you would think electronics would fail in a liquid..... crazy stuff...
http://www.pugetsystems.com/submerged.php

 

haha thats sweet,, one of my old homies I went to school with, Is a school teacher now and hes experimenting with a fish tank PC.. Kinda Crazy dont know how it will go . but he got a grant from the school for the project...

 

mineral oil will break down and become conductive...

Trichlorflouroethane, Flourinert, or whatever that stuff from 3M is called is the best solution, though the priciest at ~$500/gallon.

I prefer distilled water myself, but I don't submerge my stuff. I only watercool mine

http://picasaweb.google.com/Notarat/Quaddy#

 

Actually it's not really a new idea. Back in the 90's there was a company called Craig Supercomputers that made, well, super computers. These computers ran so hot because of the amount of power in them that they had to be submerged in a non-conductive liquid/gel tank to maintain temperature. The catch is to find and use a non-conductive liquid that won't corrode the components. If you can find that and can afford it then you can cool your system more effectively then any other type of cooling on the market.

 

Oil baths have been around for as long as I can remember - and that's a long time.

In water cooled systems, btw, use de-ionised water rather than just distilled.

The sixties and 70s supercomputer was Cray, not Craig. I think it's still going. A more commercially minded competitor was Amdahl, set up by the guy who designed the IBM 360 architecture.

Trike is a solvent used in dry cleaning and was once used in cleaning electronic components. It's carcinogenic and should not be used. It's not a coolant.

You can run electronics in any dielectric. THe trick is to find one that will also remove heat. Oil isn't particularly good for this but is OK in a pumped circuit with a cooling radiator. What it is very good for is providing a dielectric capability in high voltage applications where air might break down and allow flash over.

Rob

 

well i thought it was cool..... when ya walk into best buy you dont see computers in aquarium tanks filled with oil.......

 

thats badass... i like alienware's liquid cooling system, i wonder if theyll start screwing around with this

 

ya. oil initially is anhydrous and therefore dieelctric.

voltage breakdown and breakdown onset by heat thru eddy-current induction (in a metal core, like in a trandformer) it what makes it conductive. as well as water introduction thru condensation.

and yes, believe it or not, water is an insulator to a point. BUT, its typically distilled and deionized water, like a few others said.. its the impurities in water that make it conductive.

oil for use as a high voltage insulator, arc flash suppresant (to eliminate arc chutes on HV breakers and switchgear) and lubricant for linkage in switchgear and as a coolant has been around for almost as long as electricity has been around.

any time you see a big transformer with the fins on the side of the can or box, its an oil-cooled transformer.. the core is oil cooled, but the electrical connections are in open air.

i can't remember the designations for the groups, but theres forced air, oil cooled convection, forced air convection and forced air oil cooled. no oil pump tho.. works off convection.


HV switchgear (in substations) typically is oil-bathed. it almost HAS to be in order to handle 750KV and still be somewhat compact.

othewise, a code-mandated 16' is nessesary at that voltage between conductors and ground sources.

actually more like 15'6 i think, but, at THAT voltage, i'll take a little space.. lol

but, could you imagine a 50' wide breaker!? lol

(50' being for 3 phase and 16' per phase)

very cool tho..