TEC Chiller Project

http://cgi.ebay.com/Heavy-Copper-Hea...ht_1005wt_1165 - Somehow this huge 72lb Cu heatsink got me thinking about TECs. Good lord you could dissipate some major heat with that thing... If only it weren't needlessly 4" tall and $350. Let's see... Cu peaked at around $4 per pound in the past 5 years, and right now it's $2.60/lb. I suppose if it were selling closer to its spot value and shipping weren't another $60-$75 then I'd pick it up.

Anyway: At the 20C delta T range, can TEC chillers match the COP of phase-change equivalents? Assume best case scenario for both setups (better blocks, more TECs, lowest voltage necessary to achieve 20C Delta).

If I can get 20C below ambient coolant temperatures at a COP of above 2.0 and at least equal to that of phase-change, then I'd consider making a setup. I figure I'd have to run a little lower voltage.

Nice setup BTW.

Yes they could, but on the downside it would NOT be cost effective, TECs work much better with multiples. So yes you could get a COP similar to phase change but it would not be cost effective unless you got high Qmax TECs in bulk with a bulk discount (at that point you'd have to be a dealer or something).

Quote:

---

Originally Posted by Ultrasonic2

well the T-bB switches at 820hz

---

Absolutely, so its perfect for the job:)

Quote:

---

Originally Posted by flak-spammer

Yes they could, but on the downside it would NOT be cost effective, TECs work much better with multiples. So yes you could get a COP similar to phase change but it would not be cost effective unless you got high Qmax TECs in bulk with a bulk discount (at that point you'd have to be a dealer or something).

---

It is possible to get Qmax 300W TEC's singularly for $25 and cheaper for multiples without being a dealer. Even with everything ekse you need to buy these are pretty cost effective when compared to phase.

Buying in bulk, assuming I have the idea of bulk correct, would be pointless for a hobbyist when you will, set up correctly, rarely require more than 10. Indeed get too many and you will find it extremely difficult to get rid of the heat coming off the hotsides.

As far as copper goes I never buy new at full price, means you have to be patient and diligent but it pays off. The world price has fallen for sometime but the dealers like us to think it is still ultra expensive.

True enough, what I meant was more directed to people who wanted phase COP and the ability to move a large heatload. At that point then TEC is no longer economical unless you have a way to get a large supply of TEC's cheap. even 10 at $25 is still $250 + the PSU and the cooling adds up to $600-$1000. Small end phase change systems end up being quite a bit cheaper then.

Quote:

---

Originally Posted by flak-spammer

True enough, what I meant was more directed to people who wanted phase COP and the ability to move a large heatload. At that point then TEC is no longer economical unless you have a way to get a large supply of TEC's cheap. even 10 at $25 is still $250 + the PSU and the cooling adds up to $600-$1000. Small end phase change systems end up being quite a bit cheaper then.

---

I think your still going a bit high but I am not going to argue. If you bought 10 TEC's they d be less than $25 etc etc

The problem stems from phase builders tackling commissions as a hobby and not commercially as they should. They rarely charge enough if any for their time and because they treat it as a hobby rather than a commercial contract firstly didn't charge enough and secondly lacked the commitment/drive to see the project through because it was not under contract. In reality a decent phase should not be purchasable for less than $700-800 absolute minimum betwen friends or otherwise !!

Ahh yes. For High heat load I would probably just go phase unless I could get a nice deal on a TEC setup. My interest in TEC-chilling is for the silence and size really.

However I would be using gigantic radiators on the hot-side anyway... (Think 2(1.5x Mora2)) - that might work passively at idle.

Any tips on purchasing copper? As in... where to?

Quote:

---

Originally Posted by Archmage

However I would be using gigantic radiators on the hot-side anyway... (Think 2(1.5x Mora2)) - that might work passively at idle.

---

That's what everyone says but it depends on your proposed TEC. Single large TEC's used at high power can give off a frightening amount of heat off the hotside.

What people often misunderstand with TEC's is that the heat coming off the hotsides can easily be 2.5 times the cooling (or more.).

Often multiple TEC's are planned for mega cooling - one just needs to remember that if the TEC's cool to the level planned there will also be mega heat from the hotsides. If you plan 1000w cooling should your TEC's attain this - you will have 2.5Kw's heat from the hotsides....mega rads indeed !

Quote:

---

Originally Posted by Archmage

Any tips on purchasing copper? As in... where to?

---

If you live near metal workers (try the phone book.) who regularly use copper -just put it about your interested in off cuts (don't forget to tell what thicknesses your interested in.) or try the main metal suppliers on ebay. It will take time to get what you want. Otherwise you pay the inflated dealer prices and get it now. I have been nearly 18 months getting the stuff I will eventually use in my current project but I paid no-where near dealer prices for it.

Archmage - you might find your COP of 2.0 a little tricky to get.

Quote:

---

Originally Posted by Archmage

Anyway: At the 20C delta T range, can TEC chillers match the COP of phase-change equivalents? Assume best case scenario for both setups (better blocks, more TECs, lowest voltage necessary to achieve 20C Delta).

---

What delta you talking about here ? In TEC the delta T is the difference between the TEC hot and coldsides...which I believe is a bit different to what your thinking of. In TEC you need the larger deltas and a low hotside temp to get good cooling But the problem is a TEC delta is affected by multiple specs and one doesn't have as control over it as one would like. For instance lowering the operating voltage and a low hotside temp both lower the delta to varying degrees, reducing the potential cooling. A 20ºC TEC delta will only get you to 20ºC less than the hotside temp which will obviously, in most cases, be greater than the ambient.

To get a COP greater than 1.5 you will generally have to power the TEC at less than 40% max and have a delta LESS than 20ºC.

Quote:

---

Originally Posted by zipdogso

To get a COP greater than 1.5 you will generally have to power the TEC at less than 40% max and have a delta LESS than 20ºC.

---

which is why number of TEC's is very important.

Also which u guys see why im trying to have them in numbers greater the 4.

Gonna salvage this guy. As soon as skinnee sends it back to me, i'll do some testing with larger TEC's but 1 less.

Gonna try to get max CoP.
http://i125.photobucket.com/albums/p...s/ntc22ss7.jpg

Quote:

---

Originally Posted by Archmage

However I would be using gigantic radiators on the hot-side anyway... (Think 2(1.5x Mora2)) - that might work passively at idle.

---

That would depend on how low you want your coolant to be held. Personally, i dont think a typical TEC setup will exceed 600W of power, which means almost 900W of heat on the hotside. If you dont mind a 7-10C delta, thats not hard on a quad rad with acceptable noise.

NaeKuh, how many TECs were on that? 8? Seems like it would have to have some pretty good cooling and very good insulation for it to work properly.

its only 5 at the moment.. but im gonna reduce one to 4 x 50mm.

You want to run them as low as possible for the max QoP. This is the value thats most important along with heat load.