water chiller experiences

[nealric]

I was wondering what kind of experiences people have had with a minifridge as a waterchiller. I am interesting in doing this, but I would like to see what people who have done it think of it.

 

[Paxmax]

Aren't mini-fridges made with TECs(peltiers) ?
They aren't all that powerful though, maybe 120-150W TECs are used. The performance will be like that.

I dude stuck a bucket inside his freezer(regular compressor powered), but it never got cold in the long run, since the heat transfer was through air. So, he took the evaporator off the freezer and folded it inside the bucket. That worked pretty good.
I tried to find the link to on overclockers.com, sorry.
/Paxmax

 

[aenigma]

Yeah I have done that with a chest freezer, but the air to evaporator heat transfer isn't good enough.Later after ditching it I made a heat exchanger for the evaporator(copper pipes zip tied to the evap).
Mini fridges usually won't have enough power.Older ones might though...

 

[nealric]

The idea I was having was to have a radiator inside. My resis big (5gal) and it wont be running 24/7 so im hoping that if I can pre chill it to 0c or so at the begiining the fact taht its not that powerful wont be a problem. Btw i wont be useing a pelt powered fridge (i know better than that )

 

[aenigma]

Won't work, its not just the fact that the compressor may not be powerful enough, but the lack of cooling on the liquid.

I had a radiator and copper coils and about 10 gallons of liquid in my chest freezer that I would let prechill, it would warm up very fast after turning on the computer.Oh and not to mention 2 170mm 235cfm comairs circulating air.

 

[johnnyw]

I had already thought about the reason why peolple who uses mini-fridges as the way to cool the water in these watercooled systems, first don´t let the water get colder by using a home made rad ( like a spiral copper ) and after that, yes , take it to the mini-fidge. In that way better results will be achieved.

 

[aenigma]

I had already thought about the reason why peolple who uses mini-fridges as the way to cool the water in these watercooled systems, first don´t let the water get colder by using a home made rad ( like a spiral copper ) and after that, yes , take it to the mini-fidge. In that way better results will be achieved.

What?

 

[nealric]

Im not sure what you mean by lack of cooling on the liquid. Are you saying that heat is not being let out of the water so the water just keeps heating up? Right now I am using a 5 gal res with no rad. Even if i run it forever, it never reaches unsafe temps. I still dont see why it would be heating up too much. Keep in mind that I am not looking for artic temps. I live in houston where humidity is usually in the 90-100% range. If I get it too cold, condensation will be impossible to control. Im looking for 20-25c temps- not subzero.

 

[aenigma]

Im not sure what you mean by lack of cooling on the liquid. Are you saying that heat is not being let out of the water so the water just keeps heating up? Right now I am using a 5 gal res with no rad. Even if i run it forever, it never reaches unsafe temps. I still dont see why it would be heating up too much. Keep in mind that I am not looking for artic temps. I live in houston where humidity is usually in the 90-100% range. If I get it too cold, condensation will be impossible to control. Im looking for 20-25c temps- not subzero.

Ohhh I thought you were shooting for good temps.20-25 isn't cold at all so that will probably work.

 

[tft]

I ran a fridge water chiller unit for about 8 months and had great results. The refridgerant was r-12, and all I did was get a bucket, fill it with 70%-30% anti-freeze/water, and put the evaporator part of the fridge in the bucket. I had the lines going through the side of the fridge through fittings I put in holes I drilled (1/2 intake and output). I got great temps with this, liquid temps dropped around -30 w/ no load, and rose to about -12 w/ load. I accidently covered my temp probe in the socket w/ glue, but the temps I got most of the time, even with it being covered were around 3-4 degrees, so the actual temp must have been lower. The only reason I got rid of it was space considerations, the mini-fridge took up all of my legroom

I think if done correctly, water-chiller cooling can deliver temps that compete with phase change style cooling

 

[aenigma]

Uhm, phase change cooling is the same principal water chillers use.(refrigeration)
I think what you mean was a direct die setup(evaporator on the cpu), and no a water chiller won't compete with a direct die setup.

 

[tft]

no a water chiller won't compete with a direct die setup.

At least from what I've seen, you can get really good temps with a water-chiller, and its much easier to make than phase change. If you have a good enough setup, you can get your liquid down to well below -30. A water chiller won't necessarily be as good as a properly set-up phase-change, but it can get competitive temps within a 5/10 degree celcius range

 

[aenigma]

Again, phase change is REFRIGERATION, there is no differance.
Phase change doesn't mean direct die, it means the change from gas to liquid and vice versa.
And no, water chillers won't compete with a well done direct die setup.Water chillers are still good, but not compared to direct die.

If your making a water chiller, and not just butchering a dehumidifer or something.Then making a direct die system isn't that much harder.

 

[nealric]

I read into direct die and it seems MUCH harder than a chiller. Especially for someone who has next to no tools. Making a block would be out of the question- plus chilled wc is much more failsafe as with direct die- if the compressor fails, you procs toast.
I just got a minifrige today- I will find out how it works as soon as I can go over to my friends house who has a drill press.

 

[aenigma]

I said if your making a water chiller from scratch vs. a direct die system from scratch.They are both pretty easy if you know what your doing.I think you thought I meant a waterchiller using a dehumidifer, window a/c etc. and just dropping the evap in a tub.I was talking about building it from scratch.
I just made a block a couple days ago with no drill press.It is very easy.

I wouldn't have high hopes for this minifridge, most of them can't handle a load.For instance the new ones that use R134a have 1/20hp compressors.If this is an older one, it might work ok.Just don't expect much.

 

[johnnyw]

hmmm , this seems pretty nice but has someone got any idea of how to cool a dual with a water-chiller? Th fact is that I hope to have soon to cellery´s 500 and would like to take them the closer I can to 1 gig each and would like to know how to use a chiller in this case....
I mean, a "T" connector would do the job pretty nicely but take into account that there the heat would be doubled or something similar.
Any advice?

 

[nealric]

Hmm- I think it would be all about how much wattage it could dissapate. I think a cely 500 wouldnt put out much heat compared to my xp1600 @2v- I think it would be about the same load.

 

[johnnyw]

could you give me an explanation?
does this have a reason?
would two 80 watts tec ´s do the job?

 

[nealric]

I think somewhere you can find a wattage calculator- one has been posted a few times in this forum. I know older celys put out a lot less heat (therefore less wattage) than an xp. An 80watt should do the job for an older celly.

 

[aenigma]

http://www.benchtest.com/calc.html