1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

CPU Cooler

  1. Aug 16, 2009 #1
    Hi all.
    I want to create a Cooler for my CPU to better over-clocking
    Because heat has a negative affect on cpu performance and because of my fancy.

    I speak about a closed tube that a liquid is injected in it. It should be has a little vacuum, because when the liquid is heated and therefore expanded the tube has extra space to prevent exploding. Also a pump is inserted in this closed circuit somewhere to create flow.Both side (Cooling side and CPU side) has heatsink.

    I want to know can it cool the CPU. Is it better than a single fan?

    Is there an other liquid that was similar water but without the conductivity of water?Because water-if leak- can damage boards.

    What is ur suggestion about this system to enhance it.

    Thanks in advance

    Also excuse me for my bad English

    Attached Files:

  2. jcsd
  3. Aug 16, 2009 #2

    Vanadium 50

    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor
    2017 Award

    You can buy liquid coolers for CPU's. They have gone through all the design issues already.
  4. Aug 16, 2009 #3
    oh no
    I want to create it myself because of my fancy.
    Have u any idea or opinion?
  5. Aug 16, 2009 #4

    Vanadium 50

    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor
    2017 Award

    In that case, I would still start with the commercial coolers - look at several, make sure you understand the design, and can reproduce the stated design parameters - in particular, can you calculate the cooling, and do you get the same answer the manufacturer claims? Only after that can you be relatively confident that you are ready to design one yourself.
  6. Sep 29, 2011 #5
    You should probably learn about how air conditioning works, as it works in a similar manner. Additionally, if you're going to do it completely yourself, you should learn the thermodynamics math that goes with it. Otherwise, you might cool too little or cool too much, wasting power/money. Your local library should have a book on this, if it's well stocked, or you could look online.

    I also recommend you purchase a cheap computer to test it on, in the event it doesn't function as planned. It would truly suck if you forgot one little thing and you overheater your overclocked computer. This way you can work out any design/construction bugs before you risk damaging your expensive computer.
  7. Oct 16, 2011 #6
    You are a brave soul :)
    There is a liquid for this purpose, but it's very expensive: Fluorinert, by 3M
    I hope I spelled that correctly...

    Anyway, it was used to cool some Crays and is still used in the electronics industry for researching part failures and developing high voltage power supplies.
  8. Oct 16, 2011 #7


    User Avatar
    Gold Member

    An idea that I never pursued was the cooler system driving small generators.
    As Mike mentioned, something along the 3M product lines that produces a high pressure gas, use that to drive a series of two or more small, high speed turbines, (air tools used in dental practice) a place to start research.

    I have seen no attempts in this area, but I have not kept up with new technology for the last few years.
    The right design would actually cool the entire computer, just make sure condensation is properly handled inside and out.

    Jump on it.:smile:

  9. Oct 16, 2011 #8
    I would try it on something else before using it in your computer, though. That could be expensive if it goes wrong. Unless you want to build a computer from scratch as well.:wink:
  10. Oct 17, 2011 #9
    oooh, I don't know about the bolded part. If you create a vacuum the liquid will probably boil off and fill it...no vacuum. Perhaps some space filled with air for the liquid to expand into, but that might cause pumping issues. Either way I don't think that's much of an issure. I imagine that properly chosen pipes should be able to easily expand to adjust to the tiny liquid thermal expansion.

    Unless the liquid (water) freezes...

    But yes, liquid cooling is generally much more efficient than air cooling.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook