• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

A good cooling liquid?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

vuo

New Member
Joined
Mar 6, 2001
I don't want to use air cooling because of the noise, so liquid cooling would be good. Many of the overclockers use water as the cooling liquid. It has a high thermal capacity, it's abundant and easy to replace, but it is conductive, may cause corrosion and lifeforms try to live in it. Leakage is extremely dangerous. The Silentium used instrument oil, which is flammable and volatile like every organic solvent. Other organic solvents are also poisonous when evaporated. The car radiators have glycol-water mixture, which combines the perils of organic solvent and water.

So the question is: Is there a non-conductive, high-thermal-capacity, non-flammable and non-volatile cooling liquid? There's of course the fine electronics testing fluids, which are so expensive that I don't even have so much money.
 

Metaxas

Senior Member
Joined
Dec 25, 2000
Distilled water. Redline water wetter (corrosion prevention, and cooling enhancment) Some anti-freeze to keep things from growing in your cooler (yuck) and its cooling properties, and that's it. Well, a tiny splash of blue dye-lite too... :) Now, someone please tell me if I have this wrong, but anyways, here goes.

As I understand it, distilled water is de-ionized, which makes the water WAY less conductive, it still is conductive, but not like plain ole water. Well, I guess also that's why all manufactures that construct water cooling systems advise their customers, to set up the system outside of the case, and run it for 24 - 36 hours before you install it, that way you can look over it, and make DAMN sure there is no leaks. Well anyways, good luck.
 

BillA

choke man
Joined
Jan 2, 2001
Dow Corning 200 Fluid, 510 Fluid, and 550 Fluid
These are silicone fluids with fairly low viscosities and fairly good thermal properties.
We're talking -40 to -20C right ?
But a 50 / 50 mix of ethylene glycol is good for -30C, and worlds cheaper.

Anyone able to get the specs for the "inproved thermal properties" of the non-silicate Texaco/GM coolant Dex-Cool ?
(I have searched and e-mailed to no avail.)

BTW, distilled water is sterile and clean, DI water is "polished" to remove ions (which reduces its electrical conductivity); BUT it is unstable and quite corrosive (it wants those ions back). There are no metal components in a DI water system.

be cool
 

Fink

Member
Joined
Dec 18, 2000
Metaxas (Mar 06, 2001 06:11 a.m.): (snip)...Now, someone please tell me if I have this wrong, but anyways, here goes.

As I understand it, distilled water is de-ionized, which makes the water WAY less conductive, it still is conductive, but not like plain ole water. Well, I guess also that's why all manufactures that construct water cooling systems advise their customers, to set up the system outside of the case, and run it for 24 - 36 hours before you install it, that way you can look over it, and make DAMN sure there is no leaks. Well anyways, good luck.

Distilled water is not always de-ionized. Deionization is a chemical process involving ion exchange resins ect ect, distillation is just a physical process and may not remove all the ions. Although, you will remove 95% of the ions during a distillation process, to a chemist this is not sufficently deionized, but for a watercooling rig, distilled water water is deionized enough.

Yes, removal of the ions from the water increases the resistance within the water.

Manufacturers recommend that you run the system external to your computer mainly to check for leakage. Cheers.
 

Metaxas

Senior Member
Joined
Dec 25, 2000
BillA (Mar 06, 2001 07:26 p.m.):
Dow Corning 200 Fluid, 510 Fluid, and 550 Fluid
These are silicone fluids with fairly low viscosities and fairly good thermal properties.
We're talking -40 to -20C right ?
But a 50 / 50 mix of ethylene glycol is good for -30C, and worlds cheaper.

Anyone able to get the specs for the "inproved thermal properties" of the non-silicate Texaco/GM coolant Dex-Cool ?
(I have searched and e-mailed to no avail.)

BTW, distilled water is sterile and clean, DI water is "polished" to remove ions (which reduces its electrical conductivity); BUT it is unstable and quite corrosive (it wants those ions back). There are no metal components in a DI water system.

be cool

O I C...said the blind man, as he stepped into on-coming traffic.... So, de-ionized water would not be a real good choice then 'ey? And you bring up another point too, alternatives to water completly, no water at all, however, I wonder if those coolants would work well with a small pump system like this, or would they be too thick? I still think that liquid mercury would be the da bomb shiznit to cool with, IF you could guarentee no leaks, no evaporation, don't touch it, don't eat, drink, snort, smoke or whatever, and you had a MASSIVE pump, with tubing that could handle that heavy *** stuff...you would not have to worry about overheating again. However though, I don't think that would be so hot for supercooling though, dosen't liquid mercury freeze solid at around 30-40F?
 

BillA

choke man
Joined
Jan 2, 2001
Just my personal observation:

I've found that virtually no watercoolers actually want (or are able) to get into and sort out the interelationships between the coolant, punp, and system (hoses, fittings, flow control, etc.) design.
Not ment as a putdown, it just becomes very complex very quickly.

For example; off-the-shelf aquarium pumps are often selected for their (no suction or head loss) flow rate, or noise, and sometimes reliability. Information on the effect of viscosity is not available as that is not how the pumps are used. Yet sub-zero coolants use antifreeze which does increase the viscosity, the more so as the temp drops.
(I use a recirculating lab chiller at -30C.)

"Plug-and play" watercooling is not too hard, designing a system for atypical conditions is.
But its a good technical challenge and most educational.

be cool
 
W

William

Guest
OK, distilled water is NOT conductive if it is truly distilled. Over time it will as it picks up copper ions. I am using triple distilled water from a professional chemistry distillation apparatus. But, even store bought distilled water is not conductive. It will over time though for the above mentioned reasons. My chem teacher took a whif of water wetter and determined it to be a little alchohol of some sort, some sort of detergent, and something else that i forgot. You can use many things as a substitution for it however. Watter Wetters main purpose is to reduce the surface tension of water, in fact salt does this really well(also adds a TON of conductivity). The solution I am using is 10% or so of Water Wetter, about 5-8% antifreeze, and the rest triple distilled H20. And don't diss on my chem teachers unofficial analysis, she has a PHD in Organic and knows her stuff.
 

Metaxas

Senior Member
Joined
Dec 25, 2000
BillA (Mar 07, 2001 07:47 a.m.):
Just my personal observation:

I've found that virtually no watercoolers actually want (or are able) to get into and sort out the interelationships between the coolant, punp, and system (hoses, fittings, flow control, etc.) design.
Not ment as a putdown, it just becomes very complex very quickly.

For example; off-the-shelf aquarium pumps are often selected for their (no suction or head loss) flow rate, or noise, and sometimes reliability. Information on the effect of viscosity is not available as that is not how the pumps are used. Yet sub-zero coolants use antifreeze which does increase the viscosity, the more so as the temp drops.
(I use a recirculating lab chiller at -30C.)

"Plug-and play" watercooling is not too hard, designing a system for atypical conditions is.
But its a good technical challenge and most educational.

be cool

Heh, I bet that lab chiller kicked your wallet dead in the ***.....LOL ;D No really, how much does something like that cost? Not trying to be nosey or anything though... :)
 

BillA

choke man
Joined
Jan 2, 2001
no Metaxas, it was no big deal, in fact FAR cheaper than what most "high end" watercooling pieces cost.

Lab chillers cost $1800 to $2500 or more, depending on their capacity; and I've taken a bunch of flack from clowns who don't wish to investigate a bit - read "eBay".

For a Haake A82 (-40C max, 125W capacity at -10C, control accuracy 0.01C from 14 to 150C, max flow rate 25L/min) I paid $21.00 plus $41.00 shipping.
But on the downside, I also paid $130 for a unit that did not work !

Total cost $72.00 (or $200 if I cannot fix and resell the "bad" unit) and I no longer need a pump, radiator, temp control circuit, or TEC and dedicated TEC psu. I will easily recover twice the cost by selling off the surplus stuff.

There are some tradeoffs, nothing is "free".
-40C coolant is a real challenge, ethylene glycol has an odor, and the unit sounds tike a small fridge; I'm putting the whole shebang in the garage with 1/2in insulated lines to the computer. The next step is to put the computer there also and run it as a server.

Look on eBay under lab/recirculating chiller/cooler/bath etc.

be cool
 

Metaxas

Senior Member
Joined
Dec 25, 2000
Woooooo....really...killer... :) You know, that's a DANDY idea really, you wouldn't need all that extra stuff, just the circulater and the block and that its. Hmmm.... well hey, you learn something new everyday. Howver, I stay AWAY from ebay like its the plague. To much fraud. I'll just suck it down, and hope to find a used one from a more "repitubable" establishment, or just buy one outright. I'd much rather invest and pay 800 -1000 in something, and KNOW what i'm am getting, and also that it WILL work, then spend 200 on soemthing that CANNOT be guarenteed. That may sound hard headed, and if it does, well, sorry. :) I mean, really, I spend 1k, and get my money's worth, or hit eFraud and spend 200 and its just wasted, and there isn't anything you can do about it.
 

Fink

Member
Joined
Dec 18, 2000
dunno260 (Mar 07, 2001 08:51 p.m.):
OK, distilled water is NOT conductive if it is truly distilled. Over time it will as it picks up copper ions. I am using triple distilled water from a professional chemistry distillation apparatus. But, even store bought distilled water is not conductive. It will over time though for the above mentioned reasons. My chem teacher took a whif of water wetter and determined it to be a little alchohol of some sort, some sort of detergent, and something else that i forgot. You can use many things as a substitution for it however. Watter Wetters main purpose is to reduce the surface tension of water, in fact salt does this really well(also adds a TON of conductivity). The solution I am using is 10% or so of Water Wetter, about 5-8% antifreeze, and the rest triple distilled H20. And don't diss on my chem teachers unofficial analysis, she has a PHD in Organic and knows her stuff.

Right, over time it will pick up ions from the metals it is exposed to, that is why you use antifreeze or waterwetter to counter the increase in ions over the course of time .e.g the anitfreeze acts as an 'ion buffer' - without getting to techical. Your tripple distilled water is probably highly deionized, since it was distilled in a propper, all glass (I assume) apparatus that isolated it from absorbing ions from the typical metal distillation appartaus that will be used in commercial applications. I won't diss on chem teachers, I too have a Ph.D. in organic chemistry.