Sign in with
Sign up | Sign in
Your question

Would this work well

Tags:
  • Water Cooling
  • Heat
  • Overclocking
Last response: in Overclocking
Share
August 1, 2011 9:44:59 PM

I have a water cooling system in my PC and am looking to integrate a TEC.


Now I realize that a lot ot TEC's output twice as much energy in heat as they consume since they have temperature transferral from cold side to hot side but they also are expelling the energy that is pumped into them as heat also.

This is why you need another liquid cooler half the time just to cool the hot side since it is hotter than what your cpu produces.




I have a 125 watt cpu and obviously if I wanted to use a pelter diode I would have to get at least a 300 watt one making sure its at least 50% efficient. But then it would be very hard to cool.


So I though what if I buy multiple TEC's. About 6 of them and lower wattage ones which can be cooled with heat sinks and fans.

Have these 6 TEC's attached to water blocks which are put into the liquid cooling circuit.

60 watt TEC's should do for each and they can be cooled with heat sink and fans while they all share the load temp the cpu is outputting.

More about : work

August 1, 2011 9:49:13 PM

@rubix

I hope if you reply you don't tell me this wont work at all. Ah well I am kind of expecting it.


I should probably change my avatar since DNF sucked so bad.
m
0
l
a c 337 K Overclocking
August 1, 2011 10:30:04 PM

I think your water chiller was a better solution. Not many people run TEC anymore, but it's possible...I just don't know how viable it would be since most pelts actually are designed to sit directly on the CPU or GPU and cool it directly, while the hot side is watercooled. I think your idea is similar to setting up more of a 'cold-soak' application where you would attempt to force-cool a loop inversely to how a radiator removes heat.

Why not just get a big camping cooler, mod it with some 1/2" ports in and out, and place your rad(s) inside, using the ports to maintain a closed water loop, while your rads sit inside.. This will be your slush box. Add cold water or ice and keep the lid shut. This would also work if you used your water chiller idea...just drop your cold coils inside.

I think you are making this far too difficult on yourself. :) 
m
0
l
Related resources
August 1, 2011 10:56:22 PM

rubix_1011 said:
I think your water chiller was a better solution. Not many people run TEC anymore, but it's possible...I just don't know how viable it would be since most pelts actually are designed to sit directly on the CPU or GPU and cool it directly, while the hot side is watercooled. I think your idea is similar to setting up more of a 'cold-soak' application where you would attempt to force-cool a loop inversely to how a radiator removes heat.

Why not just get a big camping cooler, mod it with some 1/2" ports in and out, and place your rad(s) inside, using the ports to maintain a closed water loop, while your rads sit inside.. This will be your slush box. Add cold water or ice and keep the lid shut. This would also work if you used your water chiller idea...just drop your cold coils inside.

I think you are making this far too difficult on yourself. :) 



Well price is a thing for me or I would have just bought a phase cooler. But in reality I like making my stuff anyways.


TEC diodes are cheap but for me to buy multiple water blocks that is going to be expensive plus the heat sink and fans to go with them. Water blocks are at best $40 each and the heatsink fan combo's it will come to at least $400

All I really need is something like a water block but longer so I can put a couple of TEC's on it and one huge heatsink + fans.

Any ideas?
m
0
l
August 1, 2011 11:03:33 PM

Figured out what I am going to do. Basically I will put the TEC's on a metal box. A square oil can or something might be good. Then I will get the heat sink and fans for them.

Fill the metal box with liquid. Less is better.

Then the water coolers radiator will be submerged in the liquid without the fan of course.

I think I will have to try to get the water to move a bit to ensure it circulates.

That isn't an extreme setup but it will cool sufficiently.

And I will use 3 91 watt TEC's instead.


When I make it I will post it and see how high the 965 will go.

Since I had it stable @ 3.9 ghz just on air then I hope at least 4.6ghz is achievable.
m
0
l
!