Crazy liquid cooling ideas

[Guest]

Has anyone been motivated enough to experiment with incorporating peltier effect devices into a water cooling system? The big problem with them was always condensation, but if you incorporated them into a device outside the case, you could get around that. For anyone who doesn't know, these are little metal plates that, when you run a current through them, one side gets hot, the other cold. I don't know the scientific principle behind the Peltier effect. Anyone?
For that matter, does anyone have any wild and crazy ideas about how to squeeze every last degree out of your cooler?



-- All flea markets start from scratch. --

 

[mpjesse]

I've seen plenty of peltiers with water coolers... it's actually quite common.

-MP Jesse

"Signatures Still Suck"

 

[hatimh]

I know this sounds funny, but what would be nice every now and again for example just before you start a session of 3d gaming or something, just chuck some ice cubes into the water container. Actually come to think about it, it would make the water lower than room temperature (or case temp for that matter) and possibly cause some condenstaion around the hose pipes and water block. Am I right?

Beer is the devil's piss.

 

[mpjesse]

yes- you would have a problem with condensation. Another solution a lot of people do is put antifreeze in the water. As you probably know, Antifreeze mixed with water reduces water temps.

-MP Jesse

"Signatures Still Suck"

 

[CALV]

antifreeze mixed with water actually lowers the freezing point- doesnt make it colder


Next time you wave - use all your fingers

 

[mpjesse]

Yes- this is true. HOWEVER it also makes it harder to heat up water. Try it for yourself... hell, read the back of a Prestone container- it'll tell you. It actually raises the boiling point of water by 10 degrees. Why do think people in places like Florida put anit-freeze in their cars?

-MP Jesse

"Signatures Still Suck"

 

[Guest]

>Yes- this is true. HOWEVER it also makes it harder to heat up water.

No, it makes it boil at a higher temperature (and freeze at a lower temp). But actually, antifreeze has a significantly lower heat capacity then water. Which means that it takes less heat to raise the temperature of a gram of antifreeze then it takes to raise the temp of a gram of water the same amount. In short, it's not as good a coolant as water in that respect. That's why you don't run pure antifreeze in a car.

Considering that your CPU temp should never get even close to the boiling point of water ( and, not near freezing either), the only reason I could see for adding antifreeze to a water cooling rig would be for the corrosion inhibitors.


In theory, there is no difference between theory and practice.
In practice, there is.

 

[mpjesse]

If this is true- then why do anti-freeze variants specifically designed for water cooling work? If anti-freeze raises the boiling temp. of water, then reason dictates that it would be 10 degrees cooler at a higher temperature (compared to plain water). A 50/50 mix of antifreeze and water raises the boiling temp. of water by 10F Does that make sense? Or am I wrong?

-MP Jesse

"Signatures Still Suck"

 

[Guest]

>If this is true- then why do anti-freeze variants
>specifically designed for water cooling work?

I'm not sure I follow your reasoning. But this is physics, not philosophy, so I'll try to be more clear.

All materials have some thermal capacity, also called specific heat. This is a measure of how much heat a measure of material contains at a given temperature. Said another way, it is a measure of how much heat must be added to the mass of material to raise its temperature by a given amount.

Here are some links for the specific heat of <A HREF="http://www.electro-optical.com/material_props/water_thermal_prop.htm" target="_new">Water</A> and <A HREF="http://www.electro-optical.com/material_props/liquids_thermal_prop.htm" target="_new">Ethylene Glycol</A> (the principal component of automotive antifreeze).

At 0 deg. C:

Water: 4.225E+3 J/kg-C
Ethylene Glycol: 2.261E+3 J/kg-C

The units here are Joules (a measure of heat) per Kilogram - degrees Celsius. As you can see, the number for water is nearly twice that for the Ethylene Glycol. The ratio is actually 1.87 to 1

What this means:
Say you have 1 Kg of water, and 1 Kg of Ethylene Glycol.

If you add 4.225E+3 Joules of heat to the water, you will raise its temperature 1 degree celsius.

But to raise the temperature of the Ethylene Glycol by 1 degree celsius, you only need to add 2.261E+3 joules of heat.

Flipping the problem around:
Say you add 4.225E+3 Joules of heat to each sample. The temperature of the water will increase by 1 degree celsius, and the temperature of the Ethylene Glycol will increase by 1.87 degrees celsius.

This is all about heat capacity (specific heat). The boiling point of the material is a seperate matter.

Something else to consider is the thermal conductivity of the material. From the same links as above:

Water: 0.566 W/m-C
Ethylene Glycol: 0.242 W/m-C

This means that water conducts heat about 2.3 times better then Ethylene Glycol. Water is much more effective at absorbing/transferring heat then Ethylene Glycol.

So why use it?
It does raise the boiling temperature somewhat. But in an auto cooling system, the boiling point is substantially raised because the system operates under pressure.

It reduces the freezing point substantially.

It is a corrosion inhibitor.

Why use it in a CPU cooler? Corrosion is the only reason I can see.

>If this is true- then why do anti-freeze variants
>specifically designed for water cooling work?

Do they work?
I'd be interested to see a benchmark with pure water as the coolant, and another with pure Ethylene Glycol antifreeze (or a mix). Also, water with a wetting agent only.

<i>Cognite Tute</i>
(Think for Yourself)

 

[mpjesse]

Holy crap George- you obviously know what your talking about. If went through all those equations to tell me where I'm flawed in my logic... then by Jove your my new hero! Really. I do understand most of it though.

So, we've concluded that anti-freeze is only good for anti-corrosion in a water cooled system. That's all for now- tune in next time when ergeorge tells us the origins of the universe and who really created man.

-MP Jesse

PS... i'm joking- please don't take offense to any of it. Thanks for the help- really.

"Signatures Still Suck"

 

[Guest]

LoL

<i>Cognite Tute</i>
(Think for Yourself)

 

[yoda271828]

then why do anti-freeze variants specifically designed for water cooling work?

I believe that you are referring to products like 'Red-line Water Wetter'. They are not anti-freeze. What they do is reduce the surface tension of water thereby allowing it to absorb heat more efficiently.

 

[mpjesse]

I've seen those- but I know some people use anti-freeze because it supposedly helps with cooling performance. But I now know this is wrong.

-MP Jesse

"Signatures Still Suck"

 

[AMDkid1]

I've been using Military grade Glycol used to cool a massive radar. For over two years now, it also has non-conductive rating.
It works great with little or no evaporation as well as great cooling.

 

[delluser1]

WOW.
Revived after 9 years, fantastic.

 

[rubix_1011]

I wondered where the musty decay smell was coming from...