SPCR

i have been wondering about some things.

i seem to remember seeing somewhere that certain people are coating printed circuit boards w some kinda waterproof coating. could you do this to a motherboard and then bury the whole thing in a waterbath?


but another idea was that since deionized water doesnt conduct an electrical charge, could you put your motherboard in a bath of deionized water?

Only if you can ensure that no containmination is possible. (i.e. no dust, etc. would ever actually enter the water, causing it to start carrying a charge.)

The other thing to wonder about is... if it's a simple bath, that means the water just simply get warmer and warmer as time goes on...

So that means the water probably has to be mobile -- passing through a radiator, perhaps.

That reduces this to the water cooling idea, doesn't it?

Who says that deionized water doesn't carry a charge? Besides someone in this thread? Check out Horiba's The Story of Conductivity for a pretty good explanantion of water's electrical properties. They say the difference in conductivity between tap water and de-ionized water is about a factor of 100, with de-ionized water having a conductivity of less than 1.0 uS/cm, but necessarily more than 5.48x10^-2 uS/cm (the conductivity of distilled water). While this may seem like a poor conductor compared to copper's conductivity of 5.9x10^11 uS / cm (reference here), it's got nothing on air at sea-level, which has a conductivity of 2.5x10^-10 uS / cm (reference here). Forgive my cheesy references, but, alas, I have no CRC Handbook handy. If I did, I would also provide the conductivity of some type of oil for comparison, since I've heard stories of persons sucessfully submerging their computers in mineral oil (see here).

Anyway, my point is that I don't think submerging your motherboard in any kind of water is a good idea. Even if you've "waterproofed" it, you're just asking for trouble. Mineral oil or HFE is a lot less risky, although the latter is prohibitively expensive. And, as Ducky says, pretty much any of these submersion techniques will still require a pump to circulate the fluid, so the silence-value over traditional water cooling is nil.

cool info, thanks