Tube amp heat questions...


Hi all, I recently joined the ranks of tube amp owners by adding an ARC VSi55 to my setup. Love it, and it doesn't seem to put out a ton of heat, but it's getting cooler here in Phoenix so I may feel differently next July.

I did search thoroughly through past threads to see if these questions were asked, and found some good threads about summer amps vs. winter amps, but not my specific questions:

- Is there a direct correlation between tube amp wattage per channel and heat generated, such that for example any 100wpc tube amp will always put out more heat than a 25wpc tube amp?

- If that is the case, it is because higher WPC = either more tubes and/or larger tubes?

- If that is not the case, is heat a function of tube type, such that certain 25wpc tube amps can generate more ambient heat than a 75wpc amp just because of the type of tubes it uses?

I didn't take these questions into account when purchasing the VSi55, but we may put together a 2nd system for another room and I may take this into account for that purchase. And as others have done, having a nice Class-D amp for the June-Sept period is also a possibility. Thanks in advance!
128x128bcgator
Swampwalker, actually I agree with you on those 6C33s. They run the hottest filaments of any common power tube these days!
Bombaywalla- No offense intended to my vacuum-challenged friends here on the 'Gon. Oooops, I did it again!
I was TRYING to be nice, but I'm not very good at it! LOL. I almost said "friends don't let friends listen to solid state".

BTW, I'm sure you're right: I did hear ONE ss rig (Class D no less) in NY last April that I liked. Oh, no, I'm incorrigible. Sorry.

Good luck, have fun, and stay away from those nasty transistors ;-)
Swampwalker
Hey, hey, hey - be nice, swampwalker!! :-D LOL!
those little silicon gems are nice if you know where to buy the right kind... ;-)
Like Sfar, I can't add to the technical side of this discussion but I will tell you that, IME, in my system, there are 2 extremes in tube amp heat production. Coolest running is the Berning ZH-270. Hottest are Joule OTLs using 8, 10, 12, 16 of those big ole beautiful Russian 6CC3Bs. With some trepidation to be contradicting Atmasphere, in that case, the output transformers, being absent, have nothing to do w it. Those tubes are just little furnaces; even the Almarro 318s that use only a pair can heat a room!

I suspect that those examples are outliers with good technical reasons for being so. The difference is night and day (or winter and summer). In general, I'd stick w the advice you've gotten above. But, I will say that if I lived in the south, I'd probably never have sold my ZH-270. Out of production, so hard to find, expensive ($4-5K), and not compatible w all speakers, but probably the best tube amp I can think of for your situation. A less expensive but also hard to find option might be a high eff speaker w Berning's Microzotl, which uses a pair of 6SN7s as output tubes!! In a small room at my house at the beach, that little amp sounded pretty darn impressive driving a pair of 96 dB back-loaded horns. Now my taste runs to singer-songwriter/fold/acoustic type music at moderate SPLs, but in the right circumstances, it might be option. They show up here from time to time at just under $1K; since they were primarily marketed as a headphone amp, you might have better luck looking for one on head-fi.org. Good luck, have fun, and stay away from those nasty transistors ;-)
I can't add to the excellent technical information you've received but just as a point of interest I recall using an infrared thermometer like the one Viridian suggested to check the temperature on one of my tube amps.

In an auto-bias AES/Cary AE-25 amp running at idle in triode with a stated power of 15 watts from KT-88 tubes the temperature of the output tubes was somewhere close to 300 degrees Fahrenheit.
Bcgator, air conditioning is a brute-force method of dealing with heat. In the old days people simply had vents in the ceiling. A simple solution, inexpensive and one that works really well. Even in the 21st century, heat still rises :)

You will find that there is a loose correlation to the sound of the amplifier and the class of operation. I say loose as there are other variables that can have a big effect on the sound. Generally speaking though, amps that are class A (make the most heat) tend to sound better than amps that make the least heat (class D being an example).

It is not the filaments of the tube that make the heat in a tube amplifier. It is the class of operation and how efficient the amp is running (the two are not always the same). You can make the amp run more efficiently by using a loudspeaker that is higher impedance and higher efficiency. Almost any tube amp will run with less heat into 16 ohms for example, because the output transformer will operate more efficiently- more power makes it to the speaker and less is dissipated as heat.

Conversely, any amplifier (tube or transistor) will make more heat if asked to drive a four ohm load (assuming that loudspeaker efficiency is otherwise the same)- and it will make more distortion which results in less detail and harsher sound.

So your speaker can play a big role in this. If its easy to drive, generally speaking you will get less heat from the amp and better performance too. Something to think about...
Wow, fantastic information, thanks everyone. I had a suspicion that I was probably oversimplifying, so I'm glad I asked. There's good info in every answer, thanks to all.

Newbee, your point about buying only the power needed is very pertinent. I'm driving a set of B&W PM1, with 84dB sensitivity and they dip to 5.1 ohms. I went to the dealer to look at a Rogue Sphinx, certain that I needed every bit of their 100wpc (if not more) to push the B&W (my Marantz integrated with 70wpc did not match well at all and ran out of gas). The dealer showed me the Sphinx, then the Cronus Magnum with 90wpc, but insisted I also take a listen to the pre-owned VSi55 he had just taken in, with brand new tubes. I almost made the mistake of dismissing the ARC thinking that 50wpc would be a backwards step and waste of time and money, but I'm glad I listened to him. I'm too inexperienced to understand the technicals of why 50 watts of tube power opens up the B&W and pushes them to insane volume levels while the music remains perfectly clear and listenable, while 70wpc of solid state power turns to aural chaos at anything above 1 o'clock on the dial - I'm sure the Marantz is a superb integrated, and it was just a bad match-up. But the lesson I learned is that there's more to the story than just absolute WPC.

I like the thermometer gun idea, Viridian.

Jjrenman, your point about Class-D is duly noted - when summer heat rolls around, having one for back-up duty may save me electric bill $$ and make my den bearable if the ARC just gets too hot.

Al, thanks for the layman-friendly explanation - I needed it!

Eric
First, congratulations on your purchase, and welcome to the ranks of happy tube amp owners :-)

As the others have indicated there are many design-related variables that are involved, so the short answer to your well put questions is "no."

A longer answer: The amount of heat generated by the amplifier itself will be proportional, at any instant of time, to the amount of AC power going into it minus the amount of power being delivered to the speakers.

However, since with nearly all speakers that are intended for use in the home the great majority of the power going into them will be converted by the speakers into heat, with only a small fraction of that power being converted into acoustic energy, it is reasonable to approximate the total amount of heat generated by the amplifier + speakers as simply being proportional to the AC power drawn by the amplifier.

The relevant specs on your VSi55, for example, are as follows:

POWER REQUIREMENTS: 100-125VAC 60Hz (200-250VAC 50Hz) 320 watts at rated output, 500 watts maximum, 205 watts at "idle". 0.5 watts power off.

So you can project the amount of heat that would be generated by the amplifier candidates you consider for the second system by comparing their specified AC power consumption with the 205 to 320 watt range your present amplifier has under normal operating conditions.

If some of the candidates don't publish AC power consumption specs, ask the manufacturer if they can supply those numbers.

Finally, keep in mind that under typical circumstances an amplifier will be called upon to approach its maximum output power capability just on infrequent dynamic peaks. So most of the time you are probably operating the VSi55 such that its AC power draw is much closer to the 205 watt number than to the 320 watt number.

Enjoy! Regards,

-- Al
FWIU, the following points all have an impact on either the WPC and/or the heat generated.

~The class of operation (single ended, push pull).

~The type and configuration of the output tubes (triode, tetrode, ultra linear or pentode).

~The number of (mainly) output tubes.

~The amount of bias and plate voltage on the output tubes.

~The type of bias on the output tubes (fixed or cathode).

The bottom line is that tube amps of comparable or lesser wattage have far more excess heat than all solid state amps with the possible exception of a pure class A solid state.

Even less heat is generated by SS class D amps.
Post removed 
Probably has more to do with how the amp is designed in the first place and how far the design pushes its tubes to develop its rated power. For example you can get 80 watts out of four 6550 tubes but they are close to max output. The same tubes in a design that produces 40 watts will run its tubes cooler and they will last longer. But it is fair to say, I think, that the more power your amp produces the more heat it will produce, either in how many tubes you need or how hard you have to push the tubes. Moral of this story is if you don't absolutely need all that power why buy it in the first place.
I like tube amps so much, I am willing, and do use them in the summer, as well. I just don't sit that close to the amps and am lucky enough to have central AC which I imagine you have as well. I don't want to sound insensitive and can imagine being in a position where the heat may be a nuisance.
As for your question I don't know whether more watts produce more heat necessarily but if I was betting on it I would guess yes- it seems logical, that is for most cases. My 78 watt per channel monoblocs certainly make more heat from their 4 output tubes than my old 40 watt integrated which had two. In addition the monoblocs use somewhat larger fat bottle 6CA7s vs the E34Ls (A JJ invention).