Any real benefit to adding a water chiller to a h20+TEC setup?

[hardwarephreak]

I have (on order) a Maze 4-1, that I am planning on using with a 226W TEC I have been using in another system for awhile. I have been looking around, and I have noticed the VIA Aqua CC25 mentioned a bit.

http://www.aquastealth.com/subcatmfgprod.asp?0=243&1=244&2=-1&6=1

Any real advantages to using this with my setup? I am guessing it will increase my ability to cool the hot side of the TEC, thus allowing the cold side to do a bit better job. Sounds plausible. Granted I wouldn't spend $350 on a chiller.

Now I know I haven't given out much info in regards to my setup, but can anyone give me a generalization as to the difference in temps? I am just trying to evaluate whether it is worth the effort or not.

Thx

[saaya]

the lower the temps the less efficient the peltiers work, just put the waterblock on the cpu and use the tecs to chill the water.

[Jabo]

Originally posted by saaya
the lower the temps the less efficient the peltiers work, just put the waterblock on the cpu and use the tecs to chill the water.

What are you basing your statement at, m8?

I looked everywhere and couldn't find any info of sub ambient/sub zero TEMs performance ratings and efficiency curves.

It'd be great if you could pass some links saya, m8:)

[Jabo]

Originally posted by hardwarephreak
I have (on order) a Maze 4-1, that I am planning on using with a 226W TEC I have been using in another system for awhile. I have been looking around, and I have noticed the VIA Aqua CC25 mentioned a bit.

http://www.aquastealth.com/subcatmfgprod.asp?0=243&1=244&2=-1&6=1

Any real advantages to using this with my setup? I am guessing it will increase my ability to cool the hot side of the TEC, thus allowing the cold side to do a bit better job. Sounds plausible. Granted I wouldn't spend $350 on a chiller.

Now I know I haven't given out much info in regards to my setup, but can anyone give me a generalization as to the difference in temps? I am just trying to evaluate whether it is worth the effort or not.

Thx

From my experience with my system I found that the lower the water temp the better.

Let's say you can keep your coolant in your water loop to +10C you may expect temps of ~-10C to -20C at your cpu idle.
Myself, with ambient air +15 and straight water cooled tec's I am getting -8C idle and +4 full load (done at night with window open )
Hence my estimates (they are only educated guess).

Btw, great find mate!

[saaya]

berkut and teus and some other people told me about it. at -40°C a tec only works with 50% of its cooling power, so a 226W tec can only move 113W max at -40°C, probably 180W max at -10°C

i havent read about it anywhere either, but it seems logical to me that a tec works less efficient the less heat you have it can remove as it generates some heat itself, so it can impossibly remove 100% of all thermal enery and the closer you get to 0°K the less efficient the tecs work.

cooling the tecs with chilled water still works very good though, a 15-25% cooling power loss isnt that much, i was chatting with berkut and some other people about tec cascades and vapocooled tecs. its a huge energy waste to cool tecs with a vapo system because you would need 3prommies !!! to cool 226W tecs to ~-30°C on the hot side to get ~-80°C cpu temps...

thats 1700W of energy you need to get to -80°C with a tec/vapo combo.

thats why berkut is building a compressor cascade system and doesnt use tecs for his suer cooling project.

the lower the temp, the less efficient the tecs work, thats why the industry uses compressor coolings and not tecs whenever they need really low temps.

[hardwarephreak]

i was chatting with berkut and some other people about tec cascades and vapocooled tecs. its a huge energy waste to cool tecs with a vapo system because you would need 3prommies !!! to cool 226W tecs to ~-30°C on the hot side to get ~-80°C cpu temps...

thats 1700W of energy you need to get to -80°C with a tec/vapo combo.


Oh I forgot to mention, I have 2 cryo freezers and a private nuclear power plant at my disposal....all for my one 226W TEC :slobber:

(wakes up):D

So while it does help, the percentages outweigh the costs. Hell I could build a phase-change system for about $250 - $300.

[Jabo]

I guess I have to write to manufacturers to sort this one out :)

P.S. Berkut told me, when asked about, that he's not a TEC guru and he
thinks that TEM (thermoelectric module) loses it's efficiency at sub zero temps.

As far as I am concerned there's no logic in at all unless there's some atomic level physics going on here reducing 'Peltier Effect's' efficiency

[saaya]

heat=atomic movement...

[wolfgang]

I think a chiller with a tech would work well. With 30+c water with my tech its getting 20c load with 2vcore. I am about to go chilled water hopeing to keep the water at ambiant anyway. Then I should stay around 0c. I hope

[Jabo]

Originally posted by saaya
heat=atomic movement...

You see, atoms as such don't move. They are static in their atomic 'nets' forming substances. It's elecrons which move when energy is added by moving to higher energy orbits around proton/neutron nucleus.
This is how superconductivity works in simple terms, get electrons to 'stand still' and current will pass without 'any' interruption :).
In my opinion semmiconductors of which TEMs are built should gain on efficiency if cooled down... But I am not so good at semiconductors theory and Iam not sure about it.
TEM is just a pump moving energy from one side to the other. Enegry isn't cold nor hot ;)

But as I said, I'll drop producers an email asking for full theoretical explanation.

[Jabo]

All right, got it finally, check out this (http://www.kryotherm.ru/tech_inf.htm)
tech specs page.

They are rated to work to -50C

thank you :)