submerge whole board in flourinert?

lel4866 · 07-01-2003, 03:53 PM

OK, I'm new here and this has probably been brought up before, but...

Why not submerge the whole board in a sealed case with something like flourinert? Like they used to cool core memory with oil?

Q:What's a virgin core?
A: One that's never been fluxed... (showing my age)

Larry Lewis

sevisehda · 07-02-2003, 07:54 PM

Its been done somewhere. They actually got fairly decent results but considerign teh cost I doubt it was a long term solution.

xerka · 07-03-2003, 12:21 AM

Here is one guy who did it, click on the different subbmersion links. http://octools.com/index.cgi?caller=submersion.html

airspirit · 07-07-2003, 10:50 AM

You do know how expensive that stuff is, right? You'll end up spending well over $500 USD to do a proper submersion.

redleader · 07-07-2003, 01:50 PM

Whats the point? Watercooling is cheaper and better.

Euphorius · 07-07-2003, 02:25 PM

Well, you could just use mineral oil. It's only $55 for 5 gal., lab grade. It won't allow the crazy temps fluoro will, but it will get you below 0F.

jaydee · 07-07-2003, 02:29 PM

Quote:

Originally posted by Euphorius
Well, you could just use mineral oil. It's only $55 for 5 gal., lab grade. It won't allow the crazy temps fluoro will, but it will get you below 0F.

I am confused as to how minieral oil will get you below zero?

airspirit · 07-08-2003, 09:59 AM

I think he means before the stuff is too thick to flow anymore, though I could be mistaken. This assumes he dunks a phase change unit into it as well or some such craziness.

bigben2k · 07-08-2003, 10:26 AM

Thanks to xerka for the link: I couldn't remember the site name myself!

What should be remembered also, is that there still needs to be a heatsink on the CPU, because the immersion can't take that heat by itself.

lel4866 · 07-08-2003, 10:28 AM

When I originally posted, I was thinking the following:

1) Whole submersion cools all chips. I keep seeing: cool the northbridge, cool the southbridge, cool the mosfets, cool the underside...What's next?

2) No complicated water blocks or tubing. Solve one complicated problem once. Of course, now that I've been reading for awhile, it seems lack of direct impingement will significantly decrease the systems cooling ability on the chips that matter. Maybe the containing box needs to be designed like one big waterblock. Maybe the liquid enters through special jets over the cpu? The motherboard itself provides the turbulence.

3) I'm not concerned about the cost of flourinert at the moment - just its efficacy. How come there in't more enthusiam for it?

4) The first problem is creating a sealed box where I can bring out the connectors and stay sealed. Eventually, I see a neat looking clear acrylic box with cooling equipment, power supply, and disk drives mounted below in a sleek black box.

Larry Lewis

Dr. Strangelove · 07-08-2003, 12:26 PM

Hmm check out this thread over @ 2cpu.com it's been going on for quite some time discussing the best way to submerge your system in oil for cooling. Has to be said that the primary objective is silence, but I'm sure that many of the tings can be used for any type of submerging.
http://forums.2cpu.com/showthread.php?s=&threadid=32859

MadDogMe · 07-09-2003, 03:02 AM

I thought remembered reading somewhere of an 'equivilent' product (UK made?) that sells for much less, £50~75?...

Can't remember where I read it though :shrug: , I thought it was that site

...

airspirit · 07-09-2003, 09:44 AM

What you would want to do is take a medium flow rate pump and create a manifold so you can direct coolant flow over the heatsink on the CPU, NB, GPU, and anything else you feel will get extra warm. You will want the absolute largest heatsinks you can find for all of those chips as well. You will need an external heat exchanger hooked to a cooling loop to keep the flourinert cool (you do NOT want to just dunk a phase change coil into it). Yeah, you can do this, but it isn't easy to do it right. You must keep the oil chamber sealed perfectly from humidity, or you will end up with water contaminating your oil. You must also find a way to route your cables in and out of that perfectly sealed chamber. It is easy to do this for small periods of time, but extremely difficult to do a decent job with this for the long haul.

One method would be to keep the CD/HDD drives in the chamber as well, but keep a pocket of air in the top of the chamber for these drives to stay in. Off of the jet manifold, add a "water block" to the HDD to keep it reasonably cooled, and the passive cooling of the air over the oil will hopefully keep that cool enough to keep your CD drive cool (you will need to seal this container on the inside to prevent ANY air leaks). As far as the other connections go, you will need to install extension cords for the USB (add in a hub up top), PS/2 (does anyone use these?), VGA, and sound to points in the air gap and seal them to the outside.

This will be a pain in the ass. Good luck.

lel4866 · 07-09-2003, 10:08 AM

Now that I've read various comments, I agree that you probably have to put heatsinks on the warmer parts anyways, or equivalently complex manifolds for direct impingement.

That negates the benefits of the whole idea. Might as well just put good waterblocks on the hot logic parts and good heatsinks on the warm parts or things like the mosfets which perform pretty good hot, anyways.

Thanks for all the inputs, everyone. I, for one, am satisfied to look elsewhere for easier gains.

Now, on to the backside cooling waterblock!

Larry Lewis

07-01-2003, 03:53 PM	#1
lel4866 Cooling Neophyte Join Date: Jun 2003 Location: Beaverton, OR, USA Posts: 10	submerge whole board in flourinert? OK, I'm new here and this has probably been brought up before, but... Why not submerge the whole board in a sealed case with something like flourinert? Like they used to cool core memory with oil? Q:What's a virgin core? A: One that's never been fluxed... (showing my age) Larry Lewis

07-03-2003, 12:21 AM	#3
xerka Cooling Savant Join Date: Mar 2001 Location: Denver, CO Posts: 107	Here is one guy who did it, click on the different subbmersion links. http://octools.com/index.cgi?caller=submersion.html __________________ Folders: I've given up folding

07-07-2003, 10:50 AM	#4
airspirit Been /.'d... have you? Join Date: Jul 2002 Location: Moscow, ID Posts: 1,986	You do know how expensive that stuff is, right? You'll end up spending well over $500 USD to do a proper submersion. __________________ #!/bin/sh {who;} {last;} {pause;} {grep;} {touch;} {unzip;} mount /dev/girl -t {wet;} {fsck;} {fsck;} {fsck;} {fsck;} echo yes yes yes {yes;} umount {/dev/girl;zip;} rm -rf {wet.spot;} {sleep;} finger: permission denied

07-08-2003, 09:59 AM	#8
airspirit Been /.'d... have you? Join Date: Jul 2002 Location: Moscow, ID Posts: 1,986	I think he means before the stuff is too thick to flow anymore, though I could be mistaken. This assumes he dunks a phase change unit into it as well or some such craziness. __________________ #!/bin/sh {who;} {last;} {pause;} {grep;} {touch;} {unzip;} mount /dev/girl -t {wet;} {fsck;} {fsck;} {fsck;} {fsck;} echo yes yes yes {yes;} umount {/dev/girl;zip;} rm -rf {wet.spot;} {sleep;} finger: permission denied

07-08-2003, 10:26 AM	#9
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	Thanks to xerka for the link: I couldn't remember the site name myself! What should be remembered also, is that there still needs to be a heatsink on the CPU, because the immersion can't take that heat by itself. __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

07-02-2003, 07:54 PM	#2
sevisehda Cooling Savant Join Date: May 2003 Location: NY Posts: 234	Its been done somewhere. They actually got fairly decent results but considerign teh cost I doubt it was a long term solution.

07-07-2003, 01:50 PM	#5
redleader Thermophile Join Date: Jun 2001 Location: The deserts of Tucson, Az Posts: 1,264	Whats the point? Watercooling is cheaper and better.

07-07-2003, 02:25 PM	#6
Euphorius Cooling Neophyte Join Date: Jul 2003 Location: Paragould, AR Posts: 1	Well, you could just use mineral oil. It's only $55 for 5 gal., lab grade. It won't allow the crazy temps fluoro will, but it will get you below 0F.

07-08-2003, 10:28 AM	#10
lel4866 Cooling Neophyte Join Date: Jun 2003 Location: Beaverton, OR, USA Posts: 10	When I originally posted, I was thinking the following: 1) Whole submersion cools all chips. I keep seeing: cool the northbridge, cool the southbridge, cool the mosfets, cool the underside...What's next? 2) No complicated water blocks or tubing. Solve one complicated problem once. Of course, now that I've been reading for awhile, it seems lack of direct impingement will significantly decrease the systems cooling ability on the chips that matter. Maybe the containing box needs to be designed like one big waterblock. Maybe the liquid enters through special jets over the cpu? The motherboard itself provides the turbulence. 3) I'm not concerned about the cost of flourinert at the moment - just its efficacy. How come there in't more enthusiam for it? 4) The first problem is creating a sealed box where I can bring out the connectors and stay sealed. Eventually, I see a neat looking clear acrylic box with cooling equipment, power supply, and disk drives mounted below in a sleek black box. Larry Lewis

07-08-2003, 12:26 PM	#11
Dr. Strangelove Cooling Neophyte Join Date: Oct 2002 Location: Denmark Posts: 14	Hmm check out this thread over @ 2cpu.com it's been going on for quite some time discussing the best way to submerge your system in oil for cooling. Has to be said that the primary objective is silence, but I'm sure that many of the tings can be used for any type of submerging. http://forums.2cpu.com/showthread.php?s=&threadid=32859

07-09-2003, 03:02 AM	#12
MadDogMe Thermophile Join Date: Aug 2002 Location: Just shut up ;) ... Posts: 1,068	I thought remembered reading somewhere of an 'equivilent' product (UK made?) that sells for much less, £50~75?... Can't remember where I read it though :shrug: , I thought it was that site ...

07-09-2003, 09:44 AM	#13
airspirit Been /.'d... have you? Join Date: Jul 2002 Location: Moscow, ID Posts: 1,986	What you would want to do is take a medium flow rate pump and create a manifold so you can direct coolant flow over the heatsink on the CPU, NB, GPU, and anything else you feel will get extra warm. You will want the absolute largest heatsinks you can find for all of those chips as well. You will need an external heat exchanger hooked to a cooling loop to keep the flourinert cool (you do NOT want to just dunk a phase change coil into it). Yeah, you can do this, but it isn't easy to do it right. You must keep the oil chamber sealed perfectly from humidity, or you will end up with water contaminating your oil. You must also find a way to route your cables in and out of that perfectly sealed chamber. It is easy to do this for small periods of time, but extremely difficult to do a decent job with this for the long haul. One method would be to keep the CD/HDD drives in the chamber as well, but keep a pocket of air in the top of the chamber for these drives to stay in. Off of the jet manifold, add a "water block" to the HDD to keep it reasonably cooled, and the passive cooling of the air over the oil will hopefully keep that cool enough to keep your CD drive cool (you will need to seal this container on the inside to prevent ANY air leaks). As far as the other connections go, you will need to install extension cords for the USB (add in a hub up top), PS/2 (does anyone use these?), VGA, and sound to points in the air gap and seal them to the outside. This will be a pain in the ass. Good luck. __________________ #!/bin/sh {who;} {last;} {pause;} {grep;} {touch;} {unzip;} mount /dev/girl -t {wet;} {fsck;} {fsck;} {fsck;} {fsck;} echo yes yes yes {yes;} umount {/dev/girl;zip;} rm -rf {wet.spot;} {sleep;} finger: permission denied Last edited by airspirit; 07-09-2003 at 09:49 AM.

07-09-2003, 10:08 AM	#14
lel4866 Cooling Neophyte Join Date: Jun 2003 Location: Beaverton, OR, USA Posts: 10	Now that I've read various comments, I agree that you probably have to put heatsinks on the warmer parts anyways, or equivalently complex manifolds for direct impingement. That negates the benefits of the whole idea. Might as well just put good waterblocks on the hot logic parts and good heatsinks on the warm parts or things like the mosfets which perform pretty good hot, anyways. Thanks for all the inputs, everyone. I, for one, am satisfied to look elsewhere for easier gains. Now, on to the backside cooling waterblock! Larry Lewis

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)