Slashdot is powered by your submissions, so send in your scoop


Forgot your password?

Do-it-yourself CPU Water Cooler 207

Foss writes "This article on shows a (very) cheap and effective way of getting that usually-expensive water cooling system that many of us have thought about. There are some pretty pictures too :)"
This discussion has been archived. No new comments can be posted.

Do-it-yourself CPU Water Cooler

Comments Filter:
  • by doooras ( 543177 ) on Wednesday April 10, 2002 @11:39AM (#3316798)
    just cover your hardware in saran wrap and dump ice in the box. works like a charm to keep it cool.
    • by Anonymous Coward
      Just seal up the case with silicone and fill it with PCB. You should be able to get PCB for nothing at a toxic waste dump. No pumps or fans required, the oil will circulate in the case through convection. This is quiet and safe, since PCB doesn't burn and is pretty much non-toxic. It does cause cancer in rats though, but rats are sensitive creatures and can get cancer from anything...
  • Wow (Score:1, Flamebait)

    by spike hay ( 534165 )
    I got a water cooling sytem. OCed my 2 gig to 2400.

  • This looks reliable! ct ures/finaltie.jpg

    Also, look how dusty the case is. This guy must live in a barn!
    • Also, look how dusty the case is Case? I didn't happen to see a case. But who cares. I'm sick of my computer components being all crowded together in a little black box.
    • I wouldn't trust that set up as far as I could throw it; and yes, that includes having the monitor attached to it when it's thrown! ;0
    • I liked the multiple attempts at getting a decent seal from the hoses...the author must not have heard of hose clamps. Also, some baffles inside the block to direct flow ought to make it more effective.
    • Yeah, pretty grim.

      For the benefit of that guy, if he's reading...

      1) Most of your water will go straight in and straight out again without doing much cooling. A heatsink should force the water to go all the way round the heatsink - try adding some baffles inside the box to improve that.

      2) Have you never heard of hose clips? 50c each and 100% reliable at clamping off flexible hoses without leakage.

      3) Get some heatsink compound between the heatsink and the processor. Without it, a fair chunk of the heat is never even getting to the heatsink, so the whole point of having a nice efficient heatsink is wasted.

  • by hij ( 552932 ) on Wednesday April 10, 2002 @11:41AM (#3316816) Homepage
    use a double knot to stop the string coming undone and use only cotton string!

    These are the sorts of instructions I don't like to see in a mod! At least it doesn't mention chewing gum...

  • by SimplyCosmic ( 15296 ) on Wednesday April 10, 2002 @11:43AM (#3316830) Homepage

    Sure, for the ten minutes that the article mentions that the pump runs before overheating.

  • a way to get that AthlonXP to overclock. I wouldn't pay for one of those alrady made jobs. I would rather do it myself. At least someone has given a direction.

    Next stop, fried CPU....
  • ...but the rubber bands on the tubes just scares the hell out of me. Did this guy graduate from the Russian Navy or something? I think the shrink tubing would have worked if he'd have gotten the kind with the mastick (glue type substance) in it. That would have sealed it off great and it's still inexpensive enough not to break him.
    • I want to know why the guy didn't just use hose clamps to attach the plactic tubing? Gee, it works in your car for fuel, coolant, brake fluid, etc. Or was that too easy? Doesn't say much for the rest of his ideas.
  • by SGDarkKnight ( 253157 ) on Wednesday April 10, 2002 @11:44AM (#3316841)
    was at the London Smackdown [] tournament that I went to. This guy had to carry around an extra cooling pouch with all the stuff built into it. There are 3 pics of it here [], here [], and here []. The third pic is the best view of the pouch and the first two show you the in'erds on the computer.
    • Dude, that is WAY hardcore. I like the design. It would have been nice had they taken a picture of the innards of that little black bag. What kind of speed did he get with that thing?

  • by Davak ( 526912 ) on Wednesday April 10, 2002 @11:44AM (#3316842) Homepage

    Why go to all this trouble? This [] is obviously the way of the future. :)

    • Gee, how many times does a 3M product come into the discussion?

      This link refers to Liquid Nitrogen-cooled Fluorinert [](tm) FC-40 (Link points to 3M's page for this)...

      An expensive yet cool (pardon the pun ;-p) way to keep a system at low temperatures.
  • Eeek! (Score:4, Funny)

    by JimPooley ( 150814 ) on Wednesday April 10, 2002 @11:44AM (#3316844) Homepage
    The words 'cheap', 'water' and 'computer' used in close proximity do not inspire confidence or an incentive to try this mod out.
    Then I read about the cotton...!
    • agreed! :)))

      he could have used super-glue instead! :))
    • So this poses the question - where can one find links to professionally built water cooling kits? I have been interested in getting one for my computer, but exhaustive google searches have only turned up a few. What is really out there.

      For that matter, how about some links to some good overclocking sites with some solid information?

  • My CPU's always wanted a place to stop and chat while getting a cool drink.

    Hopefully it will increase productivity.

    Ohhh...water cooling for CPU's? My bad.
  • by sdo1 ( 213835 ) on Wednesday April 10, 2002 @11:45AM (#3316850) Journal
    Just sticking the CPU board into a mini-fridge [] would be cheap and would probably work pretty well. Plus any extra space could be used to keep your beer cold.


  • Does anyone else think it would have been infinitely cooler
    if he had used dental floss to hold the heatsink on instead of plain ol' string?

    C-X C-S
  • Does anyone else think this is a bit of an effort for a novelty system that's more likely to ruin your computer than provide an efficient cooling technique...and it doesn't even look sexy :(
  • by ch-chuck ( 9622 )
    I'm waiting for the DIY liquid nitrogen version...
  • by Tall_Rob ( 240828 ) on Wednesday April 10, 2002 @11:52AM (#3316910) Homepage
    Okay, so his water pump is only designed to run for a max of 10 minutes before overheating. Since overclocked CPUs generate heat, an underclocked CPU must be able to absorb heat, right? RIGHT? Why not slap an old 286 onto the pump, underclock it to run at, oh, say, 2 Hz (not mHz, but plain ol' Hz) and the 286 will absorb all the excess heat off the pump! Voila! :-)

    Or should he just get a water cooler to put on the pump, which would in turn need a water cooler for ITS pump, which would need a water cooler for ITS pump, which would . . . oh, wait, this is infinite nesting, isn't it? :-P
    • by markmoss ( 301064 ) on Wednesday April 10, 2002 @01:29PM (#3317753)
      No, what he needs (and no doubt he is aware of it), is a better pump. Also proper tubing fittings, hose clamps, baffles inside the waterblock to direct the flow so all of it gets cooled, ears on the block to attach spring-clip heatsink holders, and so on. But the cheap 10-minute pump is good enough for checking out whether he's on the right track -- and he claims his budget was 77 pence, which I think is about $1 US. Aquarium pumps run continuously and aren't too expensive, but on that budget you use what you already have...

      I do hope that he used thermal grease -- the article doesn't say one way or the other, but even a perfect heat sink can't cool well if the heat has trouble getting _to_ it.

      One thing that did kind of bother me: "Although it seems that copper would be best suited for making a water block, I'm not entirely convinced without physical proof." It scares me that anyone who has to ask would be doing this! Copper is indeed the best material, unless you are on a NASA cost plus 10% contract, then use gold and increase your profit. ;-) Copper is resistant to corrosion and has the second highest heat transfer rate (by volume) of any material available in bulk. Gold is better and aluminum is worse on both counts. When aluminum is used instead of copper, they are trading off a little effectiveness for considerable savings in cost and weight; if you are using water, weight better not be an issue, and for a do-it-yourself project you'd spend a lot more trying to solder aluminum than copper will cost...
  • by Anonymous Coward
    Well, some of us could never get something as ugly as that in the house.

    I have to screen all the computers I buy with my wife. If it doesn't fit with the decoration, it's not coming in.

  • Yeah, but if you get a water leak, your screwed 'cause your parents are sure gonna miss that cooler and defintely gonna miss that computer!
  • Convenience (Score:1, Funny)

    by Anonymous Coward
    No longer will I be forced to move to talk with coworkers around the water cooler. They will come to me, bwahaha.
  • Geez (Score:3, Interesting)

    by mcrbids ( 148650 ) on Wednesday April 10, 2002 @11:53AM (#3316920) Journal
    With the money he spent on this "cheap" water-cooler, he couldv'e PURCHASED A 1GHZ CPU!

    Oh, and it'd work for more than 10 minutes too!

    The things a guy will do...
  • First off IANAP but this guy should talk to one.

    He should take a look at some aquariums as well. That would take care of finding a pump that can run continually w/out overheating. I would think it has to be much cheaper than pumps built for remote controled submarines.

    Some small compression fittings would take care of the tube problem. Maybe some o-rings instead of rubberbands on the outside. I can think of a few ideas that would involve more than just pressure on the outside of the tubing. (That's why he had such a tough time stopping the leaks- he tried to tackle it all from the outside)

    Very nice box- a little tweaking and it would be very applicable for anyone.

  • And not that I don't trust this guy, but homemade is a bit scary. Maybe it is just me, but leaking is a HUGE issue, and I would want to be about 200% sure for something I am making myself.

    Still, these designs look good, and with a better pump, I would love to see some long term tests with this.

  • by bluGill ( 862 ) on Wednesday April 10, 2002 @11:54AM (#3316929)

    Water conducts electrisity. Well, pure water doesn't, but pure water will eat metal until it does conduct. That means you have to keep your water carefully sererate from everything else.

    By contrast, oil doesn't conduct, doesn't disolve metal. Fill your case with oil, and you have better cooling than air, and much easier to deal with. (Note, oil isn't as good as water for heat capacity, but it is still better than air and has all the other advantages)

    • by Wingchild ( 212447 ) <> on Wednesday April 10, 2002 @11:56AM (#3316943)
      Water cooling requires that the water, passing nearby a heat source, absorbs the heat and carries in on to a place where it can be safely transferred into the surrounding environment. The old Second Law, Energy can neither be created nor destroyed, etc etc.

      So why would we opt for water, which would be a less than optimal coolant? Because hot water makes *coffee*. Imagine! You could have a water-cooled server *and a coffee machine* all in the same rackmount!

      The possibilities are endless.
      • Water is the optimal coolant, when only physical properties are considered. It's reasonably low viscosity (easy to pump), and has an extremely high heat capacity. The only problems are that it is rather corrosive and electrically conductive, so leaks are really bad news, but the heat capacity is so much higher than most non-conductive fluids that engineers will often pick it anyhow. Other choices: Oil will take a stronger pump, thicker tubing, and bigger radiator, because you have to move more fluid at higher viscosity. Certain chloroflourocarbons are good enough at cooling and entirely safe to spill on live electronics, but they're also pretty much illegal nowadays. And distilled water is much cheaper than any alternative fluid...
    • Exactly! And when IA-64 arrives, you'll have a dual purpose computer/deep fryer!! George Forman Grill, eat your heart out!
    • Sure, subbing the computer in oil would work wonders. But I am not soo sure about the DVD drives being submerged in oil though :D
  • Why water? Why not mineral oil or something that's certain not to conduct electricity? It might not take heat away quite as fast as water does but it should still be a lot better than air.

    I wonder whether liquid nitrogen is feasible. That would be a great cooling system for a Beowulf cluster: remove the cases, hard disks etc and just stack motherboards really close together in a big bathtub filled with cold liquid.
    • Good point. This was only Rob's second attempt at a home-made water cooling block. Our next will involve water cooling around a heatsink which should help dissipate the heat much further.. Mineral Oil would be a great idea too. Thanks for the suggestion :)
      • Re:Other liquids (Score:2, Informative)

        by pkesel ( 246048 )
        You'll need a much more expensive pump to push mineral oil, or about anything but water. The extra viscosity will likely cause the pump to overheat. And the seals and such may decay with other materials.
    • Why not mineral oil or something that's certain not to conduct electricity?

      Distilled water is an insulator. If you drop your toaster into your bath, and you are bathing with distilled water, you will be fine. It is the ions in solution that carry charge.

  • Did anyone else read the part that the thing only runs for 10 minutes? Oh what a bargain. Buy all the watercooling stuff and have it work for 10 minutes at 1ghz. Woohoo.
  • In my experience rubber bands get weak, especially in warm places, and eventually break. I would not like to have a computer running when the "insulation" breaks and drowns the motherboard. Then again, this thing is supposed to take the heat out so maybe it's cool enough.
  • Personal Experience (Score:3, Informative)

    by quantax ( 12175 ) on Wednesday April 10, 2002 @12:00PM (#3316971) Homepage
    As someone who has built his own homemade watercooler, this setup is very jury-rigged and definately not recommended for any sort of serious long term cooling. The copper-tube inlets need to be replaced w/ brass fittings to avoid leakage, and the rubber bands should be replaced with cheap and much more effective band clamps. This waterblock design is not effective water-distrobution wise either as the water is not forced to flow through out the entire design. This would let the side furthest the inlets get hot due to poor water flow. An open chamber is only good for small waterblocks. Lastly, using thread to attach a waterblock to the cpu is ghetto as hell. Either quickly engineer a heatsink clamp yourself, or just look up a guide on the net for this, its pretty simple. Watercooling is one of those things that can be done many different ways, but this particular method is a little more 'amature' than is recommended to put on any piece of equipment that you value. If you are going to take the time to make your own watercooling, also take the time to make sure its engineered right.
    • by quantax ( 12175 )
      Oh yea, I forgot to mention, skimping on these sort of things for money is dumb because by using jury-rigged methods such as this, you usually end up spending MUCH more in hardware replacement costs. All it takes is a couple drops and could easily blow every component attached to the mobo (which is pretty much everything). Watercooling is not the sort of thing you cheap-out on since the consequences can rack up in cost pretty quick.
      • using thread to attach a waterblock to the cpu is ghetto as hell

      You say that as though it's a bad thing...

      • If you are going to take the time to make your own watercooling, also take the time to make sure its engineered right.

      I hear you, but bear in mind this is written for people who need to be told that "You can purchase a hammer in any DIY shop". Besides, there's that whole "guts beats hard work" mentality that's infested our psyche. How do you know who the Bad Guy is in a film? Easy, he's the one working out. The Good Guy is out partying, but it's OK, he'll beat the Bad Guy in the fist fight because, shucks, he's got guts.

      Uh, back on topic, that's pretty much the same attitude we're seeing here. Not "how you can do it right," but "here's how little you can do". Hacking for slackers. I like hacking (in all senses), but I prefer to see people hacking new toys, not just making shoddy replicas of commercial kit. I'd rather read about an innovative hack that's too hard for me to replicate than a shoddy hack that there's no point in me replicating, because it's inferior to a stock commercial product in every way. :(


    I prefer this one! :))
  • by xtal ( 49134 ) on Wednesday April 10, 2002 @12:10PM (#3317055)

    I run a watercooled machine as my primary work box. It's great, and the noise savings were incredible. No more whirrrrrr. Fits snugly into a standard mid tower case.

    I have a page up with all the details of contsruction [] for you who are interested. I've been running it for a few months, 24/7, and there have been no problems whatsoever. I took a few additional precautions, but the system as been moved around several times without any difficulties whatsoever and I highly recommend it to others who are interested.

    • Looks like you did a much better job than the other person. Excellent work! I would love to try this on my box.

  • by Hadlock ( 143607 ) on Wednesday April 10, 2002 @12:12PM (#3317065) Homepage Journal
    my friend had me mill one for him for a science experiment. actually , ended up about 5 of them. he bought 2x2x1" aluminum blocks. i milled two holes through (one end to the other), side by side. he tapped them, and attached plumbing devices to the newly threaded areas. The other method was to bore four holes - two holes one one side that went 80% through, and 2 holes on the adjacent side that also went 80% through. tapped, and attached appropriate connectors. no leaks to patch. since the tops were flat, he also took the old heat sink + fan, removed the fan, and used that to dissipate evem more heat.

    i never got the results back, but if anyone's interested, i can get the data to you, along with pics and more details.
  • Will all the memory chips huddle around it on Monday mornings discussing Sunday night's episode of Sienfeld?

    Sorry, I'll just stop now.
  • to go from .950GHz to 1GHz. Why not just blow $38.00 and buy a 1GHz CPU? Geez, and what a bodge!! The damn pump would only run for 10 minutes.

  • all he needs to make this 'kewl' is to add LEDs everywhere, i here that LEDs combined with case windows equals like 200mhz.

    this is as bad as reicer cars that have stickers and scoops to 'boost' horsepower.

  • by T3kno ( 51315 )
    This on definately wins the whisky tango (white trash) award. If only he would have shocked himself to death, he'd be up for a darwin.
  • Rubber bands? You gotta be kidding!

    A much better way to mate the hose to the copper tubing is to use a hose clamp []. These can be obtained from any DIY or auto parts store, and cost less than a dollar each.


  • Hydroponic CPU's (Score:4, Interesting)

    by muerte24 ( 178621 ) on Wednesday April 10, 2002 @12:44PM (#3317339)
    i built a water CPU cooler for $20 out of pocket costs. of course, that doesn't count the stuff i stole from work. :) it's a machined copper slug, a reservoir, and a $20 aquarium pump.

    if you turn the thing on with a cool reservoir, the CPU temp stays below 76F. but after being on for 10 hours, the reservoir temperature raises to about 113F due to my lack of money to buy a real radiator. so my equilibrium CPU temp with an Athlon XP 1600 is 123F, when the fan it came with ran it at 145F.

    you can see pictures and stuff here [].

    granted, copper slugs and machining equipment and "free" swagelock (and peltiers!) is not something everyone has, but use what you got, right?

    hope someone finds it useful or interesting.


    • by Exedore ( 223159 ) on Wednesday April 10, 2002 @01:16PM (#3317596) shed heat from the coolant. Just run the water coming from the cpu to a shower head (or similar) suspended above an open reservior. The water is dispersed by the shower head (increasing surface area) and cooled by the air as it falls into the container below.

      Granted, this approach requires an open reservoir outside of the case, but it's simple, effective, and cheap.

      Bonus: it can also replace those stupid "Sounds of Nature" tapes that people use for background noise at bedtime.

      • the problem with evaporative cooling is the increased hardening of the water. you would indeed lose lots of heat to evaporation, and the CPU would run super cool, but as more of the water evaporates and leaves behind all of its minerals, the think would start to get really crusty.

        then you would have to clean out the whole thing with CLR or some crap, and it would be a huge pain. and good system should be as closed as possible while still allowing for some thermal expansion of trapped air.

        the trick is to find a good radiator, CHEAP. or free. maybe a heater core from a car from a junkyard? maybe the heat exchanger from a junked air conditioner?

        the hard part is not getting the heat from the CPU to the water, but getting the heat from the water into the air.


  • After trying many different Peltier solutions, I became frustrated at how cumbersome they are and how painful they are to install.

    Finally I came up with a water cooling solution [] that was easy, simple, and , best of all, completely free!

    The picture isn't so good (it's a little hard to make out the details of the PC), but I'm sure everyone can do this mod too!

  • Ouch! (Score:5, Funny)

    by Verteiron ( 224042 ) on Wednesday April 10, 2002 @12:52PM (#3317409) Homepage
    "1 - 3 grams of solder should be enough, but it's always best to buy excessive amount just to cover yourself."

    Maybe it's just me, but wouldn't covering yourself with solder hurt like hell?
  • I find this mod really pathetic, mainly due to how he attached the cooling system to the processor.

    And how is there a benefit? My Celeron-2 600 sits at 36 degrees celcius and it is only cooled by a fan on the processor and two fans in the case. If attaching a water cooling system like his to mine only causes a 10 degrees difference in temperature, why should I care? Unless it got to the point of being 15 or 20 degrees cooler, I can give a rat's ass about it.

    Yet putting my system in liquid freon would be an option. How does a non-conductive cold liquid sound for cooling a system? :)
  • Besides the "it's cool!" factor (which it really isn't anymore, since everyone's been there and done that by now), why on earth would anyone water-cool their system nowadays? The difference between an Athlong XP 1500 and an Athlon XP 1800 is $14 [], and even the fastest Pentium 4 CPUs are reasonably affordable (to say nothing of the absolute cheap asking price for the fastest Athlons.)

    I guess what I'm getting at is this: why bother with any of this overclocking nonsense anymore? What on earth can it possibly buy you nowadays, other than a voided warranty and a fried CPU?

    - A.P.
    • Re:What's the point? (Score:3, Interesting)

      by xtal ( 49134 )
      Watercooling can completely eliminate system noise. In addition, in my experience, my system was much more stable when watercooled than when it wasn't. I run simulations that can run for a day or more, and stability issues can present themselves - no more with the watercooler.

      There's two reasons for you right there :). And, you can usually overclock to as high as the motherboard will allow as a side benefit.

  • He didn't really mention how many different pieces of hardware he ruined during this experiment, or what happens to his system at 11 minutes (when the cooling system fails), or my favorite question of all -- WHY? There are several water-cooling solutions on the market now, plus dozens of any other kind of cooling solution (Peltier, massive fans, etc). I understand the need to tinker and constantly tweak the equipment, but this guy must be really obsessive... Already noted here several times, but any cooling solution that works only for ten minutes at a time doesn't seem that great/newsworthy. BTW, won't cotton thread burn, create a lot of smoke, catch fire, and melt the entire lot of hardware -- oh, yeah -- it is water cooled, so you also have an internal fire-suppressant system.... Now I understand -- kewl!!
  • For it to really work you would also need a radiator to cool the water. A fan blowing air over the rad would also be required for small rads.

    The way it's currently implemented there is really no point. Good air heat sinks remain close to room temperature so they'll cool just as well as this water cooler. To actually improve the cooling you need so either cool the water or utilize a peltor with the water cooler to drop the temp below room temperature.


  • Easier Way (Score:2, Funny)

    by toblak ( 568717 )
    This might be a silly idea, but ... Why not buy a mini fridge, large enought to hold a computer, and a few choise beverages. Cut some holes for cables. And you are good to go.
  • Snipes (Score:2, Interesting)

    by John Guilt ( 464909 )
    1.) Don't heat the solder. heat the metal (with a small blowtorch). Maybe some acid flux first.

    2.) Use caulk to seal the hose to the piping.

    3.) Look around (larger Chinese groceries are good) for pre-formed metal trays, some of which have mtal lids that could be caulked shut.
  • by jhiv ( 163029 ) on Wednesday April 10, 2002 @02:06PM (#3318060)

    This story actually teaches a lot about how not to build a water-cooling system for a CPU.

    Some specific observations:
    1. The low delta T (temperature difference) for the water going through the system is a sure indication of low efficiency. The most likely culprit is poor contact between the block and the CPU. Thin sheeting was used, and ripples are clearly visible in the pictures. The block is probably only touching in a few places and there is no mention of using thermal grease. A stiffer bottom plate was clearly called for.
    2. As mentioned elsewhere, hose clamps should have been used. (String? Let's not go there...)
    3. There is no radiator to dump the heat back to the environment. The heat transfer from the surface of the tank is probably not sufficient to keep the temperatures low. The radiator should be after the pump, to dump the heat from the pump also.
    Perhaps the story should have been posted with the "laugh, it's funny" icon.

    BTW, I prefer Indium foil as the thermal gasket between the CPU and heat sink, not thermal grease. Unfortunately, Indium is usually as expensive as gold.
    • > 1. The low delta T (temperature difference) for the water going through the system is a sure indication of low efficiency.

      It can indicate that, but it's not a sure indication; in fact it can be a good thing!

      The temperature difference of the water is inversely proportional to flow rate- big flow rate gives low delta-t. So it depends on how powerful his pump is. A more powerful pump will mean that the chip stays cooler, provided that the inlet temperature doesn't rise.

      Clamps; check

      A reservoir may well be all that is required, provided the temperature stays well within spec.

      Indium? I think I start to see a pattern to your comments. ;-)

  • The guy's got the right idea, there's no reason his basic idea isn't good (I was considering doing the same thing myself) but it looks like it was implemented by Bob Vila - totally hacked together without even any real attempt to think of the best way to do it. It looks like he walked into a hardware store and grabbed the first thing that looked like it might work.

"Atomic batteries to power, turbines to speed." -- Robin, The Boy Wonder