Results 1 to 13 of 13

Thread: General questions about watercooling / heat

  1. #1
    Xtreme Enthusiast
    Join Date
    May 2005
    Location
    USA
    Posts
    563

    Post General questions about watercooling / heat

    I am no stranger to overclocking, but when it comes to watercooling / heat output, there are a few things I do not understand. (I never took physics)

    When we look at a radiator or a heatsink, it has a specified amount of heat it can remove. Can we compare these apples to apples or does the radiator remove the heat faster, thus invalidating the comparison?

    For example: If I look at how much wattage a triple rad can remove, it seems comparable to what 2 large heatsinks can remove (CPU and GPU heatsink wattage combined). Would the water cooling still be better? Why or why not?

    I'm really confused about this in general. Maybe someone can offer me an explanation!

  2. #2
    Xtreme Member
    Join Date
    Oct 2008
    Location
    Florida..Tampa and St Petersburg
    Posts
    430
    i know one thing...when it comes to rad,s bigger is better...
    i7 920 4.2ghz..6gb 1600mhz cas 7 memory..Windows 7 Ultimate Edition..Windows XP..650 watt PS..Water..3DFX Voodoo 3

  3. #3
    Xtreme Cruncher
    Join Date
    Nov 2006
    Location
    'burb of Cleveland, Ohio
    Posts
    2,871
    I think part of why water cooling works so much better than air cooling is because the water picks up the heat and can carry more heat than the air... so take the heat away from the component and carry the heat to the radiator, where there is more surface area (fins) than an air cooling heatsink to disperse the heat to the air. THe radiator has more air moving through it than a heatsink does, and has more fin area than a heatsink.

  4. #4
    Xtreme Addict
    Join Date
    Jun 2007
    Posts
    2,064
    Watercooling is bout redirecting heat from the CPU/GPU/NB/SB/Mosfet to a "specialized" heatsink (aka radiator) to be cooled
    Watercooling can be upgradable in case users decide to have more radiators
    Watercooling can help by separating heat zones and cool zones, which traditional aircooling lack off
    Watercooling can both be noisy and quiet depending on the amount of fans e.g 3/6/9/12/15/18 fans .. but definely air cooling cannot place that amount
    Watercooling is wayyy more aesthetic than aircooling

  5. #5
    Banned
    Join Date
    Apr 2008
    Location
    Perth, Western Australia.... mate
    Posts
    740
    i didnt have a girlfriend before i got into watercooling....



    ... now I have a cat.... her name is mittens

  6. #6
    Xtreme Member
    Join Date
    Jul 2008
    Location
    Melbourne Australia
    Posts
    111
    Quote Originally Posted by b@llz0r View Post
    i didnt have a girlfriend before i got into watercooling....



    ... now I have a cat.... her name is mittens
    hahahaha thats gold!

  7. #7
    Xtreme Addict
    Join Date
    Jun 2007
    Location
    Melbourne, Australia
    Posts
    1,478
    Quote Originally Posted by b@llz0r View Post
    i didnt have a girlfriend before i got into watercooling....



    ... now I have a cat.... her name is mittens
    Classic. Dibs on sigging that one!

  8. #8
    Registered User
    Join Date
    Sep 2008
    Location
    New Zealand
    Posts
    67
    Quote Originally Posted by derektm View Post
    I am no stranger to overclocking, but when it comes to watercooling / heat output, there are a few things I do not understand. (I never took physics)

    When we look at a radiator or a heatsink, it has a specified amount of heat it can remove. Can we compare these apples to apples or does the radiator remove the heat faster, thus invalidating the comparison?

    For example: If I look at how much wattage a triple rad can remove, it seems comparable to what 2 large heatsinks can remove (CPU and GPU heatsink wattage combined). Would the water cooling still be better? Why or why not?

    I'm really confused about this in general. Maybe someone can offer me an explanation!
    Ok, wall of text coming, and probably a lot of simplifications (my limitations more than anything).
    A heatsink and a waterblock are just the ways we transfer heat. The better the waterblock or heatsink, the more heat is transferred.
    So when you're aircooling, you're using the air to cool the heatsink to cool the CPU. So basically you're using air to cool the CPU.
    And when you're watercooling you're using water to cool the CPU.
    Let's think about aircooling specifically for a little.
    Temperatures stablilise when you reach a sort of dynamic where the heat produced by the CPU is the same as the amount of heat transferred by the heatsink to the air. More heat is transferred at higher temperatures due to Newton's Law of Cooling, where a higher change between ambient and the temperature of the surface area results in a greater transfer of heat.
    Now what the heatsink uses is air. Air isn't very dense, and the air can be heated up pretty easily, so (these numbers are made up) the air going into the heatsink could be about 25 degrees Celsius, where the air leaving the heatsink could be about 35 degrees Celsius. The air is saturated basically, and since you're using the air to cool the CPU, hotter air means hotter CPU. This statement just above could be picked apart easily, but I just say it for a simplification, because going into the whole thing is difficult for me to say, and I'm guessing here using basic physics. Anyway the heatsink would also be a lot cooler then the CPU, due to the limitations of things like thermal paste, heat transfer of heatpipes, imperfect heat conduction.
    Water/liquid is much more dense then air/gas. Water is also much better at transferring heat then air, so more of the heat can be transferred to the water, then can be to air, at lower temperatures. The amount of energy it takes to increase the liquid one degree Celsius is more, about four times as much energy. So the water going into the waterblock is staying cooler, and is more difficult to saturate. The waterblock is more efficient, less distance between the CPU and the removal of heat means better removal of heat and greater cooling efficiency. The way the heat is removed from the water is by a radiator, which has a very large surface area, so can cool much better if we consider Newton's Law of Cooling again.
    So basically watercooling makes sense in those ways.

    It is not prudent to think of radiators of heatsinks removing specific amounts of heat, because there are too many variables involved. The ambient temperature, the mount (including thermal paste, pressure), the volume of air flowing. It's never really apples to apples, even between similar set ups (and that's why people dislike the 'post your temperatures' topics).
    The radiator moving heat faster. The radiator doesn't really remove the heat faster, it's just cooling the water, which is cooling the CPU. Since the water takes much more energy to heat up (and we basically assume that the temperatures in a loop stabilise as they are moving fast enough), it takes more for the CPU to heat up the water so the CPU stays cooler, and the dynamic between water temperature and ambient is at a much lower temperature due to the large radiator surface area. I guess the CPU block is moving the heat faster into the water, there's only a very small layer of copper and thermal paste to transfer between.

    Hopefully that helps some, no doubt I'll edit it to improve the nonsense. A little.
    Last edited by Ttocs; 12-05-2008 at 04:23 AM.

  9. #9
    Xtreme Mentor
    Join Date
    Oct 2007
    Location
    USA
    Posts
    2,622
    Air cooling is viable and pretty much has reached it's peak. It works great with normal overclocks, proper case selection and fan usage. Temps will be acceptable. Noise CAN be an issue.

    Water Cooling is viable and pretty much has reached it's peak. It works great with higher overclocks especially more voltages on the CPU/NB/GPU. We are moving past the standard overclocker. The one who will put a new bios on the GPU (more complicated by a long shot than just new GPU drivers). And will spend 3-30 hours playing with the CPU bios for that last 1 mhz and stare at the screen for 2 hours while prime runs. Never mind playing a game with friends online, mhz is where it's at. Yea, over extreme, but to set a point. Temps will be acceptable. Noise will NOT be an issue.

    Buying a good case and spending MAX $250 for aftermarket CPU/GPU/NB cooling for air. Or buying a massive MM case for $450, two WC loops to cool that $900 CPU and two 4870 x2 at $900+, and then the $1000 to WC it all. Your not at that budget level, no worries, it's your life, have fun.

    Whats more fun? What's your budget? Whats your needs? Whats your tolerance for tinkering?

    Too many variables for a simple answer. Lookies at my sig. Quiet, good temps, I have the $$ to spend. All my games run fine. Could it of been air? Sure, but water is cool, in more than one way.

    We WC for better temps and less noise. It works better to do that at the detriment of our wallets. And because it's neat. There is no other reason, IMHO.

    Lastly. WC is great, but never ever does it even become good unless you go quality, and quality means $$. You can't afford good WC? Stay air.
    Last edited by Conumdrum; 12-05-2008 at 05:16 AM.
    All stock for now, no need for more, but it's gonna be soon methinks.
    Giga Xtreme 58 mobo i7 965 ES D0 step Corsair 1600 6 gig
    SLI GTX470 EVGA
    EK HF nickle blue top CPU block (free from Eddie)
    Koolance 470 waterblocks
    One big loop, two 120x3 rads. Pa 120.3 and XSPC RX 120x3. Swiftech 35x pump with V2 restop. GT AP15 fans.
    Banchetto Tech Station
    120 GB SSD, and a few other drives.
    1000W UltraX3 PSU, 900 watt (1500VA UPS
    23.999" Acer GD235hz and 24" Acer H243H

  10. #10
    Xtreme Member
    Join Date
    Nov 2008
    Location
    Sweden
    Posts
    152
    Baseline is that you want the heat in the processor to get transfered to the air.

    - Air cooling
    Heat is conducted from the processor using aluminium/copper or heatpipes to transport the heat away from the processor, and then using air to cool the heated metal/heat pipes.

    - Water cooling
    Water is used to transport the heat away from the processor, with a thin layer of metal in between, the water is then cooled. Water have a significantly better at transporting heat then Air. Seen a figure of 5 times better, don't remember if its by volume of mass though.

    The advantage with water is that the maximum amount of heat transported from the processor is larger then using alu/copper+heatpipes, due to the conductivity of copper and space restriction, you can only fit so many heatpipes. You also got greater freedom in cooling the water as you are not limited to cooling close to the actual processor.

    I hope that makes any sense to you.

    English is not my native language.

  11. #11
    Banned
    Join Date
    Jun 2006
    Location
    in a Red Rage D....
    Posts
    3,839
    it will be quieter in most case with medium and slow speed fans cooling the current rad favorites... except for those delta bois...

  12. #12
    Xtreme Enthusiast
    Join Date
    Mar 2005
    Location
    Cologne, Germany
    Posts
    591
    Quote Originally Posted by derektm View Post
    I am no stranger to overclocking, but when it comes to watercooling / heat output, there are a few things I do not understand. (I never took physics)

    When we look at a radiator or a heatsink, it has a specified amount of heat it can remove. Can we compare these apples to apples or does the radiator remove the heat faster, thus invalidating the comparison?

    For example: If I look at how much wattage a triple rad can remove, it seems comparable to what 2 large heatsinks can remove (CPU and GPU heatsink wattage combined). Would the water cooling still be better? Why or why not?

    I'm really confused about this in general. Maybe someone can offer me an explanation!
    Unless the dissipation figures are given for the exact same conditions, comparing them is a waste of time. Some manufacturers release completely irrelevant figures that don't specify air/water-flow rates, temperature range, water/air delta etc. used as a basis for the calculation. Lacking a universal standard, these parameters can be tweaked to come up with any number the marketing department finds acceptable.

  13. #13
    Xtreme Enthusiast
    Join Date
    May 2005
    Location
    USA
    Posts
    563
    All of your explanations were very helpful. I now have a much better understanding of the technical aspects of water cooling. This info should be in a sticky!

    I was hesitant to post at first (didn't want to sound like an idiot), but it seems the WC community here at XS is full of great members!

    I really appreciate all of your responses. Thanks!

Bookmarks

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •