Exploring Below Ambient Water Cooling

Page 9 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar


What exactly are you thinking, what do you suggest?
 
The only cheaper DIY solution I can suggest is a chiller built from an A/C unit or fridge/freezer. It would take some work, but then just run the cooling coils into your cooler and viola!

Otherwise, I like the idea of the manufactured chiller.
 



I agree, I do not want to get into making a unit from scratch, and have no intentions of exceeding 500 bucks to get an aquarium chiller, so if it will exceed that amount of money to do the job, then I'll stick to the ice jugs.






Oh I'm waiting on you to get that chill box Moto, and put it into operation to see how it's going to perform?

That is a cheaper option, but will it do the job?
 
Meh, easier/cheaper/less chance of popping a pipe and breathing in whatever the gas is in them these days :)
it could be done, but I for one Cba with the hassle/mess/expenditure hehe
@Ryan, should be buying it on next months payslip (28th) and then I'll get to drilling some holes in it and seeing how its made :)
Moto
 
Looks like a direct contact TEC over a 4-pass (or more) exchanger? I'm wondering if you couldn't build something like this based on your cooling needs. The only thing I'm wondering about is the amount of power draw required and overall cooling ability in watts. Looks like the one that is a 200w capacity is ~$600, although is the concept of 200w @ 0C temp difference a different concept of heat dissipation than what I'm thinking? Does that mean it has the ability to dissipate that heat load without any change in loop delta? I'm reading through the listed documentation on these to see what their capabilities are...Quite an interesting find, Ryan...
 
Yeah and the 600 price for the 200w model does not include the temperature sensor control or power supply, I'm trying to determine if a standard computer power supply could run it as I have that on hand.

Water pump wise I could split the output flow from the MCP 655 and let it feed both the heat exchanger and the CPU at the same time, cooling the CPU with one branch and circulating through the heat exchanger with the other branch.

Under those circumstances it would alway have chilled water running through the heat exchanger so it's cooling job would be lessened, In my brain theory that is! :lol:

Or use the return from the CPU directly into the heat exchanger to cool before returning to the cooler, that would probably be the best option.

Duh for me! :)
 
Should it take the 200W unit to do the job, (Speculation?) most are thinking of the 130w heat output but actually the water flowing through the water block on the return side is still cool, the CPU did not transfer 130w of heat to the chilled water.

So does whatever chiller used have to even equal the 130w range, a lower powered unit may very well do the job???

What I would be satisfied with is maintaining a 10C water temperature in the cooler, that is 50F and above the condensation formation range, it's almost a perfect running water temperature.
 
Should it take the 200W unit to do the job, (Speculation?) most are thinking of the 130w heat output but actually the water flowing through the water block on the return side is still cool, the CPU did not transfer 130w of heat to the chilled water.
Click to expand...
It's likely this might be a bit lower, but water has a very high specific heat and would need a significant amount of watts to raise the overall temp of the total volume by 1C. While it would still be absorbing and moving this much, the temp wouldn't change much.

So does whatever chiller used have to even equal the 130w range, a lower powered unit may very well do the job???
Click to expand...

This was my speculation as well...
 


Ryan,

I would think that a solution like that would enable you to ditch the cooler all together or shrink it. I do get the ice bucket method, but if dropping this kind of dough, why not just go tec and be done with it. Apologies if I missed/forgot the overarching goal from earlier in the thread, but at 500 we could be looking at a internal near 0 or even sub-0 solution that takes way less energy and maintenance to maintain.

CJ

Edit...The aquarium chiller would also probably allow you to run closed loop as well now that I think about it...aside from the aquarium drop in coolers, which may be a bit more cost effective...google aquarium drop in coolers.
 


Hey Ryan

Note that only the lowest rated cooler is operated with 12V PSU. All of the other ones require 24V PSU.

My thoughts are that if you use a lower rated cooler to lower your temps in your ice chest to say 10C then use as you normally do the thermal mass might just carry you all day. Remember that once you set the temp at which the cooler operates ( all the time while using PC i.e.) that it will be still contributing to your water mass ambient even if it can not keep it at 10C.

So...., if your set-up will run all day on 1 jug of ice at a 10C ambient before it reaches say.., 18C, then you only need the same cooling power that 1 jug of ice provides.

Sorry I don't have time to reseach what that would be in terms of calories but rubix can probably come up with a really close estiment for you.

Good luck my friend.
 
That was some serious storms last night a lot of cloud to ground lightning more than usual, I think the heat wave contributes to some very serious storms, thankfully we only lost power!

Now what were we talking about? :lol:

Ry
 
Here"s a thought to share of a possible experiment.

I already have tank temperature monitoring.

http://www.frozencpu.com/products/2408/exp-01/245W_Potted_Peltier.html

I'm considering adding a Y in my pump output line one branch continues directly to the CPU the other branch cycles through a Swiftech Apogee XT water block with the peltier, cold side to water block hot side cooled by a heat pipe heat sink and cooling fan, powered by it's own power supply.

That loop would be independent of powering the computer, even though when the computer is powered up water will be circulating through it, because the water in the insulated tank will always be colder than ambient room temperature, so the cpu will always be receiving cold water.

The peltier loop will be cooling the insulated tank water.

What do you guys think?

I already have everything except the peltier, so it's not that big a money risk to conduct the experiment.

Here are my concerns the peltier measures 50mm x 50mm, it's a perfect fit for the Swiftech block, but very few heat pipe heat sinks have that large of a base?

Thoughts?
 
honestly a chiller is not the complicated of a system.... wait for it.... in therory lol. you prolly know how they work but for those who dont the run down is:

compressor=> heat exchanger(rad)=> throttling valve => medium being cooled

you compress the liquid in which case it heats up, then running it through the radiator cools it down but still under pressure then when running it through the throttling valve pressure drops and so does the temp of the water.


EDIT: I missed Ryans post COMPLETELY. ignore post.
 
The peltier is ordered I am going to conduct the experiment I listed above, it will be powered by a Silverstone 1000w power supply, it will be a chiller branch as listed above, looping from and returning to the insulated cooler.

So disregarding all speculation I'm going to find out what actually will happen! Ry
 


We got slammed here in Atlanta GA Ryan...from what part of the world do you hail???
 
I'm not sure the return is going to be good enough for the investment Ry, but I've never been one to discount a theory or stop someone from trying something new, the splitting loop seems like sound logic and I like the 'always on' part too, gently gurgling water is a very relaxing noise :)
**Edit, unless its coming through the roof 😛**
Moto
 


South Carolina, I think we shared some of the same storms, I looked at the radar and it was tracking diagonally across both states.

We dodged the bullet with that one we only lost power, they lost trees 10 miles away from us.
 
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom