My Experiance with an aquarium chiller
 

Well I have worked my way through my second aquarium chiller! The Via Aqua CC25. After some tinkering I managed to bypass the thermostat and set the unit to stay on all the time. It ended up achieving a max water temp of aproximately 0 degrees C as measured before the intake of the pump. With insulation and everything installed I was able to hit 20 degrees C at the socket according to the onboard temp sensor. I put insulation over the center of the socket so the accuracy of this temp probe may not be so good. I also forgot to install my own temp sensor and now have to much gunk on the socket to put one in. This unit is only capable of 1500 BTU's. With that information it is obvious why the unit doesn't provide additonal cooling. I think my next plan will be to use an Air Conditioning unit rated at 5,000 btu's. :D

Out of curiousity, how does a temp difference of 20 degrees between the water and the socket look. Do other people tend to have closer values or is a 20 degree diference average. Also, should the pump be insulated? It is the only exposed component of the system. I couldn't figure out if it was better to keep the cold in and the pump heat in, or let the cold out and some of the pump heat as well.

System Specs:

Maze 4 WB
Swifty MCP600 pump
1500 BTU chiller
ASUS Deluxe E Mobo NF2
Barton 2500 Chip
3/8" tubing with 1/2" insulation.

I'm new to the forums here, but reading a lot lately. I am curious, at what temp on the chiller did you get condensation?

Would it be beneficial to just stay, say 10C below ambient to avoid the condensation problems?

Seeing all the work you did, and successfully, I am interested in your thoughts on the Via chiller unit.

How are noise levels on it? etc.

No way to guess your flowrate through the chiller... but yes a better block will yeild better performance. http://www.procooling.com/html/pro_testing.php is your friend.

I know about the comparison graph, its very handy.

My goals in my current system were quite a bit different than my questions about the ViaAqua chiller unit.

Is it loud? Does it heat up your room? What OC did you manage on your mobile chip?

Is there any kind of guess at a MTBF for that thing if its on 24/7?

Or is this a bad idea on a system thats on all the time?

So far I have more or less got the chiller working at 10C and below. This is of course on only the CPU (no northbridge or GPU for now) The chiller has been running great for the past 4 months. I have even left town for a few weeks (leaving it in the hands of my mac using roomate). It stays on all the time and so far so good. But, as you mention this thing isn't made to run 24/7 (w/o temp controller at least). I anticipate that with this type of use it may only last 1 to 2 years. heh, hoping for two. At the moment I am building a new case that will include and AC chiller. The heatwave in Seattle has let up and my AC unit is now in pieces and ready for the big time!!

Definately consider a window AC unit. They're much cheaper, several times more powerful and designed to run for extended periods of time. And since they're not TEC based, they'll use less power.

So much for Ohm's Law!!!

Out of curiosity (because I could never spend that much), is it really worth the $300+, instead of spending $400 and getting a used Prometia?

Ohm's law relates the voltage to current and resistance in an ideal circuit. Since we're talking about heat engines, the relevent law is Carnot's. Or was that not aimed at me :confused:

Well a basic chiller will be a lot cheaper then 300 dollars, although unless you regas it properly it won't cool as well. You could convert an existing watercooled setup to a chiller for less then a hundred bucks, but you're not likely to get much below -15 or -20 without a regas (which can be done for virtually nothing but requires a torch + some idea what you're doing). Not to mention a prometia really isn't worth it without a regas anyway IMO . . .

My bad,,, I assummed you were talking about cost of operation.

Let me clarify my statement,
Even 220 watt TEC's consume much less power than 5000btu AC's. :p

5000 BTU/hr = about 220w, so no they don't. Remember phase change is 5 to 10 times as effcient as solid state cooling.

For reference:
1 BTU/hr = 0.293W
Btu to Kilowatt-hours Btu x 0.000293
5000btu = 1.465kwh
Watt-hours to Btu Watt-hours x 3.414
220watt-hours=751.08 Btu
6 220 watt pelters would equal almost 5000btu consuming 1320watts

You're confuseing power consumed with cooling power delievered. Just because work/time and heat/time have the same units does not mean they are interchangeable. Thus you cannot calculate cooling power by measureing power consumed. They're completely different things.

Heres a site explaining how the second law of thermodynamics works:

http://www.phy.duke.edu/~rgb/Class/p...51/node63.html

I want to see your TECs! Tell me how they work and I'll give you the first million I make on licenseing. :)

I guess my knowledge in this area is severly lacking indeed! I shall continue to lurk untill I'm better qualified to interject.