View Full Version : Mineral Oil Cooling
SO there I was, reading the Comdex article on http://www.tech-report.com when I saw this:
http://tech-report.com/etc/2001q4/comdex/koolance2.jpg
Not much info on it at all, except that it was running and it was filled with mineral oil. Now, am I crazy, or does this seem like an extremely easy thing to do? A watertight plexi box wouldn't be too hard to build...but how do they cool the oil? That part might be difficult, maybe a bunch of peltiers working as chillers?
Here's hoping Koolance releases this soon:D
fuzzba11
11-20-01, 01:25 AM
Awesome!!! Even the power supply in there! hmmm...
Cooling with peltiers doesn't work, and it would suck too much power, so that can't be it. Maybe it never needs to be cooled!! :p
BrianCapps
11-20-01, 01:35 AM
Pump the oil through a radiator of some kind. Or immerse the radiator in liquid nitrogen.
Co0l :D
r0ckstarbob
11-20-01, 02:59 AM
hm
awesome thanks!!! i was looking for some pictures/info of the new case they debut'd
they're also coming out with a new full tower case like the one they have now....i think withing a few weeks they're selling plexiglass windows too (ONLY 10 BUCKS...WHY THE HECK NOT!!!:D )
I was thinking of making a box for the mobo to go in. from there, have tubes that go off to a radiator. Have the mineral oil radiator submerged in alcohol, which is also run through a radiator that is air cooled. But that's just my thought.
I don't think this is intended for production; can you imagine the problems if you tipped it over? You'd be cleaning up for days! The other, regular Koolance case was looking pretty good tho.
fuzz- if they can afford all that oil and stuff, they can probably afford to use a dozen high wattage pelts running off a bench PSU :)...or perhaps they are just using a modified version of the regular Koolance cooler so that it works on oil. I bet they have the heat exchanger or whatever under the case, in the white box part...actually I bet that's it. Pity Koolance won't answer my emails.
How much money would all that oil cost? I have a feeling it's not regular baby oil...
I was thinking of mineral oil for a while now...
Do you really have to cool the oil?
I mean there is sooo much surface area for it
too escape! Well, pole pigs cool like that
so why not a computer?
Well this is one of those things I will be trying
sometime in my lifetime!
I must dig up more info!
VashTheStampede
11-20-01, 12:53 PM
Yes, some form of major cooling would be needed, the oil by itself is barely enough to keep a PII 233 MMX at the same temps seen by it's stock HSF. Not only that, mineral oil is expensive, $30 a bucket, that only filled this guy's 10 gallon aquarium 1/3 of the way:
http://www.tracelength.com/01-11-18mineral.html
Someone else also made this a point:
"Another nasty effect is that water is heavier, so any condensation sinks to the bottom and pools on the mobo."
~RT~
"Another nasty effect is that water is heavier, so any condensation sinks to the bottom and pools on the mobo."
Why would condensation form unless the watter is cooled
below the temp of the surrounding?
Even though - I would prop my mobo on it side.
You could maybe have a drain hole so when the watter
(condensation) came about you could just "drain" it.
- You know like those valves on coolers.
What do they mean by the viscosity of the mineral oil?
Witch viscosity is best?
SavageHenry
11-20-01, 01:49 PM
Originally posted by YMAN
What do they mean by the viscosity of the mineral oil?
Witch viscosity is best?
Viscosity is a measure of how 'thick' a liquid is. The more viscous, the harder it is to pump. The lower a liquid's viscosity, the more the liquid tends to flow when pumped.
Another nasty effect is that water is heavier, so any condensation sinks to the bottom and pools on the mobo."
hmmm, I think that's backwards, I'm pretty sure oil is heavier...but either way, how is condensation going to form in the middle of a mass of mineral oil? The water will have to come from somewhere, it won't just appear there. The oil should be a barrier to water...
$30 a bucket though, too bad. Even a small case will need lots more than that...maybe goto a school chem lab, ask the teach? There's not too much surface area exposed, the oil will need some form of cooler.. thanks for the linkage!
Ok, idea time!- run a regular water-cooling system, with peltierised water as the coolant, then submerge that system in the chilled oil...frozen! Except, it can't freeze. Seal the whole thing airtight, fill to 100% with oil, and not even a slight risk of condensation. Should be able to run it as cold as you want, no problems.
Intraveinous
11-20-01, 03:24 PM
Originally posted by Monster of Rock
Ok, idea time!- run a regular water-cooling system, with peltierised water as the coolant, then submerge that system in the chilled oil...frozen! Except, it can't freeze. Seal the whole thing airtight, fill to 100% with oil, and not even a slight risk of condensation. Should be able to run it as cold as you want, no problems.
Now there's an idea... Same cooling methods we've seen before (pelted water cooling, phase change, whatever) but with zero risk of condensation... The oil provides only the barrier against condensation. No worries about finding a pump to move the oil, etc... Also, the oil would be cooled by the pelt/phase change setup, and would therefore cool your mobo, vidcard, etc...
hmmmmmm...
Now I'm thinking... :D
Peace
John
Should be able to run it as cold as you want, no problems.
But what about the condensation outside the case?
You want to be able too see the beautiful components,
dont you?
But what about the condensation outside the case?
lol good point man! well I guess we bust out the Rain-X or get the squeegee....:D
Intraveinous
11-21-01, 10:51 AM
Would the oil NEED to be chilled??? if not, you don't have any problems with condensation on the outside either... with a decent volume of oil, wouldn't convection take care of cooling it on it's own? Also, from back in the EARLY days of watercooling, where people had giant heatsinks with fans dipped into their resivoirs, you could do something similar...
(I'm at work so a crappy .bmp will have to do, but it's less than 100K)
Convection would take warmer oil to the top where the heat would be absorbed by the large heatsink(s) and cooled by fans (L1A 120mms or something quiet) Regular pelted watercooling keeps the chip cool, and runs to a radiator and resivoir outside the aquarium or other sealed container...
Would it work???
Peace
John
The Overclocker
11-21-01, 11:29 AM
mmm i wonder how well the hard drives and stuff would perform in the the oil?
Originally posted by the overclocker
mmm i wonder how well the hard drives and stuff would perform in the the oil?
If you were going to submerge the hdd drives you will
need to seal all the vent holes, it will not work in fluid
because it uses air to ride over the disks. I also doubt the
magnetic fields the HDD drive uses would be powerfull
enough to break through the oil. http://www.ytec.d2g.com/emoticons/notepad.gif
yeah, hard drives look like they are sealed air-tight, but they aren't- you'd need to keep them out of the oil.
I suppose you don't actually need to cool the oil, but it will pick up plenty of ambient heat- the CPU isn't the only heat-producer in a PC. Eventually it would get pretty hot without any cooling. The oil in this situation replaces the ambient air around a normal system- hot oil would be like having your PC in an oven.
That picture you made seems to make sense to me- only way to find out if it works ok is to try it!:) Although HSFs working in reverse like that have proven to be pretty inefficient.
ButcherUK
11-21-01, 11:52 AM
I did a few checks on oil, it;s definitely less dense than water so any water that gets into it would sink to the bottom, don't believe me? get a cup of oil from your kitchen and pour some water in.
I did a quick check on oil viscosity and pricing, looks like you can have some with a viscosity of about 5 (that's dynamic viscosity 5 at 40C, about 8 times thicker than water) for about 10 bucks / gal. There are thinner precision instrument oils but these cost more, more like about 40 / gal. Also oil royally sucks for heat capacity compared and water of course. Still a lot cheaper than flourinert :)
I have comprised a theory on how to cool
the fluid using wattercooling.
The wattercooling that cools the fluid would be cooled
by standard means of wattercooling.
Ok here is how it works,
ALL OF THE STUFF IN THIS DIAGRAM IS SUBMERGED IN THE FLUID!
http://www.ytec.d2g.com/WEB~PAGE/SUB/A.jpg
Green = Fluid (Mineral Oil etc.)
Blue = Watter from wattercooling system to fluid cooling system
Red = Watter from fluid cooling system to watter cooling system
Brown = Movement of fluid
Purple = Fluid in a pipe/tube
Yellow = Pump
Dark Grey = Box that diverts fluid downward through the radiator
Light Grey = Submerged radiator - below the box
Summary:
The fluid of choice, would get pumped up by the pump
to a small box above a radiator, the box would channel the
fluid through the radiator were the radiator will extract the
heat from the fluid and the transfer it to a wattercooling
system. The the cooled watter will exit the bottom of the
radiator back into the system.
it's not a big deal, I believe you- so we raise the motherboard 1 inch off the bottom of the oil-filled case, and install a drain valve in one corner at the lowest part to drain off water. No more water problem. It's sealed airtight anyway, so no water can get in anyhow.
Condensation is still the least of our worries-I still wonder about a spill or a leak...what about cleaning components after they've been in the oil? Will long-term oil submersion do anything to the components?
It seems like the big bonus from this oil-cooling method would be to slightly chill the oil to about 10C or so, and run water thru the CPU cooling system. The cold oil will be like running your system outside on a cold day- very low ambient temps to start with. It should make all components more tolerant of a stressful overclock, as all components are cooled.
ButcherUK
11-21-01, 12:09 PM
Oil is known to do bad things to capacitors. Namely the sealing gaskets swell and can cause the case to rupture. If a capacitor ruptures it's completely toast, your board is likely to just stop dead then and there possibly taking out half the rest of the comp too.
Originally posted by ButcherUK
Oil is known to do bad things to capacitors. Namely the sealing gaskets swell and can cause the case to rupture. If a capacitor ruptures it's completely toast, your board is likely to just stop dead then and there possibly taking out half the rest of the comp too.
Then spraypaint the capacitors with some type of sealant.
- not a big deal
^^^ Dont forget to give your words on my theory! ^^^
ButcherUK
11-21-01, 12:23 PM
Originally posted by YMAN
Then spraypaint the capacitors with some type of sealant.
- not a big deal
hmm lets see there are only a hundred or so to coat, shouldn't take any time ;)
I suppose you could mask off the areas you wanted direct cooling onto and then just spray the rest, YMMV though.
Originally posted by ButcherUK
hmm lets see there are only a hundred or so to coat, shouldn't take any time ;)
I suppose you could mask off the areas you wanted direct cooling onto and then just spray the rest, YMMV though.
Yes, just mask around everything elts, and give those capacitors
a good 5 coats!
ButcherUK
11-21-01, 12:34 PM
Originally posted by YMAN
Yes, just mask around everything elts, and give those capacitors
a good 5 coats!
It'd have to be done carefully - the gasket is on the underside of the cap where the leads exit, this is in contact with the board so effectively you'd have to seal the caps to the board.
Originally posted by ButcherUK
It'd have to be done carefully - the gasket is on the underside of the cap where the leads exit, this is in contact with the board so effectively you'd have to seal the caps to the board.
Yes thats true, another thought...Be carefull when you pull
the tape off, pull it off super slow & use the tape with the
weekest glue, just to prevent accidentaly pulling off the traces.
nobody posted about my theory. http://www.ytec.d2g.com/emoticons/cry2.gif
ButcherUK
11-21-01, 12:43 PM
If you use masking tape trace pulling shouldn't be an issue they're not particularly easy to pull off as they're coated already. Also your theory seems like it might work well... You'd want both the pump and rad towards the top though - hotter oil would rise so then you suck it up and push it through the rad cooling it and causing it to fall. If you set it up carefully you might even be able to get away without a pump, just convection.
Intraveinous
11-21-01, 12:45 PM
OK, so how well is a typical Eheim or other popular pump used in water cooling going to handle pumping oil since it's so much more dense and viscous? Would a car oil pump work or would it be too loud? Obviously, the point isn't just good cooling, but good quiet cooling as far as I'm concerned... Also, what would be a good way to cool the oil down to say 10C? Peltiers are costly, both in purchase price and running price, plus you'd need a bench supply and something to gid rid of the heat they produced, adding complication, cost, and noise. It would be possible to pump it through a refridgerator or something wouldnt it? though you'll still have the problem of the thermal load becoming too much for the fridge compressor. How about an Oil bong??? :D It wouldn't evaporate, but you probably wouldn't acheive the 10C you're shooting for. :eh?: Also, as far as spraypainting capacitors, would you use conformal coating like you do for (supposed, I've never tried it) condensation protection? I've got a lil 5 gallon aquarium somewhere out in the garage, and an old mobo... This sounds like fun, but I wanna make sure I don't kill anything unneccesarily.
Peace
John
Originally posted by ButcherUK
If you use masking tape trace pulling shouldn't be an issue they're not particularly easy to pull off as they're coated already. Also your theory seems like it might work well... You'd want both the pump and rad towards the top though - hotter oil would rise so then you suck it up and push it through the rad cooling it and causing it to fall. If you set it up carefully you might even be able to get away without a pump, just convection.
They are coated but I would not take any chances.
You are right, you dont need a pump if you set it up correctly
but I figure the pump would cause it to be more efficient.
ButcherUK
11-21-01, 12:51 PM
If you're oil is thin enough it should work fine with an eheim or similar, assuming it doesn't attack the seals (unlikely to), probably be best to get comthing high GPH like an ehiem 1060.
As for cooling the oil 10C, you might try pelts but that probably wouldn't work too well, nor would a bong. A phase change setup is a possibility, that'd be quite noisy and expensive though.
Just had a thought for a setup that might work well... You could have the pump take oil off the top (where it;d be hotter) and pump through an external radiator (I'd rec. a tranny oil cooler) then back into the bottom of the case. This would elimitate needing a separate water loop. Also you could try putting a peltier chiller assembly on or something but that'd raise noise a lot.
Originally posted by Intraveinous
OK, so how well is a typical Eheim or other popular pump used in water cooling going to handle pumping oil since it's so much more dense and viscous? Would a car oil pump work or would it be too loud? Obviously, the point isn't just good cooling, but good quiet cooling as far as I'm concerned... Also, what would be a good way to cool the oil down to say 10C? Peltiers are costly, both in purchase price and running price, plus you'd need a bench supply and something to gid rid of the heat they produced, adding complication, cost, and noise. It would be possible to pump it through a refridgerator or something wouldnt it? though you'll still have the problem of the thermal load becoming too much for the fridge compressor. How about an Oil bong??? :D It wouldn't evaporate, but you probably wouldn't acheive the 10C you're shooting for. :eh?: Also, as far as spraypainting capacitors, would you use conformal coating like you do for (supposed, I've never tried it) condensation protection? I've got a lil 5 gallon aquarium somewhere out in the garage, and an old mobo... This sounds like fun, but I wanna make sure I don't kill anything unneccesarily.
Peace
John
What about my idea with bigger radiators and use a bong!
Id say you could get if fairly cool.
ButcherUK
11-21-01, 12:52 PM
The problem with a separate water loop is that extra loops add more inefficiencies, having a single oil loop would be preferable.
Originally posted by ButcherUK
The problem with a separate water loop is that extra loops add more inefficiencies, having a single oil loop would be preferable.
Yes but with my sytem you are capable of using a bong witch
also enables you to have cooler temps in a simple way.
ButcherUK
11-21-01, 01:07 PM
Originally posted by YMAN
Yes but with my sytem you are capable of using a bong witch
also enables you to have cooler temps in a simple way.
Assuming losses for having the extra loop aren't greater than the gains of using a bong.
These are the words of overclockers.com member Scottland.
- He is having probloms posting, so I am posting for him...
-------------------------------------------------------------
HEY!!! think about this
say you get an aquarium right?
fill it full of mineral oil and make it air tight
make sure its got a good seal
and you have removed ALL THE air out of the auquarium
now, you would have to somehow make a whole out of the top
for your cables, and caulk aroung them for a air tight seal
now buy a standalone freezer and SHOVE the aquarium in there!!!!!!!!!!!!
no condensation would form because there is no air inside the aquarium
and mineral oil has an extremely low freezing point.
now you have a EXTREMELY cold computer that wasn't all hard to do
Also i don't know if it does but..... i wonder if a aquarium that is open would have condensation in the freezer? cause think about it.....
Water is Polar.
Oil(and all other lipids) are NON-Polar
Therefore (as you all know) Oil is NOT soluable in water
So the water would not sink to the bottom of the tank
because the condensation could never enter the tank because of the oil
Either way... a smal standalone freeze only runs like 50 $ on sale
+ an aquarium (10 $) and the oil say (30$)
you have a case/ extreme cooler for like 90 $
which is all a good case costs.
one more thing
i would mount a heatsink like a mc462A on there w/o the fan
just to take the heat off of the processor
and let the oil do the rest
and if you really wanted to get ambious you could create a resivor that sat outside the tank inside the freezer and pumped like acholol onto a waterblock .... this might work too
Just some ideas for you guys to work with
~Forgotten~
Have you guys thought about "corn oil"?
Its super cheap!
Yes, it sounds corny but thats what it is! http://www.ytec.d2g.com/emoticons/lol.gif
Intraveinous
11-21-01, 01:29 PM
You'll still have the problem of the thermal load overloading the freezer compressor. Keeping the oil cool seems to be the biggest problem. The pump from the top to outside rad to bottom seems to be the best idea I've seen, the problem there is cutting a hole in the glass at the bottom of the aquarium. If you used lexan it'd be far easier, but you've got the problem of making the lexan box completely liquid tight. I could be done (no different than an aquarium except it's lexan rather than glass) but how many of us are experienced aquarium builders?
This is a great discussion, keep the ideas coming and I'll see what I can throw together...
Peace
John
Originally posted by YMAN
But what about the condensation outside the case?
You want to be able too see the beautiful components,
don't you?
YMAN U have to seal the case anyway to preserve more cold inside or U will be losing so much power for heating and cooling ur room man U have to think more U put a camera if U vana peek inside and isolate the case that's most reasonable but the oil is cool cause some oils can take as low as -40 degrees Celsius without freezing so .... U c if U pump it right U can have wherry oc system rock solid and no extra cooling though U have to cool the radiators :) the idea is nice and 90 bucks 4 oil which u don't have 2 exchange (if U don't burn it ) is nuthing the oil is for lifetime and U can build all ur systems there its breakthrough idea man this is kewl im exited
u cant use other oils cos they might not be resistant for eleticyti and u cant cool it 2 down cos it will let the electricyty through or as much oi fiure im not a fuscian educated correct me if im wrong dont forget fusican laws and that the one oil can was about 30 $ but u have 2 boy @ least 3-4 to go with the pump and radiator and to get ur board fully cowerd
ok, I have more ideas too- the main heat producers in a PC are the CPU, PSU, and GPU, right? Cool those with standard water cooling, and run the hoses so that the rad is outside the aquarium by a good margin. Ok, now the really hot components are cooled separate from the oil itself.
Now, with the major heat sources already dealt with, the oil will only absorb heat from secondary components such as modem, Northbridge, etc, as well as a bit of leeched heat from the CPU and others cooled by water. A very strong but SLOW pump can be used to move the oil thru a large passive radiator on the side of the case. I believe that passive radiation will be more than enough to cool the oil if the primary heat generators are already dealt with.
If the system is left running foir a very long time at full load, the oil may get too warm- add a few 120mm fans at 7V or so to move some air over it.
I do not think that the pumps we are used to using for watercooling will work for oil- but oil pumps do exist, every car has one in it!
We need to know the long-term effects of oil on a PCB. As well as which type of oil is both readily availible and cheap enough, as well as has good (decent) heat transfer properties.
would a peltier submerged in oil work correctly? would the high energies of the pelt thru it's open sides change the oil somehow? I don't know, but if so I have some ideas for cooling the oil.
ButcherUK
11-21-01, 04:09 PM
Originally posted by Intraveinous
You'll still have the problem of the thermal load overloading the freezer compressor. Keeping the oil cool seems to be the biggest problem. The pump from the top to outside rad to bottom seems to be the best idea I've seen, the problem there is cutting a hole in the glass at the bottom of the aquarium. If you used lexan it'd be far easier, but you've got the problem of making the lexan box completely liquid tight. I could be done (no different than an aquarium except it's lexan rather than glass) but how many of us are experienced aquarium builders?
This is a great discussion, keep the ideas coming and I'll see what I can throw together...
Peace
John
Yes I thought about that when I posted and though I didn't mention it I was thinking you could have both tubes running out through the lid - the inflow pipe would just run all the way to the bottom of the tank.
robertm
11-21-01, 04:45 PM
Dipped caps don't need to be sealed but the Can caps like the ones used around the socket for the on board PS would need some sealing.
Just don't let the oil get to hot it smells bad :) (I know I have burned it once)
Mr_Goat
11-22-01, 03:20 AM
This is a very interesting topic and here are my ideas:
The tank can be made very cheaply either out of Lexan or Lucite.
Lucite is a trade name for a special type of acryilic plastic. It is transparent, much cheaper then Lexan and tough.
Sealing the tank is not really a problem if you use the correct adheisve. For Lexan you would use a polycarbonate adheisve and for Lucite, a acrylic adheisve.
Both adheisves are inexpensive and I sure that if you call 3M and tell them that you'd like a sample they would send you more then you would ever need.
These adheisves work by bonding the plastics on the atomic level, what they do is melt the plastic where it comes into contact. So all you would have to do is coat the two sides that you want to bond and stick them together with light pressure for 10sec. These adheisves are not messy and have a consistency equal to water.
Secondly, some guys have been tossing the idea of using a car's oil pump to pump the fluid. This rather difficult because the pump is mechanically driven and you would need to spin it at around 200+ rpm to get any flow. If you want 0.5-1gph the you need 2000-3000rpm. You can look for electric oil pumps but those are usually high perforamance aftermarket pumps which equals $$$.
of this oil ? and how is about di electri... damn what is it called im not english speaking (dielectricitie?)
adamtekh
11-24-01, 07:35 AM
coolanol :burn: ;)
http://www.exxon.com/exxon_lubes/tigerbytes/documents/brochures/bro053.htm#flammability
adamtekh
11-24-01, 07:40 AM
im going to try the coolanol stuff on my 200 mmx weeeeee
http://www.maximausa.com/products/misc/coolanol.asp
:cool:
Intraveinous
11-24-01, 10:20 AM
Originally posted by adamtekh
im going to try the coolanol stuff on my 200 mmx weeeeee
http://www.maximausa.com/products/misc/coolanol.asp
:cool:
I don't know if that's the same stuff as the Exxon link you gave before... I'd check into it more before dunking anything. The link above to MaximaUSA is for antifreeze and deionized water pre-mixed, while the Exxon stuff is a Silicon Ester dialectric cooling compound... I'd say keep looking for a source before trying to dunk your system in the MaximaUSA stuff.
Peace
John
Mr_Goat
11-27-01, 09:20 PM
Originally posted by adamtekh
coolanol :burn: ;)
http://www.exxon.com/exxon_lubes/tigerbytes/documents/brochures/bro053.htm#flammability
Very good stuff...the only problem is that the tank has to be hermetically sealed...the this stuff absorbs moisture very quickly and that is a bad thing...
It's also a diaelectric...which means that it stores electrical charge when a voltage is applied across it. I'm not to sure of the significance...maybe someone with a background in electrical engineering should take a look at this.
adamtekh
11-27-01, 09:51 PM
dialectric means it dosent conduct electricty i do belive , and yes u would ahve to seal the case , no big problem .
its doable
I will conferm that, Diaelectric means it DOES NOT conduct
electricity.
adamtekh
11-29-01, 08:52 PM
i have liltle money to test it out right now , im saveing for a water cooling rig on my Athlon :cool:
arhines
01-08-02, 01:30 AM
Originally posted by VashTheStampede
"Another nasty effect is that water is heavier, so any condensation sinks to the bottom and pools on the mobo."
~RT~
...but of course in koolances version the motherboard is on it's side. I like the idea in general, but I think that a water-loop waterblock is still needed to cool the cpu. Oil simply doesnt transfer the heat well enough in my opinion.
I remembered reading dr. Ffreeze stuff back in 98 (wow that was a long time ago!), and being disappointed because he gained practically nothing from the experiment...he did manage to get water droplets on the motherboard itself though...but luckily a layer of oil protected it. Heres the link to the water droplet nightmare :) http://www.drffreeze.com/Nightmare.htm
BinarySuperman
01-08-02, 04:53 AM
is mineral oil dense, like goo, or does it feel like water??
It seems we are on the same thought I posted a thread earlier about the dunk tank. It doesn't look that hard to make. I'd think room temp oil some peltiers it could fly. So who's up for a little testing? My friend and I were thinking about it.
The Overclocker
01-08-02, 11:14 AM
this case is just used to show off what they can do at the mineral oil is very expensive, they bring it to everything. but i am sure if you asked them they might make one or at least give you contacts to get the stuff
Originally posted by BinarySuperman
is mineral oil dense, like goo, or does it feel like water??
I think it is thin (somewhat like watter).
--------------------------------
As for mineral oil being expensive...I am still curious about
vegetable oil. :eh?:
u can ask from electrichan for condenstaor oil itl do the trick btw its freezing starts from -45 degrees celsius.
ButcherUK
01-10-02, 02:55 PM
Originally posted by BinarySuperman
is mineral oil dense, like goo, or does it feel like water??
YES!
The thickest mineral oil is pretty much a solid grease.
Then you go to the thicker gloopy types such as motor oil and stuff.
The you get into compressor oils, they're usually somewhat thinner.
Also there is prescision intruments oil and such, they're often very thin, maybe 2-3 times as thick as water.
In general the less colour the oil has the thinner.
u cant use cornoil it doesnt have electrical resistance and i think it doesnt be as good
only - for oil is the cleaning oh and it may uncontact ur connectors flowing between em so ur cpu may be all the suden be desconected ! if u know what i mean! peace ! the minus for transistor or condensator oil is some S in it that will make H2SO4 in ur comp and that is never good!
Originally posted by YMAN
I will conferm that, Diaelectric means it DOES NOT conduct
electricity.
You are correct YMAN, a "dielectric" is any material or substance which creates a non conductive barier between conductors or to prevent electrolysis between dissimilar metals. :)
rivercom9
01-11-02, 09:38 PM
I cant imagine having to upgrade that machine. You wont beable to grab anything too well cause it so slippery. And cleaning up an accident would not be very fun.
ButcherUK
01-11-02, 09:54 PM
Using an organic oil as opposed to a mineral oil is a bad idea as they go rotten after a month or two.
im planing to do so :
1 i bild (from plastik using warmclew tec) a case i isolate it with some stuff wich holds the room temp away from my bilt case.
2 i get a freaon freezer (irs usually in freezers even in home ones) btw freaon is gas wich can get the temp as slow as -30-40 degrees cesius.
3 then i need oustside radiator for freaoon coling.
4 i get pump its sed floor heater pumps are good they pump oil originaly why not my oil
5 i make special radiators (biger then they are) for cpu for mem and graphix
6 the coolest oil must be injected instantly into radiators mobos other stuff is sevondary when my cpu and agp is kooled -45 i think np wikl acure oc ing for that oi need special channels seperated from general cooled oil (if u know what i mean) pipes ofcorce the floor heater pump can pump as much as 60-100 liters in minit so itl do
7 then havte make secondary cooling for other components not mentiond befor and special boh for hdds
im geting oil from powerplant i hope itl be cheap i havto filter it but i think i can do and my com wil be idle like in -40 - -30
this is my warplan
rivercom9
01-12-02, 12:39 PM
Samzik, what type of oil will you get from the powerplant?
i get condenstaor cooling oil or i dont know the wor in engl .
Originally posted by arhines
...but of course in koolances version the motherboard is on it's side. I like the idea in general, but I think that a water-loop waterblock is still needed to cool the cpu. Oil simply doesnt transfer the heat well enough in my opinion.
I remembered reading dr. Ffreeze stuff back in 98 (wow that was a long time ago!), and being disappointed because he gained practically nothing from the experiment...he did manage to get water droplets on the motherboard itself though...but luckily a layer of oil protected it. Heres the link to the water droplet nightmare :) http://www.drffreeze.com/Nightmare.htm
Very good point. Even if you have water int the mixture, it will be covered by a thin film of oil (thick enough to completely insulate it I believe).
Intraveinous
06-04-02, 08:52 AM
OMG, this thread is back... Anyone try anything or find any more info since January? :D
Peace
John
packratbob
06-04-02, 09:09 AM
oil cooling wat an idea. when u want KFC u just run ur xp for a couple or hours with out cooling and throw the chicken pieces in. yummmmmm.
u mean mineral oils r non conductive? i never know that or do they seal the boards??? someone explain to me how this works i'm all lost but it looks so freaken kewl
I want one I want one
Intraveinous
06-04-02, 09:39 AM
Correct, mineral oils are non-conductive, however, they have a tendancy to absorb into seals on capacitors and the like and eventually burst them because of the higher pressure. Also, you have problems cooling the oil and moving it around since it is so much more viscous than water...
Read through the entire thread and you'll get a pretty good idea of what's going on...
Peace
John
Anyone try anything or find any more info since January?
lol, yah I've found that I STILL can't afford that much quality oil :D
Odd that the case on page 1 hasn't shown up anywhere else. I guess it really was just a one-time experimental thing from Koolance.
technoweasel
06-04-02, 01:27 PM
FWIW At work we have some 80k volt power supplies
immersed in peanut oil.
The same oil has been in them over 10 years and never broken down, they are sealed, but have been open for days for repair.
They have no cooling other than the oil bath.
Haven't priced peanut oil but it might be cheaper than mineral
oil.
To protect the caps ,seems like you could coat the whole mboard with an electronic sealer like Krypton Crystal Clear. Cheap like
$5 a can. Mask off the card slots and cpu socket.
How about an aluminum case (tub) with fin to radiate the heat.
Definately would want to try on an old beater system first.
Just some ideas.
Technoweasel
As soon as school is over I am going to try this- at first only runninga p1 system submerged for a few weeks to test it out and see how it works in practice. And that idea with alumimum (or copper for that matter) sounds feasable.
youthemandan
06-05-02, 06:50 PM
Ok guys this is my idea. What i am gonna do is fill my mini freezer with mineral oil and sumerge my components and in doing this i won't have to worry about condesation.
Will this work???
This whole water sinking in oil buissness can be easially solved.
Even if you mounted your mobo vertically, your cards would still stick out horizontially, and you would have dead cards (assuming the thin film of oil dosen't work out) rather than a dead mobo.
What I was thinking was to take your fishtank (or whatever you use) and mount your mobo UPSIDE DOWN on a piece of plastic in the middle. The piece of plastic will have holes on the side to let the oil flow, and if any water got in for some reason, it would "drip" down the holes safely out of the motherboard's way onto the bottom of your tank.
Here's a pic describing it...
Originally posted by JigPu
This whole water sinking in oil buissness can be easially solved.
Even if you mounted your mobo vertically, your cards would still stick out horizontially, and you would have dead cards (assuming the thin film of oil dosen't work out) rather than a dead mobo.
What I was thinking was to take your fishtank (or whatever you use) and mount your mobo UPSIDE DOWN on a piece of plastic in the middle. The piece of plastic will have holes on the side to let the oil flow, and if any water got in for some reason, it would "drip" down the holes safely out of the motherboard's way onto the bottom of your tank.
Here's a pic describing it...
I was thinking about using a tilted sheet of plexiglass for that, but the holes might work as well.
A question : Can I submerge the cooling element of a fridge in the oil (I know that it will have a bit of trouble dissipating the heat but still)?
Ocelaris
06-09-02, 01:45 AM
ok, I have a fridge system on my KR7a board... I can get the reservoir down to -30*C. I don't see why I couldn't just immerse my entire board in oil slightly to prevent condensation. or I have a couple extra fish tanks lying around that would love to have a computer in them! anyways, my KR7a is dead I think due to condensation, so I'm getting a new AX-7 hehehe, but I have a K6-2 400 that would like to take a little dip in some mineral oil :)
what about like a model boat propeller to move the cooled oil around the tank? what about pure alcohol? I think that soon somebody has to find a good cheap dialectic liquid for these projects... back to water cooling till the majik likuid is found. bill
Look at this page:
http://www-ehs.ucsd.edu/lab/0808.htm
Mineral oils have a flash point of 93C while for example ethanol has it at 37.8C, which is not so high for the CPU. Burning your board is just as bad as shorting it out because of condenstation (plus having a tank of burning liquid in your room is generally a bad thing)
hehhee i was away from forums snif ;) but im back ... or i think i am my project vent down the toilet, due i couldnt get my hands on the oil and iv been so damnd bizzy, but one idea to test the theory
.... ods here it comes: u put any electrical instrument (somthing werry ceap) for example an logitecs blinker (the little thing vich comes vith logitecs optical mouse) onto oil to test its conductivity. When the little thingie is workin in oil take some old psu (vorkin one ;) ) and put into oil and put a went on or some diod, so u can c if it works or burns.
One good thing with condensator and transformer oils is they will come off easily ;).
ha u didnt know that hehhehe i knew a long time ago buit i have to confirm it. Point is that what hapens to water when it reaches 100 deg cels. it turns onto gas yes ? ok but oil turns to gas in much lower temps so if u have heater at home u can put ya mobo onto heater or some air blower and it drys !
so i see no problem here the real probklem is the oil it has to be quality oil to not to ruin your mobo and so on ;) ok peace.
i was checking my mail today when my hart wooped woot somone has responced to my topic ok i thought lets have to look what guy has to say and woops there was abuout 5 tresponces on let me know of ya tests and im trying to katche some oil from plant.
ok im out keep on coming with this stuff ;)
ergh, am i the only one who realy knows much of anything about this?
Alright, Mineral Oil (from my experience) was rather runny, but it just might have been that type.
for a case you can use an aquarium if you want. A styrofoam cooler (like you get from the gas station and store beers in for going to the <whatever>) has been used.
the oil will NOT cool itself (as has already been established) so you need a way to cool it. Pumping the oil through a radiator is not a good idea, kills your pump much faster. A Secondary cooling system is required. A simple water cooling rig (pump, rad, and water) will suffice. You get some copper tubing and bend it several times so you get a kind of an S (except it would be more than 3 passes) 5-6 is a good number of passes to make. you hook the copper tubing to the watercooling system in place of a waterblock. Submerge the Copper tubings in the tank, and wallah.
For a phase-change system you would do more or less the same thing, use the tubing for an evaporater.
If you're nervous about blowing the capacitors you can simply silicon it up, I belevie there are spray on conformal silicon products if you want to cover the whole board.
ajrettke
06-15-02, 06:03 PM
Isn't rubbing alchohal non conductive? I'm pretty sure there are lotsa of alchohals that arn't conductive. Make sure you get 99.9% rubbing alchohal though, junk at walmart is 30 H20 and 70 rubbing alchohal.
Penguin4x4
06-15-02, 06:35 PM
Originally posted by parkan
Any suggestions, no?
If its diaelectric, go for it, if you don't know, see if it'll arc, :D Mineral oil that is used for cleaning food surfaces is like liquid soap
________
TUBE 8 (http://www.****tube.com/)
Note the low flash point for alcohol. It's rather nasty if it goes up in flames. Oh, and I am building a custom "aqarium" case (got the shcematics, just need the money. I really want to be the first to do this!
Penguin4x4
06-15-02, 09:54 PM
DO NOT USE ANY ALCOHOLS
________
Herbalaire review (http://vaporizer.org/reviews/herbalaire/)
Captain Slug
06-15-02, 10:10 PM
A. The base of the caps could be sealed by simply caulking around the edges of them with a plastic dinner knife.
B. The oil would have to be kept turbulant in order to cool anything.
So here's my diagram recommendation...
Captain Slug
06-15-02, 10:12 PM
Oh yeah, I forgot to doodle in the secondary submerged low gpm pump to "turbulize" the oil around the RAM.
muddocktor
06-15-02, 10:31 PM
I work with drilling fluids in the oilfield and we sometimes use a synthetic oil base drilling fluid. The synthetic we generally use in these systems is an isomerized alpha olefin and cost between $7-10 per gallon. As far as I know, they are highly dielectric and essentially non-conductive. They are also essentially non-toxic and are also fairly thin, about like varsol or kerosene. There are several disadvantages to them too in that they don't conduct heat nearly as well as water and they also give some people a severe skin rash( like me:( ) and they are definitely rough on rubber and some plastics. We have high volume, high pressure piston type pumps used in the drilling industry to circulate the drilling fluids and when we go to a synthetic based fluid system we have to use specially designed pistons for the pumps due to the reaction that these oils have with normal rubber compounds. Also, the drilling fluid temperatures will also run about 10-30° C higher with the synthetic based fluid as compared to a water based fluid.
I'm not saying that oil cooling a rig is not doable but it will be a bear sealing all the caps and other components well enough with a compound that won't break down in your cooling oil. Good luck with the project though, if you decide to go ahead and try.
muddocktor
Thanks for the input! Do you know how it will react with silicone rubber?
One more question (this has been asked before I guess): After looking up the prices on http://www.sciquest.com/ the price for the oil that I had in mind come out to be around $110 for 20 liters. Would it be possible to get it cheaper from a local store/elsewhere on the net? For my first tryouts I do not really care about the quality.
muddocktor
06-16-02, 12:13 AM
I think that silicone rubber is a lot more stable in the synthetics than a rubber based mainly on conventional rubber. As far as the oil, if you live near a region that does oilfield drilling you might be able to get 5-6 gallons of their isomerized alpha olefin from a mud company (what drilling fluid is called in the industry, hense my name here on the forums:D ) for free if they are feeling nice that day. The three biggest mud companies are Baroid Drilling Fluids, MI Drilling Fluids and Baker Hughes Inteq. I think that the synthetics that they use are actually produced by Chevron, but that is just what I heard through the grapevine.
Penguin4x4
06-16-02, 12:21 AM
Silicone is far superior to rubber. They've got 5,000Hp Top-Fuel Dragsters using it to seal gaskets. Need I say more?
________
No2 vaporizer (http://no2vaporizers.com)
thegrouch187
06-16-02, 04:00 AM
i actually read an article while surfing about a guy who actually did a cooling project with mineral oil and he had like a ac unit or something and there was a plate that was apparently cooled somehow and he had the mineral oil pumped out from the case onto the cooling plate and the mineral oil would drip back into the case much cooler. That is one way and another would be to just adapt a water cooling system and continously pump the mineral oil around through a radiator. it seems like a good idea but it would make a big mess if it were to spill and the mineral oil would eventually start to insulate as heat diffuses off the computer into the oil the oil would start to get hotter unless some sort of cooling fan cooled the mineral oil. conventional cooling methods make more sense than this but its really akward seeing a computer immersed in a liquid and still working
thegrouch187
06-16-02, 04:02 AM
ohh and also i would be interested in knowing the freezing temp of mineral oil because if it was low one could just put it in there refrigiratior and run the cables out some how.
for freezing temps read in front of thred i soon put my idea of thingi out ul see i have gave some of thought about it ;) in past 6 months :)
Muddocktor, I live in New York (should probably put that in my "location") and I am not aware of any drilling buisness nearby. Any suggestions? Thanks in advance. :D
Captain Slug
06-16-02, 05:01 PM
Jeez, why didn't I notice that before.
The motherboard in that pic is the same one I got for free from a friend. It was a pretty cruddy Socket A setup but It overclocked like mad.
I got rid of it because the PCI slots wouldn't let me fit a Voodoo 2.
Who cares what the long term damage is to such a crappy piece of hardware :D
Everything but the caps would be able to survive such an experiment.
h2sammo
06-28-02, 01:18 PM
Any news on this yet?
:cool:
Intraveinous
06-28-02, 01:23 PM
Caps are Capacitors... Freezing point of oils is different depending on the composition, but it is much lower than water. Oil becomes more and more viscous, ie, thicker, the colder it gets, this could pose a problem if you plan on pumping it.
Peace
John
Oh i c thanks. But u dont have to beware of the temp of oil if its about -30 its enugh and ots normal viscosity. I know there are some oil pumps used in floor heating systems so there ya go u get 200-400 liter/minit pump for oil and they arnt werry expencive.
zeroKlvn
06-29-02, 04:47 PM
Has Koolance replied to any emails yet? Oh and pardon me if I didnt see it, but how did you get this pic anyways?
-zero
Just some thoughts
Use insulating material for all but one side ( the window ) every bit helps.
Rip the freezer element out of the freezer/fridge, and put it in the oil.
Use a big slow fan, well, more like a paddle wheel, to stir up the oil.
Mount the mobo as in the pic, so both the mobo and cards are vertical.
Hope you find these idea's useful :beer:
HaywirE
06-30-02, 03:02 PM
This just crossed my mind.
Who would go with Mineral Oil Cooling?
Yes intresting.
As far as we are is H20, And refigerators.
If you think about it, who needs a refigerator, sure when you look at your ASUS PcProbe and your CPU Is -5 Or 50F. That's a good feeling, but do we really need the cooling?
H20 Is all we need for now, even a really good overclock wont take you over the edge of 115F (ex 1.6A @ 2400)
Sure 115F Is high, your processor will last.
Arguments about 115F Tempatures
My Processor Wont Last as long Okay, do you plan on keeping it 10 Years[/b]
Okay, if you already have a 2.4GHZ P4 Or a 2000Xp, why overclock, the speed is sufficent.
Okay, why overclock if you get economical parts, but a 200 dollar watercooling system, and you could replace the CPU for cheaper, and not even risk the processor burning up
Yeah sure, there's many arguments to this.
Fact, you should only overclock if you're under the 1Ghz Barrier.
I suppose if you have a extra box laying around..
Peace.
HaywirE
Happily Crunching at 3972 MHz, Now a new member of O/C Seti@HOME
Fact, you should only overclock if you're under the 1Ghz Barrier.
lol I think most of this forum would disagree with that statement:D
Any bit more speed you can get out of your system is worth doing. Mineral oil cooling, yah its really extreme, but along with getting you more speed it's a really freakin cool project :)
Going totally over the top can be a lot of fun!
omg im just intrested would this work no need for it due my 733 is faast enugh for my needs so im just curious thats why i started the thred
.oO sam out Oo.
im olmost in biznes one guy is cheking the oil for me from a company hehhe wel c what wil hapen :D
could you use the fact that water will drop down in oil to use a mixed system ? mount the mb flat about 1/2 way down in the res. with a sheild of plexi over the mb to block water and the pump at the bottom. pump out to rad. bong. fridge.ect. back to top of tank just under oil (no air) let it fall and start over:eek: would look kinda like a lava lamp:)
vBulletin® v3.8.7, Copyright ©2000-2012, vBulletin Solutions, Inc.