consumer electronics that would sell hundreds of thousands of units, rather than a small number if guitar amps. That would mean a customer for hundreds of thousands of RCA tubes. We may love to overdrive a 5E3, but when Leo made those amps, no one was running them that way. SO when people see 400v on 6V6 plates where RCA says 300v max, they shouldn't freak out.

[Wes]

The voltages I gave were from cathode to plate, not ground to plate. The cathode resistor is 150 and had about 11 volts across it.

So, yes it is high, but as G-one pointed out- it doesn't look as bad as what the schematic calls for. Heh.

Interesting.

[Chuck H]

The voltages I gave were from cathode to plate, not ground to plate. The cathode resistor is 150 and had about 11 volts across it.

So, yes it is high, but as G-one pointed out- it doesn't look as bad as what the schematic calls for. Heh.

Interesting.

To quote myself on the matter, re el84's: Smoke 'em if you got 'em.

[okcrum]

Huh?? That is exactly what I DID say.


My beef? Got no beef. But a lot of people worry that some amp runs tubes at way higher voltages or dissipations than the RCA boom recommends as though the RCA numbers are some kind of absolute maximum. The numbers in the RCA books is from an older era, and the last thing on RCA's mind when those specs were made was a screaming Marshall stack. Did they make Marshalls in the 1960s? Yes, of course, Jimi Hendrix comes to mind. But that is not what RCA cared about. They wanted GE or someone to design a piece of consumer electronics that would sell hundreds of thousands of units, rather than a small number if guitar amps. That would mean a customer for hundreds of thousands of RCA tubes. We may love to overdrive a 5E3, but when Leo made those amps, no one was running them that way. SO when people see 400v on 6V6 plates where RCA says 300v max, they shouldn't freak out.

So, you were looking to alleviate my worries?

That's so very kind of you. Do you have friends who worked at Sarnoff, too?

[Chuck H]

So, you were looking to alleviate my worries?

That's so very kind of you. Do you have friends who worked at Sarnoff, too?

Is that supposed to be funny or a sarcastic slight? Just asking because I don't know. Intention is everything on this one regarding where you stand right now.

[okcrum]

I'm wondering where all the chest-thumping is coming from. Dude seems intent on schooling me about a handbook I've been using for small stuff for decades. I actually have friends from Sarnoff.

[Chuck H]

I'm wondering where all the chest-thumping is coming from. Dude seems intent on schooling me about a handbook I've been using for small stuff for decades. I actually have friends from Sarnoff.

The chest thumping comes from me being an animal. But I'm a smart one. Just part of the human condition. I accept it readily.

You're the one that started swinging. Enzo pointed out that the parameters in the RCA book don't always apply to the guitar amp circuits we're familiar with. And this is true. And he also pointed out that there were no overdriven guitar amps when this book was written. This is also true. RC 30 is an update. But just to further the position, though I haven't read it, I'll wager that RC 30 doesn't address many specific circumstances found in overdriven guitar amps either. There's nothing wrong with your observations and suggestions in post #5 EXCEPT that it doesn't apply to the classic VOX design. If the classic design over dissipates the screens, and that's the sound we want, then it shall be. So sayeth the tone gods. You can't expect the tonal mojo of these inherently flawed circuits to remain if we "fix" them. Personally I would take your advice and see how it sounds. Solid cautionary info. There was nothing wrong with what you said and Enzo wasn't schooling you. He was pointing out another angle to be considered. In other words, he wasn't saying that you're wrong, he was saying that there's more than one right answer.

Please try to get your head around what's actually going on in this situation. You really don't want to get it wrong.