Sign in with
Sign up | Sign in
Your question

TEC build for controlled water temps

Last response: in Overclocking
Share
a b K Overclocking
October 21, 2012 8:22:56 AM

Ok so I have been thinking of building a TEC system for some time now and I have compiled a list of components for the project that I am now starting to assemble and test

245 Watt TEC’s X 4
Koolance CPU-370 Block X 4( hot side)
600 Watt 15V power supplies X 2 (TEC's)
1250 Watt OCZ power supply
1000 Watt OCZ power supply
800 Watt 30 Amp PWM motor controllers X 4 ( TEC's)
March Series 1 pumps X 2 ( hot side)
Swiftech 655 pumps X 2 ( cold side)
Watercool MO-RA3 9 x 120mm LT Extreme Radiator X 2 (hot side)
Swiftech QP 360 Radiator X 1
EX radiators X 12 (cold side)
3/8 ID 5/8 OD tubing (hot side)
3/8 ID ½ OD tubing ( cold side)
Case XSPC H1 Hive Ver.2
Insulating wrap 1 ½ IN. and 3 IN. ( cold side)
Dragon skin insolating coat epoxy ( cold side)
Lots of fittings

Well this is the just of the build anyway I will have pics of the progress and tests done to show how the system is configured and operates.
a b K Overclocking
October 21, 2012 8:40:15 AM

toolmaker_03 said:
Ok so I have been thinking of building a TEC system for some time now and I have compiled a list of components for the project that I am now starting to assemble and test

245 Watt TEC’s X 4
Koolance CPU-370 Block X 4( hot side)
600 Watt 15V power supplies X 2 (TEC's)
1250 Watt OCZ power supply
1000 Watt OCZ power supply
800 Watt 30 Amp PWM motor controllers X 4 ( TEC's)
March Series 1 pumps X 2 ( hot side)
Swiftech 655 pumps X 2 ( cold side)
Watercool MO-RA3 9 x 120mm LT Extreme Radiator X 2 (hot side)
Swiftech QP 360 Radiator X 1
EX radiators X 12 (cold side)
3/8 ID 5/8 OD tubing (hot side)
3/8 ID ½ OD tubing ( cold side)
Case XSPC H1 Hive Ver.2
Insulating wrap 1 ½ IN. and 3 IN. ( cold side)
Dragon skin insolating coat epoxy ( cold side)
Lots of fittings

Well this is the just of the build anyway I will have pics of the progress and tests done to show how the system is configured and operates.


TECs are nasty IMO, they put out a lot of unnecessary heat. If you're looking to do some extreme overclocking you might as well go with a phase change cooler instead
a b K Overclocking
October 21, 2012 8:46:31 AM

Yes that would be a better approach and would give greater results, but this is for the learning experience of controlling the water temperature directly. if the system is running to hot simply lower the water temperature by rising the TEC’s amp draw or voltage and the system will cool down. it simply sounded like a cool idea to try out and see if it works the way I hope it will.
Related resources
a b K Overclocking
October 21, 2012 8:58:02 AM

My only objective is to maintain a constant water temp coming out of the cold side radiators of 13C to 18C with one CPU and two video cards.
a c 231 K Overclocking
October 22, 2012 10:05:06 PM

toolmaker_03 said:
Ok so I have been thinking of building a TEC system for some time now and I have compiled a list of components for the project that I am now starting to assemble and test

245 Watt TEC’s X 4
Koolance CPU-370 Block X 4( hot side)
600 Watt 15V power supplies X 2 (TEC's)
1250 Watt OCZ power supply
1000 Watt OCZ power supply
800 Watt 30 Amp PWM motor controllers X 4 ( TEC's)
March Series 1 pumps X 2 ( hot side)
Swiftech 655 pumps X 2 ( cold side)
Watercool MO-RA3 9 x 120mm LT Extreme Radiator X 2 (hot side)
Swiftech QP 360 Radiator X 1
EX radiators X 12 (cold side)
3/8 ID 5/8 OD tubing (hot side)
3/8 ID ½ OD tubing ( cold side)
Case XSPC H1 Hive Ver.2
Insulating wrap 1 ½ IN. and 3 IN. ( cold side)
Dragon skin insolating coat epoxy ( cold side)
Lots of fittings

Well this is the just of the build anyway I will have pics of the progress and tests done to show how the system is configured and operates.



Have you actually ordered all of this stuff and is it in your possession?
a b K Overclocking
October 23, 2012 12:39:45 AM

over half of it, yes, i am testing with it now.
a b K Overclocking
October 23, 2012 12:55:03 AM

i bought a total of 20 little rads, so i have had a 8 stack to play with for a while now.
a c 231 K Overclocking
October 23, 2012 9:09:04 AM

So what do you actually have in your possession to test with?

You say you have 4x 245w peltiers what size mm x mm and amperage load are they?

And what are your intentions for the hot side heat cooling and cold side transfer of the peltiers?

a b K Overclocking
October 23, 2012 1:09:33 PM

i have 2 245w 50mm x 50 mm i am considering getting 2 545W 62mm x 62mm

the 245W are 18 amp and the 545W are 28 amp i have a TEC water block and a CPU water block for the hot side i am seeing with works better, and i am using the little rads for the cold side, after being wraped in insoulation the little rads can get quite cold.
a b K Overclocking
October 24, 2012 5:30:44 AM

As far as the hot side I intend to try and keep water temps between 160 F and 190 F with the 1080 rads cooling the hot side water it seems to be able to maintain that. Without a heat source to the cold side I do not know yet what the equilibrium temp will be but I will try to maintain a water temp on the cold side of 60 F when fully loaded after install, that is my goal anyway.
a c 231 K Overclocking
October 24, 2012 10:46:11 AM

toolmaker_03 said:
As far as the hot side I intend to try and keep water temps between 160 F and 190 F with the 1080 rads cooling the hot side water it seems to be able to maintain that. Without a heat source to the cold side I do not know yet what the equilibrium temp will be but I will try to maintain a water temp on the cold side of 60 F when fully loaded after install, that is my goal anyway.


I have already discovered cooling the hot side with water is a bad idea you'll reach temperatures your water pump and your water block was not designed to handle!

Their capabilities are limited compared to the temperatures you are planning on running through them.

You need to take into consideration when peltier cooling was experimented with in the early days regarding CPUs the peltiers were very small approximately 10mm x 11mm or 12mm, with a very low amperage draw.

Simply because the CPU dies were exposed in those days and it only required a very small peltier to cover the CPU.

Those peltiers were easily cooled with water because they did not produce the heat a larger 50mm x 50mm peltier does.

I've already learned the hard way water cooling the larger peltiers can and will lead to pump failure, that's why the industry is cooling them with finned heat sinks using air flow to cool with.

My radiator got so hot I couldn't touch it without burning my hand, a finned heat sink with a cooling fan can deal with those temperatures with no problems what so ever, I seriously urge you to reconsider trying to cool the hot side with water.

Quote:
I will try to maintain a water temp on the cold side of 60 F


You're going to all this expense for 60F, I would at least be shooting for 50F or 10C.

It may not be where you live but where I live 10C is still above the condensation forming point, for my setup condensation begins to form as micro fine drops at 8C, by the 5C temperature range condensation is a well formed drop, airflow over the water block and tubing inside the computer allows some leeway but 10C should be safe even for you.
a b K Overclocking
October 24, 2012 1:28:38 PM

That’s the reason for the March series 1 pump operating temp range goes to 190 F but you will notice that the pump motor is a very familiar motor size
a c 231 K Overclocking
October 25, 2012 9:23:19 AM

toolmaker_03 said:
That’s the reason for the March series 1 pump operating temp range goes to 190 F but you will notice that the pump motor is a very familiar motor size


How about a link to the actual pump you've bought there's a few March 1 series pumps?

At one point I was approaching 200F the boiling point of water is 212F standard water cooling products are not designed for those temperatures, 190F for that March series 1 is the Max Temperature, isn't it?

If you cooled the hot side with heat pipe air cooler you will be just fine, I urge you to at least reconsider!
a b K Overclocking
October 25, 2012 1:36:14 PM

yes with the single 360 rad I saw the same thing so I have tried to design this like a car radiator system but I do see what you mean with air cooling I may even do that for additional cooling as well. by adding air coolers on top of the water blocks to try and lower temps on the hot side further the closer to ambient that you can keep the hot side the better the cold side works, if your blocks are tight enough to the TEC when activated they try to separate so you have to have them tight to prevent this from happening.

i bought it from US plastics corp. it is the 97021 1A-MD-3/8"
a c 231 K Overclocking
October 25, 2012 6:10:20 PM

toolmaker_03 said:
yes with the single 360 rad I saw the same thing so I have tried to design this like a car radiator system but I do see what you mean with air cooling I may even do that for additional cooling as well. by adding air coolers on top of the water blocks to try and lower temps on the hot side further the closer to ambient that you can keep the hot side the better the cold side works, if your blocks are tight enough to the TEC when activated they try to separate so you have to have them tight to prevent this from happening.

i bought it from US plastics corp. it is the 97021 1A-MD-3/8"


You will not need the water blocks to cool the hot side at all use an air cooler directly on the peltier, if there's a size difference between the peltier and air cooler base make a cold plate out of aluminum flat stock, to make up the difference.

It is absolutely imperative you fully cover the hot side of the peltier, if any of the hot side is exposed, you will burn it out.

You can then control the fan speed on the air cooler to determine how hot you allow the hot side of the peltier to get.
a c 231 K Overclocking
November 2, 2012 11:50:58 AM

??? Earth to Toolmaker, Is everything OK?
a b K Overclocking
November 2, 2012 12:50:12 PM

Training for the past few weeks so I have not had much time, lots of stuff to do.
But will get back on to the testing and building soon, I now think that I will only need two Tec’s about 400W for this build as they have the potential to cool the water rather quickly, when the rads are insulated. I still need to figure out how to simulate a heat source into the system like that of a processor, I will need to tweak on that idea some more latter. Any ideas?
!