[FizzledFiend]

ok here is the deal...after a long discussion with a family docotr friend of mine we stumbeld across an idea that would allow us to use left over LN2 from minor surgeries done @ his office (wart removal). they recive aprox. 2 gallons of LN2 every 3 days he says that most of the time the stuff doesn't get used..so theres an occasion for 2 gallons of LN2 to be used in a more scientific way

we want to be able to really work with this ln2 cooled rig more than once so condensation is a real issue. I can't think of a way that would totally isolate the cpu from the air to prevent this condensation...I first thought that I would braze a square metal tube to a heatsink then liquid foam the whole thing...but this restricts further tweaking of the board, and I don't think we could actually get enough foam on it.

so here is the idea...we take a socket from a burnt board and then take wires from the board we want to use and send those wires to the spare socket where we would mount our cpu and heatsink.. this would totally isolate the cpu from the board and condensation won't be a risk anymore to the board.

now my question is would using say a foot of wire to relocate each pin of the cpu be too much resistance for the cpu to work?

I ain't much of an electrician so thats why I am asking..besides the fact of all those little wires getting to the right spot what problems could you guys forsee with this idea?

[CamH]

I don't know about the socket extension, but here's an idea I had the other day:

Why not throw the motherboard in oil (or flourinert if you can afford it)? It will keep condensation away, and it will help cool things also.

[gusgizmo]

no, the latency caused by the extra length of wire would cause your board to go nuts. even if it did work, the extra latency would ofset alot of your gain by ln2 cooling. camH's idea is good, you put a tube on the processor for the ln2, and submerge the whole thing in oil. make sure you dont put the drives in the oil though. and probably better not put the psu in, the caps in it may or may not react badly to the temp, better not risk it, when there is no benefit. in any case, good luck!

[snvpa]

also by oil make sure you use pure mineral oil and not petroleum based oil. petroleum based products would corode your board very fast.

[stamasd]

Just so you know 2 gallons of LN2 won't last too long when used in an open circuit... maybe a day or so.

[Big_Baller]

why don't you use the standard silicone + dielectric grease + and some neopreme around the socket area?

You could also simply get a lot of dielectric grease and put it all around the socket everywhere. It comes off later with that electronics grease cleaner from radio shack.


http://www.octools.com/index.cgi?cal...ion/intro.html

or

http://www.phase-change.com/index.ph...s&id=25&page=1

[sandman001]

you want a big tube, with a flat plate for a base. Then you mount that like a heatsink, put in your LN2, and some kind of liquid to help.

[stamasd]

Don't forget that neoprene and other insulators become extremely brittle at LN2 temperatures (put a piece of rubber into LN2 for a minute, then take it out and drop it on the floor - it will break into pieces as if it were glass). And also that thermal contraction/dilation at such extremes become important. You would want to thermally insulate the motherboard and other components from the LN2 temperatures. I guess a solution would be to make a socket extension that would raise the CPU up, not by 1ft but 2-3 inches, so it could be completely immersed in the LN2. If you see what I mean.

BTW, if you try this use first a CPU that you wouldn't be sorry of losing just in case you end up with a cracked core... it has happened before.

[I.M.O.G.]

How many times more than once do you want to work with this LN2 rig?

If you are thinking twice, I would tell you your pushing it, and if you are thinking more than that then I would just wish you luck.

Remember guys, oil is not freeze-proof. He's going to have to go the insulation route - he isn't going to be able to get his hands on a fluid that can handle LN2 temps.

[sandman001]

You can't raise up the cpu socket.

You will also need to insulate everwhere from atleast the agp slot to the top of the bard, and likely from each side of the board, as the cold will go down the traces.

[CamH]

Originally posted by IMOG

Remember guys, oil is not freeze-proof. He's going to have to go the insulation route - he isn't going to be able to get his hands on a fluid that can handle LN2 temps.

Of course it would freeze, but I don't think that would hurt anything. The oil would only be there to keep water away from the motherboard itself. Oil is more dense than water, also, so even the condensation that gathers on top of the oil will simply sit at the top.

[sandman001]

Unless it runs down off the oil onto another component that doesn't have oil on it.

remember, you wil have some little chips that need air and can't be coated in oil.

Like the mosfets, and some chips on the motherboard down by the southbridge.

[stamasd]

Originally posted by sandman001
You can't raise up the cpu socket.

Yes you can. It's been done before. Several inches shouldn't pose a big problem. Remember the slotkets? I have a P3/1400 running on a slotket, which adds at least 3 inches to the traces.

You will also need to insulate everwhere from atleast the agp slot to the top of the bard, and likely from each side of the board, as the cold will go down the traces.

A solution to that would be to blow warm dry air over the back of the motherboard.

[tgp]

ln2 has been done alot before. the problem that arrises is that its so cold it cant be stopped that easy with normal anti codensation methods.

ive seen a guy ln2 cool the gpu and just in a few min it had the whole card about 1/4" of ice over it. there is nothing cheap that you can put over that can stop something that cold. so what they do is just not even put dialetric grease or neoprene or anything, whats the use?

i think most end up with a frozen cpu in the board.

[stamasd]

For those of you wanting performance stunts like this but who can't get LN2, there's always the old chemistry lab trick of "rapid freeze" bath of acetone (or methanol) + dry ice. It will bring the temperature down to about -50 degrees (compared to -77 for LN2).

[Tirbyren]

IS there a max low temp a computer can handle before it just freezes up, literally?

[Gerfey]

umm... stamasd, liquid nitrogen is a lot colder than -77, like another 120 degrees colder than that, if you're dealing in celsius...

Anyway, if you're just doing it to see how much of an oc you can get, then just let it freeze, just make sure everything stays really cold while you're doing it. As long as the ice doesn't melt, nothing is going to short out. Just make sure you turn everything off before you run out of N2(l), and then let everything dry thoroughly before you power it up again. That should probably work

[I.M.O.G.]

Originally posted by CamH


Of course it would freeze, but I don't think that would hurt anything. The oil would only be there to keep water away from the motherboard itself. Oil is more dense than water, also, so even the condensation that gathers on top of the oil will simply sit at the top.

But is frozen oil more dense or less dense than fluid oil?

Frozen water is less dense and it expands, and if frozen oil does, I wouldn't think you would want it expanding around and inside of the socket, CPU, and cooling mount.

I guess maybe it wouldn't hurt anything, but I'd rather state this as an uncertainty, than say it wouldn't hurt anything and be proven wrong when he tries it.

[Ven0m]

I've read somewhere that guys cooled CPU with some kind low-freezing-point fluid cooled by nitrogen - it run almsot ok, but that fluid was frozen by nitrogen - not completly, but hardly moved.

Another attempt was just nitrogen block and comp didn't work properly.

Anyway it would be nice to see pics & results of your attempt.

[CamH]

Originally posted by IMOG

Frozen water is less dense and it expands, and if frozen oil does, I wouldn't think you would want it expanding around and inside of the socket, CPU, and cooling mount.

I guess maybe it wouldn't hurt anything, but I'd rather state this as an uncertainty, than say it wouldn't hurt anything and be proven wrong when he tries it.

Oh, right. I didn't think of that. I guess it would be a good idea to test first...

Or maybe the components will be strong enough to hold it off. Liquids (and oil will still be kind of liquid while turning to ice) will normally go to the area with the lowest amount of pressure.

[stamasd]

Originally posted by Gerfey
umm... stamasd, liquid nitrogen is a lot colder than -77, like another 120 degrees colder than that, if you're dealing in celsius...

Anyway, if you're just doing it to see how much of an oc you can get, then just let it freeze, just make sure everything stays really cold while you're doing it. As long as the ice doesn't melt, nothing is going to short out. Just make sure you turn everything off before you run out of N2(l), and then let everything dry thoroughly before you power it up again. That should probably work

Oops, you're right, got my numbers all messed up. I was thinking 77K.

Still an acetone+dry ice rig would go to -55 and below, for temporary stunts that should provide a cheap starting point.

[stamasd]

Originally posted by IMOG


But is frozen oil more dense or less dense than fluid oil?

Frozen water is less dense and it expands, and if frozen oil does, I wouldn't think you would want it expanding around and inside of the socket, CPU, and cooling mount.

I guess maybe it wouldn't hurt anything, but I'd rather state this as an uncertainty, than say it wouldn't hurt anything and be proven wrong when he tries it.

Water is an anomaly, almost all other fluids become denser when frozen. To directly answer your question, yes, frozen oil is heavier than liquid oil and falls to the bottom. I've done that experiment personally.

[stamasd]

Originally posted by CamH


Oh, right. I didn't think of that. I guess it would be a good idea to test first...

Or maybe the components will be strong enough to hold it off. Liquids (and oil will still be kind of liquid while turning to ice) will normally go to the area with the lowest amount of pressure.

Put some water in a glass bottle (say 1/4 full) and leave it outside in the winter or put it in a freezer. Look it up the next day. What happened? The bottle broke. That's because water expands when frozen, and no, it doesn't "go to the area with the lowest amount of pressure" while doing so, it freezes and expands in place - and as a result, the bottle breaks due to increased pressure.

Fortunately oil doesn't behave like water though.

[CamH]

Originally posted by stamasd


Put some water in a glass bottle (say 1/4 full) and leave it outside in the winter or put it in a freezer. Look it up the next day. What happened? The bottle broke. That's because water expands when frozen, and no, it doesn't "go to the area with the lowest amount of pressure" while doing so, it freezes and expands in place - and as a result, the bottle breaks due to increased pressure.

Fortunately oil doesn't behave like water though.

I'm not sure, but I think that if you don't put some sort of lid on it, it will expand upwards a bit and freeze... I dunno.

[stamasd]

No lid. Bottle breaks. Period.

[Fu_Man_Chew]

All I know is water expands a lot! when frozen. and under cretin circumstances can be capable of a lot of damage.

ever see water freeze in an engine block... or crack a rock the size of most 1/2 ton trucks.

Defiantly find out how oil reacts when frozen....

Nitrogen boiling point -196 Celsius/-320 Fahrenheit "on the chilly side"

I like the idea of moving the socket up off the mobo if you could some how make a box around the cpu out of a really good insulator...

O.K. Now I always wanted to do this. Make a case that is air tight... AIR TIGHT!! Fill the case with some form of gas. I was think nitrogen they use it in race car tires due to the fact it expands very little when heated compared to air.(may have the same property's when cooled) Fill the case with this gas I think 95%+ should be fine. the main point is we do not want a souse of condensation in the first place. (no water vapor no condensation) the cpu will be seperated from the rest of the case by the raised socket (only need 2 inch or so) and be out side the case immersed in ln2.

case will kinda look like this

http://www.geocities.com/cid_2120/newbitmapimage.JPG

O.k. now the ln2 will cool the gas in the case cuz as we all know its really really cold. If the cpu cup part is metal should conduct the heat out of the case quite nicely. and a plus to using nitrogen as a gas the ln2 will not freeze it. Also you might need to have a gutter or something like that around the cpu cup if the ln2 condenses the n2 gas (Don't think this will happen but it might). In the end one cold computer!!

[gusgizmo]

it wouldnt be hard to do that. make some gaskets for the case to sela it, and then just inject ln2 into your block that you have made for the job. voila, the case is filled with ln2, and your processor is nice and chilly. i dont think youd need a socket extender for this.

[FizzledFiend]

ok so the majority say it can be moved but only a nominal amount....I can't see a real easy way to do this with the size wire need and it be strong enough to suport a cpu/heatsink/gallon of LN2 so the second Idea was install everything into the board cpu, video, ram,ide cables etc. etc. then submerge the whole thing into a plasti-dip bath. Plasti-dip for those that don't know it is like rubber paint, it dries to solid rubber. Only problem I forsaw with this would be the covering of like the chokes, transistors, and mofsets...stuff that produces heat that cools thru ambiant air. With everything coated in the plastic it might have a hard time staying cool, but with ln2 cooling the immediate area it maynot be so bad. the pluss side is that EVERYTHING is insulated from the air so no wory of condensation...heck were talking about a water PROOF mobo LOL.


second question for the day...would I need to cover the cpu with a heatsink or heatspreader or just construct a cup around and up from the socket and just straight LN@ onto the cpu...worry about the instant freeze and cracking of the chip.....

this looks prommissing and I look forward to trying this out

[I.M.O.G.]

Originally posted by stamasd


Water is an anomaly, almost all other fluids become denser when frozen. To directly answer your question, yes, frozen oil is heavier than liquid oil and falls to the bottom. I've done that experiment personally.

I'm not concerned with rising or falling in the fluid... I'm just considering the effect of expansion/contraction of a rapidly frozen fluid around the socket.

I read your next response though also, so do you know how much the density of oil changes when frozen?

[stamasd]

Originally posted by IMOG


I'm not concerned with rising or falling in the fluid... I'm just considering the effect of expansion/contraction of a rapidly frozen fluid around the socket.

I read your next response though also, so do you know how much the density of oil changes when frozen?

Well, rising or falling is a direct consequence of volume variation, given Archimedes' law. A fluid that expands when frozen will rise to the surface; a fluid that contracts when frozen will fall to the bottom. As for how much, I don't have numbers (I'm not setup to measure densities of oil crystals), but I'd appreciate not much; certainly less than the volume variation of water.