OT - installing fans in an amp head

[Paul Marossy]

I would have it so that the fan draws air thru the grille cloth (if applicable) or in such a way that it blows out the back of the amp. If you have a grille cloth on the front, it will act as a filter of sorts. I would stay away from the fine mesh - it will act as a dust collector and will inhibit the flow of air. Just my veiwpoint based on my observations. It will get dusty inside the amp a little faster than normal, but it's not going to be caked on like you think it might. And I live in the desert, it's very dusty here - haven't had that much of a problem with a buildup of dust...  

[Manolo Dudes]

I put a 220V fan in my Marshall combo but fed it with 110V and blowing the air out of the cabinet.

See details here. You may find interesting how I managed to fasten the fan to the rear grill using bicycle chain links.

In spanish but with lots of self descriptive photos.

[petemoore]

As much of a drag as cleaning a fan and amp surfaces is, it's probably a good idea...you can lose big % of cooling with a thin layer of anything on surfaces expected to transfer/dissipate thermal energy.
 Extensive testing results on diminished capacity of motorcycle cooling fins to do their job when covered in 'something' is probably applicable.

[Paul Marossy]

Hey, that bike chain link thing is very clever!

[ESPm2M]

Hey, that bike chain link thing is very clever!

I agree. What a great way to mount it - nice job!

Paul, and others, please read my post titled "... fan mod info continued ..." that is seven posts back up the thread.  I am curious what you think of the NTC thermistors and fans at that link. thanks

[R.G.]

Tubes have to dissipate heat from the glass envelope. That's the only way heat gets out. If no heat gets out, the temperature inside can literally go to infinity - or at least until the glass melts, the air comes in, and it ceases being a vacuum tube.

Glass is an almost-insulator for heat. What it does in a tube is receive infrared radiation from the plate, get hotter then re-radiate the heat. It is not a selective radiator, so it radiates heat to the outside and to the inside. So some of the heat the plate is radiating to the glass is re-radiated right back if the glass is near the temperature of the plate.  This is yet one more thing that can overheat plates, boil off gas molecules adsorbed on the metal, poison the cathode, and lead to thermal runaway.

To keep the tube living longer, you keep it cool. The glass is ineffective at cooling the plate; about all you can do is keep re-radiation down, and that is done by keeping the glass temperature down. Glass that is cool will not re-radiate heat to the plate, so the plate does not get into thermal runaway.

Moving air is an effective way to do this. Even a wisp of air across a pair of fully screaming output pentodes will usually keep them cool enough to touch. Those of you with tube brands burned into your fingers will appreciate this. The makers specify that tube glass envelopes should remain under 200C - and water boils at 100C. Human skin sizzles at about 190C.

Cooling your tubes will not affect the tone. The idea that hot tubes are good sounding is a red herring in the technical jargon, originating as noted in the idea that one can bias tubes hot or cold. Higher bias currents (hot bias) does smooth out crossover distortion and many people prefer it. The tubes do get hotter this way, hence the term hotter biasing.

However, the internal operation of the tube will be the same at higher bias currents if you keep the glass cool - and the tube will live longer.

It is possible to bias tubes so hot that they produce so much heat that even cool glass will not keep them from melting internally. But if you have properly biased tubes, cooling will just make them last longer than they otherwise would for that bias level. And they will sound like the bias level tells them to, not affected by the glass temperature.

As for dust, there are fan grilles for computer cabinets that will hold a small filter. You can also get a foam filter membrane for household airconditioners at Home Depot and glue a small square on the intake side of the fan. The air the fan blows is then somewhat cleaner than the ambient air and should keep dusting down.

Fan cooling amps is a Good Thing.

[Paul Marossy]

Fan cooling amps is a Good Thing.

I agree. That's why ALL of my amps have fans in them!  

[Hal]

RG - thanks for the great info!  Right now im thinking of what i can do to increase the air circulation in the preamp section of my Peavey Ultra 112.  They don't get as hot as the power tubes, but it seems like a stupid design - while the power tubes are outside, the preamp tubes are in a small, restricted area with a panel over it...

is there _any_ cause for concern with respect to the overheating of preamp tubes?

[Paul Marossy]

is there _any_ cause for concern with respect to the overheating of preamp tubes?

I wouldn't be too concerned about it. Preamp tubes last several times longer than power tubes do. They just don't have to work anywhere near as hard, and they don't deal with the kind of bias currents that power tubes need to do their thing, which is part of what wears them out quicker.

[ESPm2M]

R.G, it was good of you to sort that out for those reading through this thread.  Well put indeed!

So can anyone tell me about the "NTC Thermistor" at this link http://www.bgmicro.com/lmad.asp  it is the third item down on the page.  Can it be used to automatically control fan speed?  What do you all think about the fan on at that same link (first item on the page)?

thanks!

[ESPm2M]

*Bump*

any comments on my last post?

[runmikeyrun]

If i posted what i'm about to tell you on another tube amp board, i would probably be castrated...

I had an old ampeg V-4B that i was running like it was an SVT.  I kept having problems with it melting down so after about 3 times my tech caught on that the board was getting so hot the solder was melting and the plate resistors (?) were falling out (Ampegs are laid out upside down compared to a Marshall).  If you're not familiar with the cosmetic layout of an old ampeg, it has a metal angled faceplate at the top and a piece of wood with grille cloth over it underneath.  I found out that the piece of wood was the perfect size for a fan.  I removed the grille cloth, carefully cut a pattern out of the wood, installed a fan, and replaced the grille cloth.  The back of the head is partially open to facilitate airflow to the tubes, so i had a perfect exit path for the air to go.  I drilled holes in the  chassis pan where the items were desoldering themeselves and riveted little airfoils to the chassis to direct air into the holes and towards the problem hot spots.  The rest of the air cooled the preamp and power amp tubes nicely as it exited the amp.  


I chose to use a 117vac fan for several reasons:  for one, i didn't want any added strain on the filament circuit with a 12v fan.  Also, the only 12v fans i could find were dc and i didn't want to get crazy adding a rectifier circuit.  I was also afraid the brushes in the motor would induce noise.  The 117 fan was easy to power: tap right off the leads where the power cord comes into the amp.  It was also brushless.  If you're concerned about fan noise during recording or whatnot, it would be very simple to install a spst toggle switch to turn off the fan whenever necessary.  

I'm sure a thermistor would work as well but you would have to first find out what temperature your components are at pre-fan, and what item is most important to keep cooled.  Then determine what an optimum temperature would be to run them at.  Then figure out where to mount the thermistor to keep your important items in check.  And after all that, you still might have problems if your fan doesn't move enough air to cool your items down your thermistor's temperature; your fan might still run all the time.  Since you wouldn't have to worry about your tubes being too cool, there's no harm in running the fan full tilt all the time.  However, being able to tell people that the fan is thermistor controlled scored major cool point...  

As for dust, if your fan doesn't actually blow air into the chassis, it would be very easy to clean any dust the accumulates on the tubes or trannies.

Incidentally, i sold my ampeg shortly after i installed the fan.  I never heard from the guy, so i'll have to assume it didn't melt down again.  

hope that helps,
Mike

[Paul Marossy]

The 120V fan isn't a bad suggestion. The Seymour Duncan Convertible uses one, which is an EG&G Rotron, and it's pretty quiet. The only problem with the 120V fan is that it might be hard to find one that is quiet, if that is a concern...

I personally think the fan ought to run at a constant speed, and continuously.  

[runmikeyrun]

yeah, i would think mouser carries 120v fans, if not mcm electronics should still carry them mcmelectronics.com

[R.G.]

The chances that a thermistor exists that would properly control fan speed are slim.  It could happen, but I wouldn't go that way.

I would go for one of a couple of different approaches.

First, I'd use one of two different fans - 12VDC or 240Vac. The 240Vac on 120 will start and run, but slowly and quietly. The only problem with that is that you have to run power line voltages around inside your amp, and that's a potential safety hazard. 12Vdc fans can be reliably controlled by low power circuits, so they're the obvious choice if you're going to do something fancy in terms of control.

I would use something like a silicon junction for a thermal sensor and then use an opamp with a transistor follower to power the amp. That way you can set it up to start the fan at any temperature you choose.

Or, if that's too complicated, rectify filament to 8Vdc and power it all the time.

You could also use a thermal switch to flip the fan on at the thermal switch's temperature, but it's not adjustable.

[ESPm2M]

These last few posts had a lot of great information!  Thanks everyone, this is really getting interesting.

I was originally thinking of using a thermistor in-line as you might a pot as a variable resistor.  That way it doesn't ever really go off, it just ramps up when necessary and visa versa.  I can see where it would be tricky having to find a thermistor with an appropriate range/sensativity.

R.G., Can you explain more about your silicon junction idea?  How exactly would I set this up?

... I'm getting excited about this project. Maybe we can design a simple circuit to contribute to the schematic section for future reference because this topic has definitely come up before in one form or another!

[R.G.]

See http://sound.westhost.com/project42.htm for one example.

[Torchy]

Have a look here for a proven (and quiet) thermistor-driven fan supply. I've used it myself, using 12VDC brushless fans from PC powersupplies (nice and quiet).

http://www.uoguelph.ca/~antoon/circ/auto-fan.html

[ESPm2M]

Now thats what I'm talkin' about!  :wink:  Thank you both  

Where can I find high quality glass enclosed NTC thermistors (I guess 50k) as mentioned by the article?

I will definitely let you all know how this project turnsout, and post some pictures.  I will buy the fans and thermistor and build the circuit to test it, but it won't be able to install it in the Mesa head for at least a month and a half I'm sorry to say.

[Paul Marossy]

www.mouser.com ?
They seem to have everything...