Liquid

[Brent1]

Hello everyone I have a question for you. I have built a chiller with a TEC and I have the hot side been cooled with water. The cold side cools water to send cold water to the heat sours witch passes through another water block. Here is my question what would be the best liquid mixture to use for the cold side I have used glycol mixed with water so it doesn’t freeze but is there anything else I could use such as something clear so I can use dye to color the water better? Or even if there is something else that doesn’t end up so thick after mixing to get the desired mix so you don’t get freezing happening.

Thanks to all
Brent

[littleowl]

have you tried 50/50 water and antifreeze or denatured alcohol??

[n00b 0f l337]

Liquid chilling is not very efficienct with most designs, do you have a diagram or plans of your loop? Probably better to run one loop and DD the tec.

[Senater_Cache]

I would assume the purerer (H20) the mixture the better. add neccessary corrosion and bacterial inhibitors in minimal amounts. Just dont let the TEC freeze the water.
I am basing this off of non-Thermo Electric water setups, so I dont actualy know that what I said holds true in the TEC world.

[Brent1]

Hey thanks guys for your responses. Just a short responses to your answers if I may Tec’s are as effective as your setup, poor setups make every one think they are not effective. That goes for water cooling a poor setup will only cool as much as passive. One thing to bear in mind a TEC is not refrigeration and can not be compared to refrigeration or phase change. I have worked a lot with all of the above and found with careful consideration and planning Tec’s are about the most cost effective solution for cooling unlike phase change or refrigeration that will break the bank, not to mention if a component of the later goes the repair cost will break the bank again not to mention fry your rig with Tec’s the most I have ever seen for repair is to change out the PSU. Again it falls to the end user on what solution fits best.

I have made my own concoction lol, with alcohol and distilled water and Water Wetter my temps on the cold side have dropped considerably from using just glycol and water mix. I have not frozen up yet so this is a good thing.

All my water blocks and rad are copper I do not use aluminum in any of my water cooling solutions for my self or my customers.
But as for the mixture for the cold side any one that has any other ideas I’m all ears. I’m trying to find a medium that will give the coldest results and not freeze other that refrigerant lol

Thanks again I would like to here more on this subject

Regards
Brent

[Brians256]

Methanol works great if you go below 32F. Above 32F, use as pure as possible water for best thermal transfer.

With water, you will need to add anti-corrosion and biocides. Water is a better heat transfer liquid than methanol.

With methanol, you need to watch evaporation. Vinyl tubing is actually quite porous and methanol (slowly) goes right through it. Methanol fumes are poisonous.

[Brent1]

Thanks Brians256 yes what I’m wanting to discuss is what would be the best liquid mixture for the cold side of a pelt in a liquid chiller loop with out going pressurized what would be the best liquid to use for cold transfer with out freezing the liquid. As for the bacteria or alga, I would not be too concerned with it using Alcohol or Methanol. Not from the research and studies I have found anyway.

Description of my chilling loop, the hot side has a water block that circulating (with the help of a pump) hot water through 2 rads for cooling the hot side of the pelt. It is a closed loop. The cold side is also a closed loop witch also has a water block on the pelt to cool, circulating water mixture that will not freeze through a pump then to a small rad with a blower to circulate and cool the surrounding air in the box then to a second water block that keeps the chip cool.

Just wanting to see what you all think it is interesting to hear new ideas. I have read a lot of your articles in this forum and I’m impressed with the wealth of knowledge that you all poses in this field of Thermal electronics, to add I’m new to this forum and it is my pleasure to meet you all I will trust we’ll have some interesting discussions on a lot of topics in the computer sciences field.

Best Regards

Brent

[Brians256]

My suggestion is to use just enough methanol to keep the cold coolant liquid at all times (you don't want ice buildup anywhere in the loop).

For the hot side, I would use distilled water and antifreeze.

[Brent1]

Ok Brians256 thanks I will give the methanol a try on the cold side. I have been using Water Wetter and distilled water for the hot side through testing it seems to dissipate heat allot quicker than antifreeze. I will let you know how it gose

Regards
Brent

[qet77]

you can buy industrial coolant products that are clear, the one we use at work for our CO2 laser is called "DOWFROST" and it is clear. they are pretty expensive but those are the only clear ones i know of . people are correct about water moving the most heat per unit mass also. i think ammonia is a little better but only at high temperatures plus if you got a leak you would probably die.

[BillA]

I'm designing a biodiesel plant having varying amounts of methanol in all of the process streams
putting anything more than ppms of methanol in flexible plastic tubing is just plain stupid
and its effects WILL make one more stupid yet
very damaging stuff to the nervous system

[Marci]

Call me daft, but easiest solution (predyed the usual UV greeny yellow) is buy some MCT-40... (??)

[BillA]

sounds fine
any choice is better than methanol, have come to be quite wary of the stuff

[bobo5195]

Um Tec’s being good has a water cooler means not very much when Tec’s are a far better cpu cooler.

Um guessing your setup is this

Wb (cold side)
Tubing
Two waterblocks sandwiched together as a heat exchanger with a pelt in the middle to act as a kill
Normal loop (rad, res etc)

This is never going to be as efficient as a pelt water block combo as the cold side loop has a thermal resistance which is greater than a direct connection. There are some reasons I can think of do this though:

1. Chill multiple waterblocks at the same time
2. Have some odd sized pelts to hand that wouldn’t work in normal cpu cooling applications
3. Want to direct die cool the blocks
4. Double TEC setup of some sort with the cold side having another chiller.

Either way inline water chiller on ebay is probably the best solution but as an exercise this only makes sense if you are planning a direct die cooling with something like FC 77 (fluorinert) on possibly multiple blocks. This would allow for direct die cooling but the benefit is likely to be marginal vs. a pelt (cool though).

Double pelt option doesn’t make much sense as they might as well be sandwiched back to back. If you have odd sized pelts it makes much more sense to buy another than spend loads on gizmos.

[Brent1]

Yes thanks for the info bobo5195 I am cooling several chips at the same time I’m not looking for a lot of cooling I know the math but it will cool better than just water cooling and that is what I’m after here. This is a project that I’m doing and having fun with I know I could go out and buy a chiller and all that jazz but this is some thing that I like to do is play and mod things as for the gadgets I have lots of left over stuff from builds I do for customers or from rigs I had bought that where upgraded or what have you. So it is no money out of pocket. I own a custom computer & cooling solutions store. We do a lot of custom stuff for consumers as well as for commercial customers. Who knows I may come up with something that will work for a customer one day but I highly doubt it lol.

Once I have everything built and if it works or not I will post a work log and pic's for your enjoyment and maybe it will give that question an answer to someone aswell as myself.

[Ls7corvete]

Well honestly there is not much being posted on what is the best in TEC cooling these days. So I think that you probably know as much as any info people can find on the forums, test different mixtures and let us know what works best.

Im with Bobo though, the best thing to put between the cold side of the TEC and the CPU is nothing but 8mm of copper.

[Brians256]

Or another TEC (if you can find two that are sufficiently powerful enough to deal with the massive innefficiencies of stacked TECs).

[pH(x)]

Here from the past to save the day!

pHx to the rescue -- lol

I am researching something similar as well, and have found 2 prime candidates! OptiCool biodegradable transformer fluid from DSI Inc., or a mixture called Fluorinert by the 3M corporation

http://www.dsifluids.com/
http://solutions.3m.com/wps/portal/!..._IcFQEAOIxDW4!