85W out of two EL34s??

[waspclothes]

Hey everyone,
I recently saw a post where someone was claiming to get 85W out of a fixed bias class AB amp with a pair of EL34's. They even provided scope shots showing a nice sine wave at 26VRMS supposedly going into an 8 ohm dummy load, which would give 85W.

I was always under the impression that EL34's had 25W max dissipation, so an amp with two of these tubes couldn't put out more than 50W safely. I understand that some amps exceed the plate voltages and stuff to get more out of these tubes, but is 85W as far as some old designs were pushing these tubes?

[Justin Thomas]

Waspclothes,

Don't confuse max signal power out with maximum plate dissipation. Most thorough tube data sheets will give a spec for both. The scope won't lie about the signal output, nor will the scope tell you how hot those tubes are idling. That plate dissipation can be exceeded for periods of time, as long as the AVERAGE is not over 25. So, they could put out 80W of signal for 5 seconds, but then go back to idling at 40W for 10 seconds. Or something like that...

There's more to it than just the plate dissipation - signal in, various voltages, bias points, TYPE of signal, etc.

Justin

[J M Fahey]

There is an official Philips 100W schematic using 2 x EL34 , class B, 800V on plates, 400V on screens, 11k plate to plate impedance, which used to be VERY popular here in Argentina in the 60s/70s as distributed PA system, think Church , train/bus station or WalMart PA system, driving a ton of roof mounted 6" speakers.
Suggested transformer, for which they offered full winding data, had only a 500 ohms secondary, which clearly showed intended use.
Fapesa Amplificador 100W en kit Kit Philips Argentina,

[waspclothes]

Waspclothes,

Don't confuse max signal power out with maximum plate dissipation. Most thorough tube data sheets will give a spec for both. The scope won't lie about the signal output, nor will the scope tell you how hot those tubes are idling. That plate dissipation can be exceeded for periods of time, as long as the AVERAGE is not over 25. So, they could put out 80W of signal for 5 seconds, but then go back to idling at 40W for 10 seconds. Or something like that...

There's more to it than just the plate dissipation - signal in, various voltages, bias points, TYPE of signal, etc.

Justin

Thanks for taking the time to reply Justin. Maybe I'm wrong, but I'm under the impression that the maximum plate dissipation can be exceeded, but only for part of the wave cycle.. so fractions of a second, not continous for 5 seconds.

[Gingertube]

EL34 Application Data:
TDSL Tube data [EL34]
Up to 100W as Juan says.
The higher power options require a more complicated power supply to run high B+ for the anodes and a lower supply for the screens.

My favourite EL34 Amp runs B+ of 380V (anodes and screens), cathode biased into Raa 3K4 Output tranny for 32 Watts - lovely effortless loping pace and more power than I ever need.

I don't get the obsession with squeezing every last watt out of a pair of tubes. The EL34s in my "little" 32W Amp will last for a verry long time and I can try just about any EL34 or 6CA7 variant in the amp without worrying.

Cheers,
Ian

[mozz]

Traynor yba-1a. Just picked one up and running 580v on the plates maybe 30 mills plate current. I can get about 85 watts or so clean and i think it was 122watts major clipping. I have now added screen resistors and a choke to drop the screens down a bit. That is with EH el34's. With the original GE's i think it was 92 watts clean. They will do it no doubt. I have about 1 hour play time and a master volume now so i can't really say how long they would last pushing them.
http://www.aguitarforum.com/threads/nutad.90883/

[Justin Thomas]

I don't get the obsession with squeezing every last watt out of a pair of tubes...

I think that's a holdover from the "Golden Age," when the entire world was centered around maximum power with minimum distortion. Now I wonder if we care about either...
I wonder if part of this trend of squeezing watts is a result of the confusion between max plate dissipation and max signal out - that is, we have to bias our tubes so hot to get so much power, when it's as you said - these types if designs require a more complex power supply, one that many of us are neither inclined nor informed enough to worry about.

I remember looking at the designs in some of the old tube manufacturer manuals... we spend years fine-tuning a single gain stage, and slap in a shoddy power supply that's the same as in 1958, while these guys were using complex power supplies with very simple preamps.

I wonder - could someone scope an old Deuce, Music Man, one of the high-power amps known for clean that uses that funky way of squeezing as much emission from a set of tubes as possible and is biased at 25% dissipation, and measure the full power out? Maybe a Triple Rec or 5150, too... (Clipping be damned...) You know, one of the amps where we always see the owner saying, "it sounds great, but I wanna bias it hotter!"

Maybe this is like the great "it's current, not voltage, that kills" fight... while that may be true if there absolutely no other variables involved (IOW, a non-existent theoretical world), the scene quickly gets much more complicated...

Justin

[Enzo]

Those cathode drive power amps are running close to class B, that is why they are biased "cold."


How does a tube fail? It melts. Think about what would make a tube melt its screen into slag.

[Justin Thomas]

So, even though a Music-Man 130 is biased so cold, it still puts out it's rated power easily (if all else is running as designed), correct? SVT, same thing - they get measured putting out 280W, no problem, when the max dissipation of 6x6550 would be 252W?

I don't understand HOW they do it, exactly, by I know that they do in fact accomplish the feat... Enzo, can I steal your line? "You have to look at the whole circuit, not just a single aspect or part." -Enzo-

Justin

[Enzo]

Steal my lines all you want, and if you find a better one, let me know, and I'll start using it instead.

[The Dude]

What, Justin- are you trying to break the record for the world's longest sig?

[Justin Thomas]

I'm still under the limit for 500 characters... No I didn't count them!

Justin

[waspclothes]

Thanks for all the info guys,

I guess this brings my next question: How would you bias an amp like this? Does it make sense to set the Idle Dissipation to 70% still?

[g1]

What amp, and how long do you want the tubes to last?

[DRH1958]

Waspclothes,

Don't confuse max signal power out with maximum plate dissipation.
There's more to it than just the plate dissipation - signal in, various voltages, bias points, TYPE of signal, etc.

Justin

I know I'm a little late to the party on this thread but this might help.

Here is a direct quote from page 96 of RC-30 RCA tube manual. It's in the section "Interpretation of Tube Data" "Plate Dissipation is the power dissipated in the form of heat as the result of electron bombardment. It is the difference between power supplied to the plate of the tube and the power supplied by the tube to the load." So power provided to the load plus plate dissipation equals total power to the tube. From this, I gather that it's basically the left over wasted power in the form of heat, so the power to the load can be much higher than the "plate dissipation" in the form of heat. This directly backs up Justin's point and here is the exact explanation from the guys who "wrote the book" on tubes.