Silicon fluid immersion?

Kreiger5

Weaksauce
Joined
Nov 5, 2005
Messages
70
Mark305TBI said:
I love this thread :) .

Seconded.

I've always wondered if this was possible, and if so, how.

Now i know, and not only that, but how much, what liquids, weights, breather solutions, pump solutions, and i now know that a r/c motor prop would look dumb inside a case
:D

bump this thread for win.

One question. Does the computer HAVE to be in a computer case? Couldn't you put it into a fish tank of some sort (perhaps one designed for colder fish to handle the temps)

You could even put fake plastic fishies into it (tho they might interfere with the various moving parts)

Ideas, gentlemen.. ideas.
 

superkdogg

Gawd
Joined
Oct 14, 2004
Messages
743
This thread deserves a bump on sheer slick nerdy idea alone. the fact that it is also effective and well thought-out is a raging bonus.
 

wetware_interface

Weaksauce
Joined
Sep 30, 2005
Messages
73
this has some great info for a slightly different idea im hashing out to chill cool the fluid and leave it cooled / maybe even in frozen state around the critical components.

however for those wanting to prevent oxidation of the fluid...

just do a sealed system with air valves at top and one just above mobo level. open top air valve and fill chasis with helium. its completely inert and won't oxidize anything nor spark or cause combustion. you could even fill with fluid first and then use compressed helium at lower valve and let it percolate through fluid (quite forcefully :D ) and it's higher pressure would evacuate room air out of the case. shut off the top valve and immediately after stop helium transfer. you wouldn't need to worry about effects of pressurized internal air as it'd be mostly inert helium.
 

best [486]

Limp Gawd
Joined
Aug 28, 2005
Messages
476
two words: cooking oil

non-conductive and cheap
and the color doesn't matter 'cause it's cheap
 

Zxcs

[H]ard|Gawd
Joined
Sep 11, 2004
Messages
2,007
best [486] said:
two words: cooking oil

non-conductive and cheap
and the color doesn't matter 'cause it's cheap
Already been brought up many times, however it does not have the heat capacity of other transformer liquids such as luminol-TR1 or midel 7131. It also smells and is pretty viscous.
 

mwin

2[H]4U
Joined
Jun 24, 2004
Messages
2,908
silverphoenix said:
Question can't you guys just use something like Fluid XP?
I emailed them and asked what they thought about using it in an immersion environment. They didn't recommend it. They didn't give a reason, though.
 

mwin

2[H]4U
Joined
Jun 24, 2004
Messages
2,908
zer0signal667 said:
Damn, this thread just won't die... :p

Did you ever find something and use it?
It's only 13 months old ;) . It ranks 5th in views in the Extreme Cooling forum, my claim to [H]ard|Forum fame...

I think I had pretty much settled on that silicone fluid, but I never actually ordered it. Bought a house and moved out on my own, so I've been too distracted. I've started pondering the idea, again...but we'll see. Building a good enough (clear) enclosure is probably the biggest hurdle.
 
Joined
Mar 15, 2006
Messages
2
Hey guys, great forum topic! :D

I'm currently interested in building a Gaming Oil PC for the living room. Silence is my highest priority, but only so long as the computer can continue running nothing above the standard expected running temperature. However, the lower the temperature the better, but again, only so long as it remains completely silent. So far from I've been able to gather from this forum, luminol-TR1 is probably the most cost-effective oil solution. It's clear, and doesn't fester like Vegetable Oil which produces an annoying smell.

I'll place the computer in a cheap plexiglass casing, or fishtank, seal it off, and using valves, fill the chasis with helium so components don't oxydize. To compensate for case expansion due to heat, I'm going to place a sealed rubber "breather tube" at the top of the case. I will be placing heatsinks on the processor, and over the graphics card chip. I'm not going to use any fans, because from what I understand they burn out because they can't stand the resistance from trying to spin in the dense oil.

Because the power supply also has a fan, I'm going to remove the fan prior to including it inside the tank with the motherboard.

The gaming computer will have quality components such as those used in the example given on the tomshardware website:

Rancid Veggie Oil Computer

The components I'll be using will be:

AMD FX-55 chip
Geforce 7800 Ultra PCIe
Creative Labs X-Fi sound card.

Mechanical components such as CD-ROM, hard drives, and other such disks will, of course, not be submerged.

This system will be running in a room with temperatures of up to 90 degrees Fahrenheit.

Here's a key issue however I've yet to find anyone not yet properly addressing. The think tank example uses a submersible pump to circulate the oil in the case. However, a submersible pump it appears to me is apt to make noise, so I don't wish to include it. Does anyone know if this would be dangerous, because of there being a potential for key components to overheat due lack of oil circulation?

I think I've dealt with the water condensation problem by sealing off the tank. However, if there's anything else I'm missing which may potentially damage the system, such as water condensation, please let me know.

Thanks.
 

carden

n00b
Joined
Jan 19, 2006
Messages
43
On a side note, could you use an acrylic clear case, then use window tinting on it for some fun cases? or would the oils/fluids eat it?
 

gesicht

Gawd
Joined
Jun 25, 2004
Messages
861
i think you would need more then just circulation from a pump.

i would think you would still need a pump with hoses to push fluid across things like cpus and gpus. need large fined and spaced heatsinks with lots of room for flow.

if you are going to go all out. i dont think you will get worthwhile cooling without directed flow across heatsinks
 

Ozone77

Limp Gawd
Joined
Oct 14, 2004
Messages
326
Anyone knows what liquid KOOLANCE is using on their "submerged PSU"?

The recent PSU review here at [H] opened the PSU and did not mention any modifications to the capacitors or other components so I am guessing it its "plastic/rubber-friendly"
 

Arcygenical

Fully [H]
Joined
Jun 10, 2005
Messages
25,057
Anyone knows what liquid KOOLANCE is using on their "submerged PSU"?

The recent PSU review here at [H] opened the PSU and did not mention any modifications to the capacitors or other components so I am guessing it its "plastic/rubber-friendly"

I would bet just regular low viscosity mineral oil.
 
Joined
May 23, 2007
Messages
8
Ok Guys,

I have been directed to this thead because i too am interested in immersion cooling for a pc, after reading a few of the posts i thought i would chuck in a few ideas to some of the problems you are thinking about.

For the problem of thermal exspansion when using a sealed system to prevent oxidisation of the fluid being used why not use a pressure compensator in the circuit, they are used all the time in my line of work on rov's to prevent electrical housings being crushed by the sea pressure and also for thermal exspansion, should cure your problems.
link for an example select the pressre compensators from left side on list http://www.ifokus.no/index.cfm?tmpl=butikk&a=list&b_kid=94108&toppid=-1

In regards to moving the fluid around for better cooling why not use a 12v bilge pump, they are very robust, and have to be compatible with a varity of oils,water etc.
link for an example http://www.whalepumps.com/marine/product_list/8/133/

Hope this was useful

Del
 

Arcygenical

Fully [H]
Joined
Jun 10, 2005
Messages
25,057
Yes, a bilge pump would be a good idea...

But most commertially available mineral oils have horrible heat transfer characteristics...
 

synaps3

Gawd
Joined
Mar 15, 2006
Messages
957
Another bump for a rather interesting thread...

Has anyone ever gotten the balls to try this with decent hardware?

Where can I get Midel or Luminol-tr? Building a system around this concept sounds like fun :D
 

synaps3

Gawd
Joined
Mar 15, 2006
Messages
957
...eww...

I'd rather not have my computer smell like a McDonald's dumpster. The idea of using cooking oil is bad anyways, because it is so dense. It would be difficult to get it to flow properly...
 

Alien42

n00b
Joined
Sep 16, 2007
Messages
5
Question can't you guys just use something like Fluid XP?
I emailed them and asked what they thought about using it in an immersion environment. They didn't recommend it. They didn't give a reason, though.
I can give you a reason:
http://www.youtube.com/watch?v=bOWLv2YHE_c

Damn, this thread just won't die... :p
necro-posting FTW! :D
 

davea0511

n00b
Joined
Aug 23, 2008
Messages
2
Ancient thread, I know ... but it's the best treatment on the subject of PC-immersion fluids that I've been able to find on the whole internet

... anyway I thought I'd add my input that the best pump for a thick liquid like some people are using is a boundary layer pump ... (a tesla turbine in reverse). Easy to make, especially if you have an old useless hard drive containing a few platters. There's lots of tutorials on the net on how to make them.

Though my personal opinion is that a pump should be entirely unnecessary if you utilize convective heat flow.
 

davea0511

n00b
Joined
Aug 23, 2008
Messages
2
So some of these people started their PC's some time ago. Would be nice to know at this point what the long term results were of having their entire PC submersed in a dialectric fluid. Did any of them fall apart ... and which fluids were the most inert?
 

Traze

Limp Gawd
Joined
Dec 8, 2006
Messages
281
So some of these people started their PC's some time ago. Would be nice to know at this point what the long term results were of having their entire PC submersed in a dialectric fluid. Did any of them fall apart ... and which fluids were the most inert?

Ya, anyone actually do these?
 

BrainEater

Gawd
Joined
Jul 21, 2004
Messages
1,017
My machine was submerged in midel7131 for 4 months.It worked perfectly.

All the components still work too....After cleaning in boiling IPA about 4 times , I use it for my router now.No problems whatsoever.....

The Puget Systems rig has been running for more than a year , it works fine too.Check thier website for more data....

-----

There are issues to be dealt with as far as submersion cooling goes , but many people are working on this , myself included.I'm quite certain this field will be expanding.

:D
 

Heather Taylor

Velocity Micro Customer Experience Specialist
Joined
Oct 6, 2007
Messages
620
Yep, ours has been running for over a year, and we're debuting V2 @ PAX this weekend.
 

flogge

[H]ard|Gawd
Joined
May 20, 2008
Messages
1,888
So has anyone tried using bubbled air or another gas to agitate their fluid rather than a submerged pump? I know there was mention of charging a system with helium or nitrogen. Why not mount your motherboard at an angle so that the bubbles come up against the face of the motherboard and video card and then have to move up the motherboard to get back to the top of the fluid.

Some of the fluids and low temps you guys are using could produce some large, slow moving bubbles but it might be worth a try. This does require the use of a gas compressor or blower of some sort,
 

Heather Taylor

Velocity Micro Customer Experience Specialist
Joined
Oct 6, 2007
Messages
620
The guys had a lot of fun with v2...if you have any questions about it, feel free to send an email and I'll forward it over to the production team.
 

OhMyGod

Gawd
Joined
Aug 24, 2006
Messages
658
Why not just use demineralized water? You would need to clean everything first to remove any contaminates and maybe a small resin (cheap) tube with a bypass to occasionally reclean the water as you pump it around.

You would also want to seal it to greatly slow down the conversion to carbonic acid.

But then you got the best liquid to use.


*edit*
you could even put a TDS (cheap) or conductivity meter in the demin water to verify it is staying clean.
 

serpretetsky

[H]ard|Gawd
Joined
Dec 24, 2008
Messages
1,783
Why not just use demineralized water? You would need to clean everything first to remove any contaminates and maybe a small resin (cheap) tube with a bypass to occasionally reclean the water as you pump it around.

You would also want to seal it to greatly slow down the conversion to carbonic acid.

But then you got the best liquid to use.


*edit*
you could even put a TDS (cheap) or conductivity meter in the demin water to verify it is staying clean.

chance of catastrophic failure is higher. Also, i remember some site tried this (tomshardware i think?) and their computer failed after a minute or two. I don't remember if they decided that ions had formed during the computer boot and created conductive areas around the board or they decided that water's capacitance had made very close electric traces affect each other.
 

OhMyGod

Gawd
Joined
Aug 24, 2006
Messages
658
chance of catastrophic failure is higher. Also, i remember some site tried this (tomshardware i think?) and they're computer failed after a minute or two. I don't remember if they decided that ions had formed during the computer boot and created conductive areas around the board or they decided that water's capacitance had made very close electric traces affect each other.


Well, in theory it will work. I imagine the problem would be with the initial cleaning. But even then, if you had a constant demineralation (spelling?) bed cleaning the water all the time I would bet it would work.

Hey, if I managed to do this... would it make me rich and famous? or at least get all the girls wanting me?
 

mwin

2[H]4U
Joined
Jun 24, 2004
Messages
2,908
I've never seen anyone ever try engine oil. Is there a reason for this?

It's probably awfully rough on the plastics and rubbers in the motherboards, cables, and everything else. Give it a shot and let us know how it works for you... :p
 

spadefoot

Limp Gawd
Joined
Oct 13, 2000
Messages
424

OniExpress

Limp Gawd
Joined
Feb 11, 2009
Messages
179
You could indeed do it that way, but it would be significantly harder in a few ways.

1. "Airtight" is harder than "Watertight" (or in this case "Oiltight" which is technically harder than watertight).
2. Absolutely no ease for doing maintenance/upgrades/tweaks. There's no remotely practical way to open the thing up without losing all of your gas. (ba-dum CHING)
3. Much poorer thermal conductivity, so whole-board cooling is pretty much out.

If you sealed the board in a container of inert gas, you could run your subzero cooling to the CPU without fear of frost (or at least frozen water) assuming you have a near-pure atmosphere inside. It's possible, but only in the sense of difficult engineering designs. Also you'd likely bleed gas at even a small rate constantly, and then there's the fact that you'd lose it all eachtime you had to access the motherboard (this mainly seems like an issue with initial setup and would require you to flush the normal gasses out several times).
 
Top