Power tubes and heat-build up

[Kit]

Im working on this DIY guitar amp, a 30W el34 thing.
The amp is finished and sounds good and im about to put it in an enclosure, when i came to think: i dont really know anything about tubes and heat-build up.

When i build SS-poweramps i can select an appropriate heatsink based on calculating with thermal R, but with tubes im kinda lost.

I had an idea about mounting the chassi at the bottom of a closed 2x12 speaker box about the size of a fender twin(maybe a little deeper). Tubes would be facing up.

But seeing as the box is sealed theres is nowhere for the heat to escape, and i wonder if this will shorten tube life?
Are there any calculations or rule of thumb for this kind of problem?

 

[amorris]

i wouldnt seal the tubes up in the enclosure. keep the speaker enclosure sealed but build a "box outside the box" like an old emt plate reverb or leslie cabinet that has a little cubbyhole type space for the amp to live in so that real world air can circulate.

 

[adamasd]

Airflow is good, the tubes need to get rid of that heat and not being able to get rid of it will shorten their life. If you want a closed back cab look at the Ampeg flip top amps or at an AC-30 for how to do that. The AC-30 is not closed back, but it could no problem.

adam

 

[Kit]

Airflow is good, the tubes need to get rid of that heat and not being able to get rid of it will shorten their life.

Ok, but by how much?
There must be a formula for this.

The only info i can find on the web (so far) is articles on forced cooling for
rf tubes in the kW range.

 

[adamasd]

RDH4 says:

As a general guide, the maximum temperature of the hottest part of the bulb under operating conditions in the equipment should net exceed by more then 20ºC, that temperature that would be attained if the valve were operated at its maximum ratings under conditions of free air circulation in an ambient temperature of 20ºC.

That is all I can find on the subject. There probably is not going to be any equation for this, since the equation would have to be a per tube basis. The main thing to remember is Convection is what is going to move the air for you, 50% of a tubes dissapation is by convection. So if that hot air has somewhere to rise out of the cab from, and there is somewhere for the cold air to be sucked in you will be fine.

adam

 

[adamasd]

I forgot to say, a pair of EL34s running in a closed back cab would pretty much turn that cab into an oven with the EL34s as the heating elements. Eventually the cab would reach the temperature the EL34s are running at since there is no place else for the heat to go. EL34s can put out alot of heat, it could be quite dangerous to run tubes like this. Speaker Cones are paper and it really does not take much heat to get paper to combust.

adam

 

[bcarso]

There will be some internal convection and then conduction through the walls to the outside world, but it is very poor for wood cabinets, still lousy for plastic, and still mediocre but much better for metal cabinets.

It is a slight exaggeration to say that there is no cooling whatsoever, but why take chances?

Rows of slots at the base of the box and at the top are way better than none at all, but may not be adequate. But since you can't tolerate this (usually) from acoustic considerations, as it will spoil your calculations for a sealed box and whoosh a lot, you will have to seal off the amp compartment anyway. At that point lots of ventilation of that chamber with suitable mesh etc. to prevent junk getting in will be appropriate.

Another approach is to have the box be open-back (with mesh etc. if you like) and if necessary boost the bass a bit---after all, it's a guitar amp/speaker, not a subwoofer. You won't need a whole bunch of boost at 80Hz even with a completely open back and a decent-sized cabinet.

 

[pstamler]

Not only will sealing the box shorten the life of the tubes; it will also shorten the lives of all the other electronic components, including resistors and, especially, electrolytic capacitors. And the speaker.

Peace,
Paul

 

[Kit]

---after all, it's a guitar amp/speaker, not a subwoofer. You won't need a whole bunch of boost at 80Hz even with a completely open back and a decent-sized cabinet

Yes, this is all true.
But what im looking for in a closed back guitar cab is a "tight" sound rather than lots of bass. I find open back cabs to be "flabby".
If that makes any sense. :?

 

[bcarso]

[quote author="Kit"]

---after all, it's a guitar amp/speaker, not a subwoofer. You won't need a whole bunch of boost at 80Hz even with a completely open back and a decent-sized cabinet

Yes, this is all true.
But what im looking for in a closed back guitar cab is a "tight" sound rather than lots of bass. I find open back cabs to be "flabby".
If that makes any sense. :?[/quote]

Get a stiffer suspension driver or drivers. There should be something out there suitable.

 

[Kit]

The AC-30 is not closed back, but it could no problem.

I´ve looked at some pictures of an AC30 and its chassis-mounting, it seems like a good idea.

 

[PRR]

> I´ve looked at some pictures of an AC30

Look at a LOT of tube guitar amps. You can't do any better than they did, but don't figure in ignorance and do worse.

No, you can't seal power tubes in amp-size boxes. This is a basic constraint on tube radios and gitar-amps. This is why the no-NFB pentode survived to the end of tube radios and many gitar-amps: you have to open the back to let the heat out, and then the un-damped bass-resonance is your friend.

Or, as in some Ampeg B-15s: mount the amp up top, optionally on a flip-board so it can be stored inside for transport.

Sealed-back speakers were most often used with separate head-amps: the amp was wide-open in back, but the speaker was closed.

 

[CJ]

You could box it with a digi themometer in there for a few hours.
Speaker might pump some air thru the cone, when it's moving, that is.
Big cab will dissiptae a lot, mucho square inches, hate to seal stuff that might need attention, like a tube swap at the show.
In general, if nobody else is doing it, it is usually a bad idea.

 

[AnalogPackrat]

[quote author="PRR"]> I´ve looked at some pictures of an AC30

Look at a LOT of tube guitar amps. You can't do any better than they did, but don't figure in ignorance and do worse.
<snip>
Or, as in some Ampeg B-15s: mount the amp up top, optionally on a flip-board so it can be stored inside for transport.

Sealed-back speakers were most often used with separate head-amps: the amp was wide-open in back, but the speaker was closed.[/quote]

Weren't AC30s notorious for overheating? Those puny vent slots seem like trouble to me. Any ventilation on the top surface of an amp is just asking for dust, dirt, beer, etc. to collect on the internals. Fender got this part right, IMHO. Marshall did OK, too. The Ampeg Portaflex things are pretty cool (pun intended), too. Older amps with the controls on top are also somewhat vulnerable to spills. Just remember, if there is a flat surface available someone will put a beer on it and that beer will eventually spill on said flat surface. You'd really like to be able to just wipe (or slurp) it up and keep playing like nothing happened.

A P

 

[Kit]

In general, if nobody else is doing it, it is usually a bad idea.

Yes, i agree it usually is.

Im not trying to do something special here, its just that i find the topic interesting.

Over the years there have been many combo open-back amps with varying mounting styles for the tubes/chassis. We can narrow it down to these i think:

1. Fender tweed. Controls facing up, tubes at the other end facing down, transformers mounted horisontally. No top vent.

2. Marshall "bluesbreaker" style. Same as tweed but tubes mounted alongside transformers horisontally. Big top vent.

3. Fender blackface. "Hanging garden" style. Forward facing controls.
Tubes and transformers are hanging at the bottom of the chassi. No top vent of course.

4. Vox ac30. Controls facing up "tweed style" but the chassi is L-shaped and the power tubes and transformer are mounted vertically inside the box so to speak. Medium top vent.

I think the reason that AC30`s blow up is that they biased quite close to class A, maybe to close. To me it seem to offer the best "cooling" in that the tubes are mounted vertically with a vent directly above them.
But maybe thats not important, maybe intake and circulation is the key...
The blackface fender has a VERY open back compared to the others but im thinking that the heat would build up under chassi as it has nowhere to go....but then again a typical BF fender would be biased colder than an AC30.

Lots of options here.

 

[CJ]

The AC 30 never got hot, it just caught fire. No biggy in a cramped theatre, eh?
Park that chrome stand under the drapes for maximum clear-out factor.

 

[NewYorkDave]

Traynor Guitar Mate. The bottom portion is sealed, the top portion (with the amp chassis) has an open back. It makes for a bulky enclosure but a nice-sounding combo.

I don't have any pics of my Traynor, but I found these on the web:

http://www.kilback.net/traynortweaks/traynorphotos/ygm1.jpg

http://www.kilback.net/traynortweaks/traynorphotos/ygm2.jpg

With back removed:
http://www.kilback.net/traynortweaks/traynorphotos/73ygm3-back.jpg

 

[adamasd]

Wasn't it the AC50 that always caught fire and the AC30 that had caps in the tremolo that blew up?

adam

 

[PRR]

It hardly matters if the tubes hang down, sideways, or catty-corner like V-8 pistons.

The Fenders, tweedy and silver/black-face, have wide-open rears. The most restrictive cases I can think of are 5-tube radios with perf-board backs. Nothing remotely like a sealed box.

I did "seal" a PC once. No exhaust fan, no vents. I did my math and decided that a 386-33 could dissipate through the skin of an IBM AT case. Including bottom-skin: it was raised off the table to get a breeze under it. It did run for a year without heat-stroke, and the reason I abandoned it was a cheezy CD-ROM drive that quit on summer days.

> im thinking that the heat would build up under chassi as it has nowhere to go...

"Heat rises" is over-rated. If it can't rise, it wanders around. If it wanders to another exit, it leaves. The minor baffling behind most Fenders is not going to "trap heat" any more than insulating your attic prevents your heating dollars from finding their way out your windows and walls.

Tubes go outside the box. Speakers can go inside or outside, but you need very different voicing approaches for a sealed speaker than an open speaker.

> Speaker Cones are paper and it really does not take much heat to get paper to combust.

Short-term, you will have BIG problems with everything else before the cones catch fire. A hot but happy tube runs at about half the temperature rise needed to ignite paper. Try it. Run a receiving tube at ratings in a "reasonably vented" enclosure, put paper on it.

In time, the paper will oxidize much faster than you would like, get brittle and fall apart. Wood cases can char. If you actually let a hot tube rest on wood for a year, it might oxidize very rapidly. In larger systems (high pressure steam), spontaineous combustion could happen, though wood researchers have not found a case where it did. (Clean dry laundry does self-combust fairly regularly, when heaped hot from the dryer in institutional carts and left overnight.)

> You won't need a whole bunch of boost at 80Hz even with a completely open back and a decent-sized cabinet.

Interestingly, many tube guitar amps roll-off a couple dB at 80Hz. Two Twelves in a large open baffle, resonating near 80Hz, with no or slight resonance damping, can be overpowering on the low string, muddying-up the bass-player's tonal area and detracting from the tenor/alto tonal area that modern guitarists want to dominate.

The Fender Twin is a fascinating acoustic beast. Leaving the back open almost doubles the transducer efficiency (I didn't say sensitivity). It forces a bidirectional pattern from bass to well up the fretboard: "Constant Directivity" long before that was a Patent. This bi-di pattern has a lot more "throw" than a box-speaker which will be omni for the first couple octaves. And if you use a speaker with a rising response (high efficiency) and tuned to ~80Hz with little amp damping, plus the 2KHz shelf-boost typical of the Fender tonestack, you get a nearly constant power response over the guitar range.

 

[Kit]

The Fender Twin is a fascinating acoustic beast. Leaving the back open almost doubles the transducer efficiency

Interesting. I´ve would have thought that a closed box would have been more efficient.
Any papers on this?