h a l f b a k e r yInvented by someone French.
add, search, annotate, link, view, overview, recent, by name, best, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
Take a Bell jar and drill a few not-quite-randomly-placed holes in it. Take a nice big channeled radiator run the inlet pipe to the hole in the top of the jar and the outlet to the one in the bottom. Place your nicely heatsinked motherboard inside the jar, run the cables out through a plugged hole
and place the entire contraption on top of whatever they call the flat gasketed plate you put a Bell jar on top of. Put a few inches of water into the jar, making sure to cover the heatsinks by a good margin. Evacuate the jar to the point where water boils at below whatever temperature you want to limit your components to, remove the pump and plug that hole.
Turn the power on and watch the water boil in the jar, condense inside the radiator and flow back into the jar.
The radiator can of course be made integral to the jar for a more "heatpipe"y-looking contraption.
What dielectric constant numbers mean.
http://en.wikipedia...onstant#Measurement
[bigsleep, Jul 31 2009]
[link]
|
| |
//covering the heatsinks by a good margin// |
|
| |
You better be using a double-sided motherboard or the graphics card is going to get wet. Not that it matters as the MB is going to get condensation all over it. |
|
| |
The idea started well. I was expecting a closed air circuit with a centralised heatsink. |
|
| |
//or the graphics card is going to get wet// |
|
| |
The entire MB is under a few inches of water so yes the graphics card is indeed going to get wet. |
|
| |
On your planet does water not conduct electricity ? |
|
| |
Indeed on my planet water does not conduct electricity. Of course washing the electronics beforehand and using deionised distilled water will save you a bundle in motherboard replacement costs. |
|
| |
I thought of using a liquid with a much lower boiling point: that way you wouldn't need a low-pressure jar at all, but I couldn't think of one offhand with a high enough dielectric rating and heat transfer properties that isn't poisonous or explosive. |
|
| |
One Google later, I'm getting similar resistivity figures for DI water and air. I'm also getting that copper could be a major problem for stray ions. Novel idea though, maybe spray coat everything with a thermally conductive but electrical insulator ? |
|
| |
never thought of the board heatsinks dissolving or oxidising; spraycoating the rest of the stuff should be enough though. |
|
| |
One more Googling session later, the dielectric rating is not important but the dielectric constant is. As you point out, resistivity is not a problem initially (until things start oxidising or CO2 starts dissolving), but the biggest factor is that water has a dielectric constant 100 times that of air. This means all those exposed lines have 100 times the capacitance between them - thereby screwing up all the high frequency stuff. Best not get your CPU socket wet methinks. |
|
| |
I think you're reading the numbers upside-down. |
|
| |
[edit] I still think you're reading the numbers upside-down. High numbers good, low numbers bad if what you want is insulation. |
|
| |
Insulation is dielectric strength, capacitance is dielectric constant. We know DI water is initially ok and won't break down, but the capacitance is a problem. |
|
| |
Knocking around we have - air 1.0, the MB itself 5.0, water 80. |
|
| |
//Take a Bell jar and drill a few not-quite-randomly-placed holes in it.// |
|
| |
How do you evacuate the jar with those holes in it? |
|
| |
While I admit to mild disappointment that apparently you'd have to spray the board connectors, I still have no clue what you're on about regarding capacitance. Yes there are capacitors, I'm going to qualify that you shouldn't use a board which has open-air capacitors; in fact while I think of it, putting a heatsink on the power supply instead of a fan might be an idea as would using an SSD and let's keep the DVD as a an external USB drive. |
|
| |
But I don't see anything that benefits from the ability to short out (in air). |
|
| |
// I still have no clue what you're on about regarding capacitance// |
|
| |
It's back to the old argument about what a capacitor is. Basically a capacitor is two conductors separated by a distance. On the MB if two tracks go parallel to each other they will form a capacitance that needs to be taken into account when designing the MB (MB dielectric constant is 5). Those little connectors on RAM modules when immersed in water effectively become 80 times as long due to the dielectric power of water. At 1Ghz FSB even the smallest extra capacitance is going to mess things up. |
|
| |
\\but I couldn't think of one offhand with a high enough dielectric rating and heat transfer properties that isn't poisonous or explosive.\\ why is that a problem? |
|
| |
//why is that a problem?//
repeatability. |
|
| |
//capacitance//
still lost... assume a total lack of knowledge of electrical circuits here: insulation good, crosstalk bad; water insulates more than air, water = good. |
|
| |
Now we're getting somewhere. The amount of crosstalk is a function of the capacitance. See [link] which shows that by putting water in the gaps you get 80 times more crosstalk. |
|
| |
umm, no, you'd get 80x *less* crosstalk. |
|
| |
Read the link. Crosstalk is proportional to capacitance. |
|
| |
you mean the part where it says: |
|
| |
"If a material with a high dielectric constant is placed in an electric field, the magnitude of that field will be measurably reduced within the volume of the dielectric." ? |
|
| |
No. The section linked titled "measurement". |
|
| |
ok, so put some caulking around the base of the CPU, though I don't see the air actually going anywhere anyways. |
|
| |
I see a little little bit of a problem: |
|
| |
You've got a radiator (outside the bell jar), with pipes connecting the inlet and outlet of the radiator to the top and bottom of the jar. |
|
| |
Since the pressure inside the jar and the radiator are the same, the temperature at which water boils/condenses is the same in both locations. |
|
| |
Any liquid water coming out of the radiator will be no cooler than the water being boiled. |
|
| |
Since all of the components are flooded in water, no component will be cooler than the hottest component. |
|
| |
So potentially, you could be doing more harm than good with this system. |
|
| |
//Any liquid water coming out of the radiator will be no cooler than the water being boiled.// |
|
| |
even if that were true then at least the heat of vaporization will have been removed, which is quite a chunk of therm. But it isn't since you've set the thing to boil at some temperature *above* ambient and the nice finny radiator dissipates the heat into the surrounding air. |
|
| |
I just saw an article in an overclocking magazine about oil cooling. They took the whole motherboard and submerged it in oil and it worked flawlessly. |
|
| |
are there additives you can mix with water to reduce conductivity ? |
|
| |