Need an extreme pc cooling solution?

Just submerge it in some mineral oil.


It seems like a cool idea to me. . . though as dust and the like begin to accumulate in the oil it could get nasty, and moving/cleaning could be a nightmare...

 

Oil is a viscose material and would hold the heat locally, and its thermal properties are not great(ie it does not conduct heat as quickly as say Copper) on top of the concerns you mentioned. Also the fact that a lot of boards are mounted vertically. Great thinking but I'd say needs more thought

 

The usual solution is inert florine, which works great, but is $100-500 / gallon. Big companies use it, but not the individual users.

Side note, this covers a wide range of compounds, some of which can be oxygenated and breathed by mammals.

 

this is amazing how come it doesn't short out. Why doesn't the oil affect it? Tell me more i am fascinated.

 

Oil isnt conductive,to me the best extreme cooling is to put your whole box inside chest freezer with holes cut for the cdrom,mouse and keyboard ect,the whole system would sit confortably at -60,keeping it on all the time your though would send your electricity bill up a few $$ a quarter,turning the freezer off everyday would kill it pretty quick due to condensation

 

Been done.
Toms Hardware for the win!
But i must admit using a aquarium bubbler to circulate the oil is a intersting idea. Given your using thin enough oil i whould imagen that aquarium filtration would work as well....

 

wouldn't the oil heat up??? A bit like cooking oil but less effective. I don't see how it would cool it better than air?

 

*Bump* still need answers lol

 

It could work, because oil is denser and should hold more heat. Then there is the boil factor, anytime a liquid boils it absorbs a lot more calories due to the phase change. Does it work? Personally I doubt it, it's big thing is coolness factor.

Inert flourine becomes a lot less viscous when heated, so it has been used, so has water, but it is a lot more expensive.

 

Also, if you look at the closeups of the PSU, you will see that between it's fan and the fan mounted above it, they actually do have some circulation going on before the bubble bar is added. in theory, you could seal the system in there (sans bubble bar) and it would eliminate the dust issue. My concern is that if you are circulating the oil, where does the heat go from there?? I know that you will loose some to the ambient environment via conduction through the glass, but I just can't see it being as efficient as a fan cooled system. plus, when you need to work on it... ewww.


~C

 

that would be my problem with this method, changing out memory, expansion cards and hard drives would just plane be too messy to work with. I like the "place it in the freezer" method myself.

 

Its a cool idea on first thought but you cant easily change any components without loosing alot of oil or making a big mess. I dont see whats wrong with a fan to be honest

 

Sometimes air isn't enough, especially when the component want to go to really high temps, and you don't want them to.

Case in point, old lasers used to use chilled water, they had a xenon tube that was continously lite with 300VDC. Left to themselves, they would shatter in seconds, so they are washed in a bath of chilled DeIonized water (aka ultrapure water), which is nonconductive. The now heated water goes to a heat exchager, where the waste heat is dumped into a house chilled water system, and the DI water is reused.

We are seeing something similar happening with CPUs, not quite as complex but the water cooling systems are really better. The other option is TEC, or massive copper heatsinks, but chilling works better.

 

is it true that CPU's actually preform faster when cooled? and if so why?

UKARMYCADET

 

I've never heard that, although I do believe I've heard about a feature that slows em down if they get too hot, to keep from burning up an expensive piece of gear. And sometimes, when the CPU goes, so goes the motherboard.

A hot piece of silicon definately has a shorter life span. When we "age" our electronic components at work for testing purposes we do it under power in an oven, where 1 day = 1 week of service under normal conditions.

Heat speeds up a lot of natural processes, and this makes it the enemy.

 

Check out the Toms hardware link above.. explains most of the questions being asked