Need to cool masive ammounts of heat?

[Lisandro_P]

This looks quite promising, though it'd look funny in an amp: evaporative water cooler. Basically, works just like a nuclear plant cooling chimney. It simple, easy to build, and can cool massive ammounts of power easily. Won't look very good in the living room though...

I'm interested in... umm.. "esoteric" ways of cooling. In any case, i'm about to spend a couple hundred in heatsinks so i'll probably end up cooling BJT's the old fashion way, but these weird approaches interest me.

Guess no one has tried anything similar eh?

 

[markp]

You might want to look into 'heat pumps' as a good way to remove the heat. They are essentially a tube filled with a refrigerant gas that has the heat source on one end and a bunch of fins on the other to get rid of the heat. NAD and Luxman have used them among others.

 

[Lisandro_P]

You might want to look into 'heat pumps' as a good way to remove the heat. They are essentially a tube filled with a refrigerant gas that has the heat source on one end and a bunch of fins on the other to get rid of the heat. NAD and Luxman have used them among others.

I had little luck finding prices for these, any idea on how much they run for? What atracted me from the evaporative device it's that it gives a terrific performance for very little money, in adition of being an unusual design.

 

[till]

problem with boiling water to cool the transistors is the case will be > 100°C.

Cheap way for loads of heat dissipation would be mounting transistors on one of these: http://www.aircoline.gr/English/eproducts/eheating3/Kermi.jpg

you fill with water and depending on size it will be good for some hundred to thousends Watt at ~ 60°C

 

[fr0st]

Bong's actually cool alot better than radiators when used properly. They work sorta like a shower, the water coming out of the head isn't 100c but it still steams up the room without an exhaust fan.
They work remarkably well and can operate in some circumstances below ambient temperatures.

 

[Lisandro_P]

Two concens; the water cycle is open to the enviroment; that means algae and bacteria clogging the flow if the water's not treated, and, working by heat release via evaporation, water will drain in the long run, which is a pain in the ***. Still, the performance just can't be beat.

The radiator idea is intriguing... you mean that just by filling it with water and hooking transistors to it (drilling a radiator or soldering a metal tab would be requiered) would work? It could at reasonable temperatures, but i don't know how much power we're talking about here...

 

[AudioFreak]

problem with boiling water to cool the transistors is the case will be > 100°C.

You are never actually boiling the water, just raising it's temperature a little and letting the natural evaporative process take it's course in the optimum environment provided by the cooling tower.

 

[till]

The radiator idea is intriguing... you mean that just by filling it with water and hooking transistors to it (drilling a radiator or soldering a metal tab would be requiered) would work? It could at reasonable temperatures, but i don't know how much power we're talking about here...

yes, thats the idea. i allready have a raditor here. i would mount the transistors only on one side, for a circle flow inside the radiator could appear.

for the amount of heat you could cool this way, the manufacturer gives specs at some temperature differences for them. i think its also possible to glue the transistors to the surfaces, so no drilling would be necessary

 

[Lisandro_P]

yes, thats the idea. i allready have a raditor here. i would mount the transistors only on one side, for a circle flow inside the radiator could appear.

Sounds good... how much did you paid for the radiators?

 

[till]

i do not pay anything for them as i have a friend who is in that business. So i take surplus of his business. But i fear this will help you not much unless you are happy with something like a 400*500mm radiator, as i doubt we will manage to send a larger on over the atlantic...

 

[Sch3mat1c]

I wonder how the cooling tower would work with a spray of oil. Of course then you have the problem of oil fumes (not much, but the finer droplets would be blown out by the fan).

Great for California but marginal, as the method goes, around here with humidity upwards of 80%.

Evaporation also demands a shot of distilled water every so often due to evaporation. You wouldn't want to use tap water, especially around here due to minerals (hard water). And yeah, algae would be a problem. How's antifreeze work for something like that? Anyone know the toxicity of ethylene glycol in air, as well as the vapor pressure? (Or just vent it to the outside, which you should do anyway so you don't fight your air conditioning!)

Tim

 

[Lisandro_P]

Originally posted by Sch3mat1c (...)
Evaporation also demands a shot of distilled water every so often due to evaporation. You wouldn't want to use tap water, especially around here due to minerals (hard water). And yeah, algae would be a problem. How's antifreeze work for something like that? Anyone know the toxicity of ethylene glycol in air, as well as the vapor pressure? (Or just vent it to the outside, which you should do anyway so you don't fight your air conditioning!)

Water evaporation is the main issue here; algae can be controlled to a degree by using water cleaning products (much like swimming pools; chlorine and soluble products), enough so it doesn't become a problem in the short term.
What's left is that you have an humidifer for an amp with this system, never mind the need to feed it with water regularly. You can't attempt to trap the humidity as the whole system needs heat exchange with the enviroment. A fine metal grid works quite well, but you're still shooting vapor out.

I can't seem to figure how to sort these issues out; otherwise, if you need to cool a lot of heat for cheap (thinking of building that 200w class A? ), it's simply the best choice.

 

[Sch3mat1c]

You could make it a sealed system that operates at 100% humidity and have a huge heat exchanger with the outside, condensation would flow back to the sump where it's sprayed back in. Basically a large, highly ineffective, corrosive (i.e. rust, etc.) refrigeration system. But it would be cheaper to hack open one of those teeny fridges and cool the die directly.

Tim

 

[Lisandro_P]

You could make it a sealed system that operates at 100% humidity and have a huge heat exchanger with the outside, condensation would flow back to the sump where it's sprayed back in. Basically a large, highly ineffective, corrosive (i.e. rust, etc.) refrigeration system. But it would be cheaper to hack open one of those teeny fridges and cool the die directly.

That would make it no better than... well, a radiator

I was thinking, was GoreTex the fabric that let air through but not moisture? That could work to a degree...

 

[Sch3mat1c]

Well, it doesn't let liquid through. Obviously since water molecules are on par with oxygen and nitrogen (slightly smaller actually), humidity will come through all the same. Might as well use a fabric filter though, ought to be more permeable. Though you won't get any moisture running down off it, unless there's enough to saturate it. But then you have the problem of enough fan pressure to blow through the moist fabric (a good squirrel cage ought to do the trick), and I don't know if that kind of velocity would just blow the droplets either off the fabric, or right on through, anyway

Tim

 

[rikkitikkitavi]

one other option could be to use the condensor/evaporator radiator
found in air/air heat pumps, f e x radiators or A/C
can probably be found for very cheap (scraped) at a workshop/firm that sells this
type of equipment .

they usually have quite a large surface, several m2, even in the smallest ones. , and of course comes in different sizes from small A/C to large installations for larger buildings.

They are also comparativly lightweight, since it is only copper tubing and thin aluminium sheet.

circulate the water through this , and the heat will be radiated off very efficently. since the system is closed, no filling of water is necessary, and using distilled or battery water prevents any fouling due to dissolved minerals.

I tested with one sizing 15*25*50 cm , area 7,5 m2 by feeding hot tap water through it. in 55 C, out 23 C, cooling power about 850 W!

Adding a fan increases cooling power many times.

If you want to go completely off the line, replace the water with som e heat transfer liquid like Fluorinert (about 1-200 US$/liter) that is inert , nonconductive, nontoxid, nonflammable e t c.


/rickard

 

[jacco vermeulen]

Does anyone have experience with closed circuit fluid cooling,
e.g. :Hugh Dean's Glass Harmony design?

I am looking for examples of fluid cooled heatsink designs, or a link to a Mathlab simulation program for fluid cooled heatsinks to estimate cooling distribution.

I am thinking of using thermal oil for cooling, some chemical tankers use thermal oil for heating liquid cargo.

The heat transported by thermal oil from the amplifiers heatsinks would be transferred to water in a heat exchanger.
For the sailing yacht i built i made an intercooler that should also work well for heat transfer from amplifiers.
On my Chevrolet i use braided lines from the engine to the radiator and the oilcooler for the TH700r4 automatic gearbox, their easy connectors should hook up easilly on an amplifier.

The waterflow coming from the intercooler i intend to connect to floorheating ( plastic tubes embedded in concrete) combined with an exterior airconditioning unit.

Any suggestions are welcome.

 

[markp]

Here is some info on 'heatpipes' that is very helpfull. It is a sealed pump-free system of liquid cooling. Luxman, NAD and many other pro amps have used this type of cooling.
http://www.enertron-inc.com/HeatPipeBasics.html

 

[jacco vermeulen]

Thanks for the link, nice to read up on.

I have used heatpipes before on a DIY amplifier.
Unfortunately for this purpose i need something that can handle 2500 watts.
At 50 Celsius sink temperature its 0.01 k/W
Heatpipes that big are not produced, and the heat is still inside the room.

Thanks

 

[TwoSpoons]

Its really hard to beat water as a heat transport fluid. It has :
a very high specific heat (4.1 J/degC/cc IIRC) = huge heat transfer per unit volume
low viscosity = easy to pump fast
zero toxicity = no spill worries and you can drink it if you're thirsty
zero cost, for those on a budget
ready availability = no UPS courier charges
excellent materials compatability = it wont eat its way through plastic tubing!