Need advice about (some strange homemade))) watercooling

peaceduke

Honorable
Dec 18, 2014
230
0
10,690
first of all hello ! and let me say that its homemade not because I ain't have money to buy a cooler
its because my water in place where I work is really cold...

here is the situation . I'm working in place where I can take water from outside (talking about only wintertime)
its kind of a village , receiving water from down below the soil .
so water is almost ice , I can't even touch it , my hand huts after 10-20 seconds
once I left hand for minute and almost had serious injure (hand was red almost all day)

I'm thinking of cooling CPU with that water , I have local folks who will weld some CPU housing from copper, see image below . (system (cooling rig) will go out of PC case , so , no MB flood will be possible , )) unless copper housing will explode .

what do you think about it?

socket.jpg
 
Solution


Heat flux is also proportional to dT, I think the coeffcient is per degree of difference in temp between the two sides of the interface. Cold water would be better, until it got warm, BUT can create other problems like condensation.


You actually can't cool with regular water.
If you were to expose the pipe of your home, you would see that the interior is coated in minerals.
Those minerals would deposit on the fine fins of your block and it would clog in a matter of days.

You can use distilled water, but even that isn't suggested as metals will eventually find their way to the water, and can cause corrosion when they go from various copper and aluminium alloys.

That is why, like your car, special coolant is almost always used. This not only prevents boiling and freezing, but more importantly prevents corrosion from unlike metals.

 
It should also be noted that the cooling properties of ice cold water and regular water are identical. Cold water would take longer to heat up, but it wouldn't provide a lower CPU temp than room temperature water. This is because the limiting factor is getting the heat from the block to the water, and that thermal coefficient doesn't depend on the temperature of the water.

It would only matter if the water was at the same temp or higher than the CPU.
 

peaceduke

Honorable
Dec 18, 2014
230
0
10,690


thanks for response !
I know about minerals and corrosion stuff , I need that cooler for 1 season
don't you think that it should live at least 3-4 month ? (I will make new housing and tubing each 3-4 month, moreover all that connection will be outside the house so I can use regular tubes) + I will put a small cooler at the top of a housing
in case something happens its not going to fry CPU immediately...

but what you said is correct of course . I had to ask that above
will those minerals and corrosion will appear only in interior of that housing ?
if its gonna affect only interior and exterior (outside body) will remain dry
I think its ok then !
no ?
 

peaceduke

Honorable
Dec 18, 2014
230
0
10,690


thats interesting ...
I thought ice cold copper is more important than water temperature itself
if copper will be cold (water will be moving very fast inside) lot faster than in regular watercooling systems
so you're saying its not going to work anyway ?
 

peaceduke

Honorable
Dec 18, 2014
230
0
10,690


because I want to overclock CPU over 5GHZ
 


Heat flux is also proportional to dT, I think the coeffcient is per degree of difference in temp between the two sides of the interface. Cold water would be better, until it got warm, BUT can create other problems like condensation.
 
Solution