Is it possible to incorperate a peltier into a WC loop to create a chilled water loop? I was thinking of making a custom res with one side of the peltier touching the water and the other half to a large heatsink outside the case. Would something like this work or even be worth attempting?
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
-
Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.
Peltier + Watercooling
- Thread starter synaps3
- Start date
the main problem with this idea is the fact that when you take the heat from the water with the peltier you are using electricity to get the heat from one side to the other. the radiator moves heat from one side to the other without the use of more electricity. also unless you are using the peltier in place of a radiator, you would be cooling the water in the loop, but at the same time raising the ambient (room) temp but the same amount plus the energy it takes to move the heat around. (in order to move 1 watt of heat from the water to the air requires more then 1 watt of electricity to run the peltier) when the cooled water got to the radiator, the temperature would be equaled out to room temp anyway. and you are adding heat to the room with the peltier so the radiator would be less efficient.
but you are on the right track... i actualy did something like this except oppositely
i took a bigass heatsink and epoxyd it to the 'cool' side of a 226w peltier. i epoxyd a 1/2" chipset block i had laying around to the hot side and mounted it on the INSIDE of my case infront of the air intake. this took the air that was entering the case, removed the heat from it, and deposited it outside the case so this heat wouldnt wash over my components. internal case temp dropped about 10C
but you are on the right track... i actualy did something like this except oppositely
i took a bigass heatsink and epoxyd it to the 'cool' side of a 226w peltier. i epoxyd a 1/2" chipset block i had laying around to the hot side and mounted it on the INSIDE of my case infront of the air intake. this took the air that was entering the case, removed the heat from it, and deposited it outside the case so this heat wouldnt wash over my components. internal case temp dropped about 10C
The rad (if I end up watercooling) will probably be in front of the case, while the heatsink for the peltier would probably rise out of the top, or be the case itself (anything is possible ). I'm just kicking around ideas right now, but it seems like it could work. I guess I would have to figure out what metal the peltier is to avoid corrosion though. I like the idea of a sub-ambient WC loop
). I'm just kicking around ideas right now, but it seems like it could work. I guess I would have to figure out what metal the peltier is to avoid corrosion though. I like the idea of a sub-ambient WC loop the peltier is ceramic, its a thin wafer that you have to sandwich between two components. if you are putting the cold side in the water loop you would need to epoxy/mount this side onto a water block and put this block in the cooling loop. the other side would have to be attached to a large heatsink and fan. i dont think you will see any noticable drop in temperature this way. maybe 1-2C if that. maybe i wasnt clear before...
say your loop looks like this...
rad>resevoir>peltier> pump>cpu>gpu>back to rad
if your room temp is say 20c and your computer is putting out 500 watts of heat raising your water temp to 40c (25watts per 1c increased). and also lets say that when water passes through the radiator the temp is reduced by half the difference between the water and air temp; 10c. so water exiting the radiator is now 30c. (although there will never be more then 1-2c variation in water temp throughout the loop)
when you turn on the peltier you are using say another 100 watts of power, 75 watts of that is used to simply move heat from the water to the air, the other 25 watts are generated as heat when converting from electricity to a temperature change. now you are adding 100 watts of heat to the air and taking 75 from the water. and the hotter the air, the less efficient your radiator will be. also remember that if, say you actualy do cool the water to 10c, when it goes through the radiator room temp being 20c (plus what you have raisied it to with the hot side of the peltier pumping out heat) the water will be 15c (x + half the difference) on the other side. so you only lowered the temp 5 instead of 10. heat is going out of the peltier into the air and back into the radiator.
*remember none of these numbers are real, only the theory behind them is correct*
its like turning on a light to use a solar powered calculator. it would be much more efficent and use far less energy if you simply used a AA battery then turned on a 100w light.
say your loop looks like this...
rad>resevoir>peltier> pump>cpu>gpu>back to rad
if your room temp is say 20c and your computer is putting out 500 watts of heat raising your water temp to 40c (25watts per 1c increased). and also lets say that when water passes through the radiator the temp is reduced by half the difference between the water and air temp; 10c. so water exiting the radiator is now 30c. (although there will never be more then 1-2c variation in water temp throughout the loop)
when you turn on the peltier you are using say another 100 watts of power, 75 watts of that is used to simply move heat from the water to the air, the other 25 watts are generated as heat when converting from electricity to a temperature change. now you are adding 100 watts of heat to the air and taking 75 from the water. and the hotter the air, the less efficient your radiator will be. also remember that if, say you actualy do cool the water to 10c, when it goes through the radiator room temp being 20c (plus what you have raisied it to with the hot side of the peltier pumping out heat) the water will be 15c (x + half the difference) on the other side. so you only lowered the temp 5 instead of 10. heat is going out of the peltier into the air and back into the radiator.
*remember none of these numbers are real, only the theory behind them is correct*
its like turning on a light to use a solar powered calculator. it would be much more efficent and use far less energy if you simply used a AA battery then turned on a 100w light.
I don't think you understand what my question was though.
Let's say I have a standard loop like you mentioned:
rad>reservoir> pump>cpu>gpu>chipset>back to rad
Now, my method does not add the heat to the loop. I was suggesting making a standard loop with a custom rad that is metal and has the peltier mounted externally. The cold side of the peltier would be physically mounted on the res, and the hot side would have a heatsink outside of the case (the res will go in the top 5.25 bay, and there will be a hole in the top of the case where the heatsink is). With this method, you get all the benefits of a cooler loop without the additional heat in the loop.
Let's say I have a standard loop like you mentioned:
rad>reservoir> pump>cpu>gpu>chipset>back to rad
Now, my method does not add the heat to the loop. I was suggesting making a standard loop with a custom rad that is metal and has the peltier mounted externally. The cold side of the peltier would be physically mounted on the res, and the hot side would have a heatsink outside of the case (the res will go in the top 5.25 bay, and there will be a hole in the top of the case where the heatsink is). With this method, you get all the benefits of a cooler loop without the additional heat in the loop.
theelectic
Limp Gawd
- Joined
- Oct 30, 2006
- Messages
- 362
If you chill your water below ambient you don't want to use a radiator. Otherwise you're making an air conditioner and putting heat INTO your loop not out. Think about it, cold water in the rad, warmer air blowing over the rad = cooling the air, heating the water.
A chiller can be used in place of a radiator, though you'd need a pretty big one, otherwise the results are middling at best (see CoolIT Freezone).
A chiller can be used in place of a radiator, though you'd need a pretty big one, otherwise the results are middling at best (see CoolIT Freezone).
Arcygenical
Fully [H]
- Joined
- Jun 10, 2005
- Messages
- 25,066
You'll need ALOT of Peltiers depending on the loop.
My loop generates about 220w of heat (E6300 & GTS) which would need at least a 300w peltier. A 300w peltier just to cope with the temperature load (VERY conservative measurement, most peltiers are only like 50% efficient).
If I wanted to go below temperature equilibrium, I'd need 2 300w peltiers, one hell of a lot of insulation and a PSU capable of supplying at least 700w of power... Not to mention the electricity cost. Now, if a CPU puts out about 100w of heat... And requires something like the Tuniq Tower is JUST able to cope with this heat load... You'd need a HSF about 2x the size of the Tuniq just to cool each peltier.
You could split it up into 5-6 100w peltiers, which would be managable, but take up a hell of a lot of room, require 5-6 really fast fans.
Next, having ANY radiator will bring heat into the loop. Period, not to mention the 25% antifreeze you'd need in your loop, lowering overall efficiency.
I've toyed with this idea for a LONG time, even built some basic setups using 2 200w Peltiers and coldplates out of monolithic copper and a drill press. It's just not worth it for any real cooling, it's a great novelty item though.
That is true about the rad-- I did not think about that.
I probably will not go with a peltier then. Looking at the Freezone and its SIX peltiers, to only cool the CPU, makes me a bit weary of even trying this. It would be fun, but a waste of my time. Thanks for the responses everyone.
I probably will not go with a peltier then. Looking at the Freezone and its SIX peltiers, to only cool the CPU, makes me a bit weary of even trying this. It would be fun, but a waste of my time. Thanks for the responses everyone.
http://www.procooling.com/index.php?func=articles&disp=84&pg=1
Better off just getting a compressor. Either use that to chill the water, or skip the water altogether and go direct die.
Better off just getting a compressor. Either use that to chill the water, or skip the water altogether and go direct die.
theelectic
Limp Gawd
- Joined
- Oct 30, 2006
- Messages
- 362
The Freezone, at least in my experience with it, is a joke. Think of it this way: it uses 56W worth of pelts, and has a regular (no heatpipes) heatsink and 92mm, roughly 60CFM fan to cool the 56W worth of pelts AND the heat load of a CPU AND the heat dump from a pump. A heatsink and fan isn't enough to cool the output of my CPU to an acceptable level, so how can it cope with the additional load of the pelts? It can't.