Test results for TEC chiller

[Brent1]

Hello I decided to start a new thread for these tests. If you want to get the rest of the story for why I have posted this you can catch up by reading this thread first

http://forums.procooling.com/vbb/sho...569#post155569

Test 1

Well I have done some tests since my last post and found some amazing results with the
Use of different coolants.

(1) the first test was with the use of 75% distilled water and 25% glycol for the hot loop
(2) cooled loop I used 75% distilled water and 25% glycol


I used the same mixture on both sides because I needed something for the leaching of metal ions not to occur. The only different metals that are in my loop would be the fittings. With this mixture testing the computer with Prime 95 for 48 hours my temps ran at hot side 62c the cold side 31c bear in mind I was only cooling the CPU for any of the tests


Test 2

In this used 2 different mediums

(1) The hot side I used just distilled water no other agent at all. I found through research that distilled water will not carry ions because there is no monticules that the ions can attach themselves to but you do need something to word off algae but for these tests I did not add any since they where short tests
(2) The cool side I used 75% distilled water and 25% ethanol


The same test was preformed as above with the use of Prime 95 for the duration of 48 hours
Test results where as fallows both the hot and cold sides yielded better results this time and stayed at 60c the cold side at 29c

Test 3

(1) For the hot side this time we used the product Water Wetter and distilled water. This product removes heat and can be cooled quickly.
(2) For the cold side this time we used 70% distilled water and 30% alcohol

Using the same tests as above we found using the Water Wetter was a good choice the hot side temps went down considerably too 56c and the cold side temps where 27c 3 degrees above room temp

The list of what was used for equipment:
(1) 156w TEC with Epoxy Pot, bare leads. Dimensions: 48mm X 48mm X 3.4mm
(2) Copper custom made water block for the hot side
(3) Copper custom made water block for the cold side
(4) Copper water block for the CPU http://www.bteckcorner.com/products/...Water%20Block/
(5) 2pumps http://www.bteckcorner.com/products/pumps/12-V%20Pump/
(6) Rads http://www.bteckcorner.com/products/...60%20Radiator/
(7) Computer System is a MSI 939 4800+ system used for the tests
(8) All tests where preformed with closed loops on both the hot and cold sides of the TEC

We also tried using a 226w TEC with the same blocks and found there was not very much gain in the cold side because we could not despite the hot side fast enough .
In order to make the larger TEC work efficient we would have to build the water block for the hot side larger and use a larger rad to cool the liquid faster to receive better results on the cold side.

Just to see what would happen with a larger TEC we went to our machine shop and did the math and built a larger water block for the cold side then put it all together the tests where done with out a computer just a dry run and the results where amazing in 5 minutes the water block that would be installed on the CPU had the temp of +3c in 11 minutes the temp was -6c so there for we found if you do not cool the hot side fast enough you do not yield good results on the cold side.

We also decided to use products that are readily available to anybody sure we could have use industrial strength product for coolants but that is giving an unfair advantage over people that do not have availability to industrial coolants. We like to do real world testing with the same products that any customer would use.

There is allot of misconceptions out there we found with our testing the use of Tec’s if configured right you can make a very good cooling system but you have to do your home work and be willing to try some thing different good luck and have some fun with Tec’s


Well I hope this is of some help to the extreme cooling junkies out there as we learned a lot and had fun doing these tests.

Regards
Brent