Koolance 1200w Liquid Cooled PSU @ [H]

FrgMstr

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
55,634
Koolance 1200w Liquid Cooled PSU - Liquid cooling your power supply? Koolance goes high tech with high power in what is a very innovative product. We give you an exclusive look at world's first truly submerged liquid cooled power supply at 1.2 kilowatts.

This is an interesting product that has its applications in certain niches however the Koolance 1200W Liquid-Cooled Power Supply is not for most users. As an exercise in the feasibility of using liquid cooling to extract more capacity from a platform the Koolance 1200W Liquid-Cooled Power Supply is a huge success. Our testing proved this extra capacity is captured at the expense of the quality of the unit’s DC output on a single 12v rail.
Click to expand...
 
I love taking on criticism, but cannot. I don't know how to argue a smiley. :D
 
no its just that having all that power at the risk of dirty electricity doesnt seem worth it with more expensive parts that use more power. nothing wrong with the review.
 
ghost6303 said:
no its just that having all that power at the risk of dirty electricity doesnt seem worth it with more expensive parts that use more power.
Click to expand...

Well like I said in the review as a proof of concept it worked very well albeit with limitations such as that and mounting issues. Had Koolance started with a base model that had less noise/ripple it is indeed possible that this method would have resutled in one of the best power supplies we have seen.....and one that mitigated many of the issues involved with cramming more hot running parts such as the new R600 based cards. Hopefully, they will work with CWT and get that issue sorted because he product is indeed interesting.
 
Damn it, Paul!

What? Did you tell Kyle I was currently working on my own Koolance review and had him rush the edit on yours? ;)

Good job as usual! Looks like a nice 850W power supply. :D I, for one, am not taking apart mine as I will like to run it for a while in my own machine to see if it's worth a crap or if the 120MM rear fan is just as loud as the internal 135MM fan I currently have.
 
jonnyGURU said:
Good job as usual! Looks like a nice 850W power supply. :D I, for one, am not taking apart mine as I will like to run it for a while in my own machine to see if it's worth a crap or if the 120MM rear fan is just as loud as the internal 135MM fan I currently have.
Click to expand...

Oh but taking it apart is so much fun........I don't think this unit works any more though.


And in all seriousness....anyone reading the review DO NOT OPEN THIS POWER SUPPLY. I know they all say that, but most can be opened without destroying the unit. That is not the case with this unit for most people so just don't.
 
I didn't see anything related to this in the review, but: Does the psu run the tubes out the back to the radiator, and the whole psu is flooded with liquid? When you took it apart, did you "just" have to drain and then disassemble?
 
Paul_Johnson said:
Oh but taking it apart is so much fun........I don't think this unit works any more though.
Click to expand...

Duh!!

;)

Really... did you kill it prior to or during disassembly or do you think draining it killed it? You don't think refilling it and screwing it back together will "save" it?

Bbq said:
I didn't see anything related to this in the review, but: Does the psu run the tubes out the back to the radiator, and the whole psu is flooded with liquid? When you took it apart, did you "just" have to drain and then disassemble?
Click to expand...

Yes. The tubes run out of the back of the PSU and into the radiator.

EDIT:

Picture here: http://www.jonnyguru.com/PSU/Koolance1200W/DSC01402.JPG

May have to reload because I don't allow deep linking on my site.
 
jonnyGURU said:
Duh!!

;)

Really... did you kill it prior to or during disassembly or do you think draining it killed it? You don't think refilling it and screwing it back together will "save" it?
Click to expand...

I killed it during dissasembly. I probably could reassemble the thing but the evaporative loss of the coolant plus the fact that it could be contaminanted (since i had to gut it at home and well.......there is no Laminar flow hood here) would make me not really want to try and run it.
 
So in essence you're paying $500 for a liquid cooled, overclocked 850W power supply? :eek:

Amazing!

Would love to see some modders here do this for less than half the cost :D
 
Crazy Larry said:
So in essence you're paying $500 for a liquid cooled, overclocked 850W power supply? :eek:

Amazing!

Would love to see some modders here do this for less than half the cost :D
Click to expand...

The CWT it is based on in TT dress is running for over half that price ATM, plus the cost of the coolant, plus the radiator and seperate cooling. It would probably be more expensive. This thing can't be hooked up to a regular cooling loop so it would be a challenge.
 
Cool concept, but I'd have a hard time plunking down that much cash for a PSU that only has 6 pin PCI-e connectors at this point.
 
From the pics it just looks like a PSU with a Koolance Hydra-Pak from a one of there harddrive coolers just sitting on the heatsinks? Well + other stuff for safety and to keep it self contained. If thats the case I'm so modding my PCP&C 1KW-SR the same way. I built an ultra quite all water cooled computer only the 1KW-SR was loud. So I modded it for a 120mm(same CFM) super quite fan. But hell, now I think i'll try water cooling it.


EDIT: Oh wait... I guess I somehow missed that the whole thing is filled. Well scrap that then.
 
I'll prob get flamed for this but liquid cooling is so... :rolleyes:

Gotta love marketing.
 
This is pretty special actually.

Using inert non-conductive liquid to literally make contact with the electrical components is pretty ballsy. This isn't the typical 'water-cooling' were used to seeing on the market.

Now if only we had a ToughPower1200w that koolance liquid-cooled... or better yet, if we had retail case products to cool our mainboards or other overclockable components that way.

I give it a thumbs up for being a good concept. Too bad it had a tough act to follow. It would have gotten more fanfare if it was the first 1200w to the market.
 
Torsten said:
This is pretty special actually.

Using inert non-conductive liquid to literally make contact with the electrical components is pretty ballsy. This isn't the typical 'water-cooling' were used to seeing on the market.

Now if only we had a ToughPower1200w that koolance liquid-cooled... or better yet, if we had retail case products to cool our mainboards or other overclockable components that way.

I give it a thumbs up for being a good concept. Too bad it had a tough act to follow. It would have gotten more fanfare if it was the first 1200w to the market.
Click to expand...

good idea. they should make a case with one side of it is a huge but totally efficient and quiet power supply with water cooling for the rest of the system.
 
The design certainly is ingenious, but there are some definite drawbacks. To put that sucker into any case I have, it would have to be supplied at the bottom of one of my rigs here. However, that particular case has an isolated wind tunnel for the power supply- in effect doing the same thing this unit was designed to do, just minus the liquid.

It does beg a second question of me- would I need a 1200w power unit? My computer barely pushes 160 at load (and my entire apartment barely pushes 160 at load when my computer is turned off ;) )
 
haysupimark said:
The design certainly is ingenious
Click to expand...
What are you talking about? Several people have done this with baby oil long in the past. The idea isn't new, and I'd argue that it's far from ingenious - interesting yes, ingenious no. You pay $500 for an 850W PSU with a 2-year warranty (which is probably the expected useful life of the non-conductive coolant). Plus, now you have a 120mm fan mounted outside your case making noise.

The first company to take the Channel Well PUC1200V-A01 and mount waterblocks that can be used with an external loop (ideally the 140mm fan could stay so that water would be optional) will earn my business.

Maybe some modders here would appreciate this idea.
 
Nice Review!

I was hoping you guys would do this power supply soon. I have just one thing I would like to point out:

When I was reading the review it did not state anywhere that you disassembled it AFTER testing. The whole time I was reading the review I was wondering if the disassembly had somehow affected the test results. It was not until I read this thread that it was mentioned that you took apart after the testing was done.

Other than that good job.
 
haysupimark said:
It does beg a second question of me- would I need a 1200w power unit? My computer barely pushes 160 at load (and my entire apartment barely pushes 160 at load when my computer is turned off ;) )
Click to expand...
pfft

you are definitely the target market for this PSU. You should buy one now :D
 
Replicant1911 said:
Nice Review!

I was hoping you guys would do this power supply soon. I have just one thing I would like to point out:

When I was reading the review it did not state anywhere that you disassembled it AFTER testing. The whole time I was reading the review I was wondering if the disassembly had somehow affected the test results. It was not until I read this thread that it was mentioned that you took apart after the testing was done.

Other than that good job.
Click to expand...

For all the reviews the units are dissected after the testing (which is why on some units you can see blown fuses, burnt PCB's, burnt wiring, etc) for that very reason that opening one beforehand introduces confounding factors. The articles are written dissection first though because it seems to make more intuitive sense to see what you are going to be working with before you work with it in a review.
 
Paul_Johnson said:
For all the reviews the units are dissected after the testing (which is why on some units you can see blown fuses, burnt PCB's, burnt wiring, etc) for that very reason that opening one beforehand introduces confounding factors. The articles are written dissection first though because it seems to make more intuitive sense to see what you are going to be working with before you work with it in a review.
Click to expand...

Ah that makes sense and I agree with you. The reader usually wants to take a look at the "guts" of a component before putting it through its paces. I am sure very few people were confused by that but you might want to consider throwing in that little mention for dense people like myself :p .
 
Nicepants42 said:
What are you talking about? Several people have done this with baby oil long in the past. The idea isn't new, and I'd argue that it's far from ingenious - interesting yes, ingenious no. You pay $500 for an 850W PSU with a 2-year warranty (which is probably the expected useful life of the non-conductive coolant). Plus, now you have a 120mm fan mounted outside your case making noise.

The first company to take the Channel Well PUC1200V-A01 and mount waterblocks that can be used with an external loop (ideally the 140mm fan could stay so that water would be optional) will earn my business.

Maybe some modders here would appreciate this idea.
Click to expand...

Your correct that the idea is not new to super computers and the modding community, but it is new to the consumer retail market.

A water cooled PSU would be sweet, but I think the nature of PSU design would make this very difficult. Since they are composed of many electrical components that produce heat, rather than a single IC with a flat surface that produces heat; wicking this heat away with a waterblock would be difficult at best. Perhaps if the largest heat producers within the PSU were watercooled, then the rest could be air clooled with little to no noise. Then again engineers are known to be ingenious, so I dare not say a 100% water cooling solution is not possible.

*EDIT* oops, missed your last link.
 
Torsten said:
This is pretty special actually.

Using inert non-conductive liquid to literally make contact with the electrical components is pretty ballsy. This isn't the typical 'water-cooling' were used to seeing on the market.
Click to expand...

not really, power companies have been using mineral oil filled transformers fo years
(all those pole mounted transformers are oil filled)
 
Very interesting from an engineering/reliability point of view. Don't quite know what to make of it. If the noise/ripple had been better (less) I would not be wondering how hard they have pushed the design and compensated with cooling and to what extreme is that a valid strategy for a product sold to the masses. Not that that is wrong, see my sig, but what if Intel took a E6300 (1.8GHz) slapped a decent HS/fan on it (as hundreds of us have) and called it an E6399 and rated it at 2.8GHz and sold it for 40% more? Perhaps Intel should have done so. Have to think about this one and that is a rare treat these days. Good article.

To me the cooling was more interesting than the supply itself. Wish there had been more about the pump and rad. Maybe temps of air in and out of rad but with all the air blowing around to dissipate the heat being generated by the loads it probally would not have been very accurate anyway. Still - Good Job !!!

edit hmm, was the coolant identified ? I will have to go back and see, wondering about the long term effects on the materials used in construction. Starting to think a 2 year warranty is not nearly enough given the cost.
 
BillParrish said:
To me the cooling was more interesting than the supply itself. Wish there had been more about the pump and rad. Maybe temps of air in and out of rad but with all the air blowing around to dissipate the heat being generated by the loads it probally would not have been very accurate anyway. Still - Good Job !!!
Click to expand...

I actually toyed with the idea and tried to get some readings but the temp sensor that was sandwihed between the incubator and the radiator kept giving temperature spikes because of the new mounting plate I needed for the setup. I will say this though. The unit did retain heat much longer than a conventional PSU as when I removed the unit form the incuabtor after it had completed testing it was very very warm and after sitting for several hours was still warm to the touch unlike air cooled units.

edit hmm, was the coolant identified ? I will have to go back and see, wondering about the long term effects on the materials used in construction. Starting to think a 2 year warranty is not nearly enough given the cost.
Click to expand...

It is a type of inert perflourocarbon. Perfluorohexane has been suggested as the exact formulation of the perflourocarbon but as it stands I haven't got that pinned down exactly.
 
Paul_Johnson said:
The unit did retain heat much longer than a conventional PSU as when I removed the unit form the incuabtor after it had completed testing it was very very warm and after sitting for several hours was still warm to the touch unlike air cooled units.
Click to expand...

I noticed this too. I put my temp probe on the housing and it was 40C in a 25C room after running it "open air" for the cold tests. After 1/2 hour, the housing was still around 35C.

I think if any power supply needs some sort of "After Spin" it's this one. The pump and radiator fan should run for a few after power down.
 
jonnyGURU said:
I think if any power supply needs some sort of "After Spin" it's this one. The pump and radiator fan should run for a few after power down.
Click to expand...

Yeah I still love that feature on the power supplies i have that have it. Wish more companies would license it.
 
Yeah... but for $1.50 per unit license fee to Zippy... eh....

With newer PSU's it doesn't seem to be so bad because they don't get as hot. But it is still nice since it does give the PSU a chance to cool down.

When I wrap up a review, I bring the load down from test 10 to test 1 or 2 to "cool down" the PSU AND the load tester. I then shut down the PSU and then don't shut down the load tester for another 30 minutes afterwards.
 
jonnyGURU said:
When I wrap up a review, I bring the load down from test 10 to test 1 or 2 to "cool down" the PSU AND the load tester. I then shut down the PSU and then don't shut down the load tester for another 30 minutes afterwards.
Click to expand...

I just cut off the PSU they are cheap....the load tester runs for 30 minutes or until I go to bed because it isn't cheap :)
 
Interesting to resolve the fluid that they used. Its boiling point would help reveal what perfluorocarbon it is.

Paul, would it have been possible to get intake and exhaust temps of the radiator with your probes, even if it was outside the incubator? At the 100V load test 4 that thing would have been putting out about 300W at equilibrium, and that's a lot even for that rad.
 
drlava said:
Paul, would it have been possible to get intake and exhaust temps of the radiator with your probes, even if it was outside the incubator? At the 100V load test 4 that thing would have been putting out about 300W at equilibrium, and that's a lot even for that rad.
Click to expand...

It may be. If I have some time this weekend I'll see what I can do about setting something up.

Also, jonnyGURU has a different setup and may not have run into the issues I did with that.
 
You must log in or register to reply here.
Back
Top