Slashdot Mirror
Considering Watercooling Your PC?
An anonymous reader writes "Thinking of taking the plunge into water cooling your PC? These guys have rounded up three systems ranging from cheap and cheerful, to stylish and pricey."
← Back to Stories (view on slashdot.org)
Water has excellent heat transfer properties, (better than the refrigrants in your fridge), is easy to handle, unlike some of the better heat transfer fluids such as liquid metals, and is non toxic.
If you want to immerse your computer, Flourinert has been around forever, though now probably banned.
I have some experience with watercooling. With proper care, it's safe and a quiet way to cool your machine. For those of you who move your computers around though, becareful what materials you use. I built my computer in a warm dorm room which meant that my copper block to plastic piping worked fine. Then I took it home to my freezing basement and water went everywhere. I think the metal shrank while the plastic didn't, and water came out of the connections.
Other than that I never had any problems. I don't use it anymore because it's too heavy to carry around all of the time.
--------
It's OK to be social, just don't tell anyone about it.
Damn slashdotted already. Which systems did they compare ? I have a corsair watercooling system and it does a great job with cooling and a breeze to setup too. The only negatives are its price (~200$) and noise. But the noise is not all that bad. Just a sort of humming sound of the fan. But its a bit more silent that all the case fans and cpu fans I used to have.
Water is a much better coolant than your typical refrigerant. The reason water is not used, however, is because its boiling point is far too high.
I suspect, it would be better then if you had a fan failure. Remember that unless its underpresser water won't heat beyond 212F unless you put enough energy in to vaporize. It also becomes pretty vicus as it becomes warm, so convection will cause it to circulate to some degree, as will very slow boil where its not so fast that a layer of vapor separates the water from its contact with the cpu. So basicly very worst case senerio your cpu hits 212F, water has a very high specific heat, so if you have a decent sized radiator and a resonably large fluid volume, I would be really suprised if it got hotter then that.
You would probably be better off..
the heat transfer equation H=h*a*(delta T)
H=heat
h=heat transfer coefficient
a=surface area available for cooling
delta T=diff between temperature of device to be cooled and surrounding cooling fluid
shows that the easiest way to cool something is to reduce the temperature of the fluid that cools it..
If you lower the air temp in your computer case by 10c, the processer temp drops by 10c, assuming the fans all stay at the same speed.
Increasing "a" is limited by fin efficiency (which is what these water cooling systems are trying to get around, but a sealed evaporator/condenser would be smaller and more efficient, there is a metric buttload of patents now on sealed passive boilers/condensers), and as air speed increases, "h" rises less and less in proportion)
If you want more info, look at the free download of the heat transfer textbook I list in my journal.
From the second page you listed:
;-)
The best candidate seems to be an eutectic solution of sodium and potassium, (NaK). The melting point is as low as -12 C. Its density and viscosity are similar to water but it has a lower specific heat and a much higher thermal conductivity. NaK can be used with nickel, chrome and steel but it is aggressive to cadmium, antimony, bismuth, copper, lead, silicon, tin, and magnesium. It also reacts violently with air and water. It is apparent that this alloy is associated with several material and handling problems. Liquid sodium has nevertheless been used as a coolant for nuclear reactors, which shows that these drawbacks can be managed.
Sure, but do we want to manage flammable liquids that combust when exposed to air in our home? Pumped through our computer? Not to mention that it eats away at silicon and most likely the PCB itself.
I can see why no one considers this.
HBI's Law: Frequency of calling others Nazis is directly correlated with the likelihood of the accuser being Communist.
Last time I tried water cooling my computer, the pipe started leaking, there was a short circuit and I accidentally set my computer on fire, which needless to say was neither cooling nor cool... After that accident I gave up altogether and do you know what I did? Instead of overclocking my CPUs, I started to underclock them. I noticed that in many cases even a 15-20% lower c;ock speed may eliminate the need of having any fan at all, as long as there is a large radiator with good contact and a reasonable air flow in the case. Sacrificing those few percents of megahertzes might sound very "not elite" but guess what? It still can display websites faster than I can read.
Sincerely,
Pan Tarhei Hosé, PhD.
"Homo sum et cogito ergo odi profanum vulgus et libido."
"cheap & cheerful" is a common British phrase which basically means "not expensive or fancy but it does the job".
---- Den ene knappen er powerknapp, den andre er Bender voice knapp "Bite My Shiny Metal Ass"
The Peltiers we tested in the lab needed the same amount of electrical power as the amount of heat they moved..So much better than 5 %, but it doubles the thermal load on your heatsink.
I think the parent poster posted this: "The melting point is as low as -12 C." Correct me if I am wrong. But that is a tad bit cooler than room tempeture let alone the running tempture of a CP and GP U's It's really not a bad idea.
The EER rating on air conditioners (a common heat pump) tell you the ratio of heat moved to power expended to move it. The units of EER are messed up though, it is BTU/HOUR divided by Watts, multiplied by some factor of 10.
First of all unless you're using some seriously exotic components it's going to be solid at the temps you'd deal with in an average compouter (15-75c).
Secondly the materials needed to contain your liquid metal aren't cheap.
Thirdly it makes inspection for any potential flaws a real pain in the ass.
Fourthly you have to consider the viscosity of liquid metals vs. water or other coolants. The amount of energy you have to use to move them is going to be substantially higher.
Fifthly, the conductivity of most liquid metals vs. water (even water with electrolytes) means that most leaks are going to = dead computer.
Oh yeah. . .also. . .I totally forgot about toxicity issues, reactivity (sodium and potassium as you've selected would be really fun to have leak out of your computer and hit, say, a patch of wet floor). . .
Find out about my new childrens book: SS Death Camp Criminal Batallion Go To Monte Carlo For The Massacre
Corrosion should not happen. Not in the lifetime of your system anyways. Distilled water + anti-corrosion type additive should see to that in closed loop system.
These systems must have UL approval right? If so, I don't think your insurance company would be able to say jack if you didn't just grossly mis-install it. But, since it involves water + electronics, I wonder about requirements of a GFCI circuit.
Vapor-phase cooling can cool a CPU below the ambient temperature, and it's not "5% efficent".
There is a 12V Vapochill system that requires only 6.6A, and it is quite effective at cooling even the P4 Prescott (not nearly as effective as the AC versions of the same product, though).
A sealed chamber with only water and water vapor in it (all air and other non-condensible gasses have been removed) will boil water at the hot end and recondense it at the cold end, at any temp above freezing for the water.
If the chamber is a vertical tube with the water and the heat source at the bottom, and fins and cooling air at the top end, the vapor from the hot end will recondense at the top cold end and run back down. (It's really a heat pipe without the porous media to move the condesed fluid back to the hot side.)
I had the same concerns alot of people on here seem to have about watercooling. I shelled out around $900 for a Koolance case a little over 3 years ago and popped in a Intel MB w/ P4 1.7. I also have my 6 300GB HDs and GF FX Video card water cooled as well. I have been in the case quite a few times, and even upgraded it to a P4 3Ghz, new Intel MB, and new Vid card. I even had to replace my old socket 423 cooler for a socket 478 cooler and no probs. To this day (3+ years later) I have NEVER had a single problem with my case. No leaks, no overheating, no problems period. Yes, it only cools as much as the ambient temp in the room, but on a really hot day that is only 100degrees F. As I type this I am running at 92f.. during intense gaming (ie. Doom 3) The temp never exceeds 110f on the CPU (which is where the temp probes are). I have never been able to get my heatsink/fan CPUs to cool nearly this good under intense loads. Just thought I would share my personal experience with watercooling.
Introduction
Despite a rather slow and shaky start, the water cooling revolution is well and truly under way. If the falling component prices aren't testament enough to this fact, the sheer number of kits being touted by a plethora of vendors surely is.
Nobody has ever doubted the advantages of water cooling as compared to air. Water is some twenty five times more efficient than air at conducting heat, which makes it an obvious choice for cooling all manner of hot running computer components. Unfortunately it also comes with several disadvantages too.
Perhaps the biggest hurdle water cooling has had to overcome is the natural fear in all of us that comes from pumping a fluid around the insides of an expensive, electrically powered computer. Common sense tells us we just shouldn't be doing it. Then there are the other dissuading factors such as high cost, tricky installation, bulk, weight, reliability and aesthetics - all of which have conspired to make water-cooling a fringe activity enjoyed by the elite and the brave.
However, things are changing. Water cooling is getting cheaper, safer, increasingly compact and more aesthetically pleasing with every passing day. The whole premise of liquid cooling your PC is now more viable than ever before, and with the trend for ever-increasing cooling requirements showing no signs of abating, it might be a case of when rather than if you make the switch to the wet stuff.
To help you make the plunge, we've decided to take a closer look at three different approaches to water-cooling, each theoretically suited to a different level of experience. Whether you're a LAN gamer who demands ease of transportation, an overclocker who needs top-notch performance, or even a case modder who values good looks as highly as good performance, there's something here for everyone.
Before we get stuck in, let's introduce the three kits on test: Asetek's WaterChill KT03A-L30, Eastar's Cool River Deluxe Version, and Koolance's Exos-Al.
--= I have the rest of the article copied if anyone wants me to post the rest =--
If you want to immerse your computer, Flourient [sic] has been around forever, though now probably banned.
Actually, it's still quite legal, albeit very expensive. We buy buckets of the stuff for our MRI scans - MRI detects hydrogen, and since Fluorinert has none, it's a perfect background substance when scanning tissue samples.
Didn't even know you computer geeks used it for cooling/cleaning 'til I started reading slashdot.
Flooding with coolant won't hurt unless it floods while power is applied or the coolant is coorosive. Lots of electronic repair places put circuit boards in a dishwasher to clean them up. Just make sure that they're completely dry before you apply power.
I wanted to chime in here. My friend and I put together HIS Zalman Reserator (not mine -- got no extra money for these things) and Antec Aria SFF PC a few weeks ago with tremendous results. The Zalman Reserator retails for $250 at Frys. Most online vendors charge more for it.
It's basically a 2.5' tall heatsink/radiator with a submerged pump. It includes a waterblock for your processor (Intel and AMD) and all the tubing/hardware you need.
You lose the ability to easily bleed the thing, although clamping off hoses and pouring nearly 3 liters of water out the top of a Reserator doesn't really seem all that troublesome to me.
Bottom line -- he dropped his idle temps by 20C and his load temps by a similar amount. This was a few weeks ago when it was slightly warmer here in Georgia. He now idles (running a P4 Prescott) at about 27C. It's pretty amazing.
For more information, see here:
http://www.hardforum.com/showthread.php?t=783557
IronChefMorimoto
I refrigerator keeps things (like food) that do not generate heat cool. If you put a constant heat source (like a computer) in a refrigerator it will heat up the refrigerator much faster than the compressor can cool it. You just invernted the oven.
Except not. Otherwise, a refrigerator would only be able to get food as cool as room temperature. :)
Go learn some physics
My blog. Good stuff (when I remember to update it). Read it.
you could take a look at my site
http://www.surfbaud.co.uk/news.php
which details experiments with peltiers and water cooling.
For those who can't be bother it boils down to this.
1/ water cooling is OK if done properly, but all the commercial "home" watercooling products are absolute shite
2/ peltiers work, but add huge heat loads to the system overall, and should never be directly interfaced to silicon.
Hopefully in the next few weeks (having just bought a vacuum pump) I can get around to playing with large scale home made heat pipes and report on those.
HTH etc.
http://slashdot.org/~GuyFawkes/journal
I've done similar to an old 486-SX board. Except first, I submerged it in water-free antifreeze. It was an experiment to see if the board could run while submerged in glycol. Answer: No.
We pulled the board out of the tub, tossed it in the sink and sprayed it off. Took a blowdrier to it for 10 or 15 minutes, put it back in the case, and it fired right up.
And it was shiny-shiny after the glycol treatment :-)
Water can only cool to the ambient temperature of the room, and not below as some people seem to think
That is true, iff you are running a closed loop water system with a typical radiator as the water to air heat exchanger.
Water will cool below ambient if you do an evaporative cooling system. See "cooling tower".
sdb
fan is still a fan, and my 120mm fans even on low rpms are loud...especially when placed outside of the case. but another thing to consider is the pump...that's a small fan or fan wheel that's at a VERY high rpm in order to move the appropriate ammount of water through the system.
I disagree.
I run a watercooled system, and my homemade case has mesh sides which do nothing to stop the noise. From my personal experience (as opposed to speculation):
* My Ehiem pump is virtually silent (seriously)
* The 2 x 120mm AC fans are quite quiet, and have a low frequency wooshing sound which is gentle on the ears.
* The cooling system is quieter than the HDD
* The whole system is quieter than the 60mm stock fan on my wife's PC.