Graphics Cooler

EnisadaEnisada Edmonton Member
edited June 2003 in Hardware
I'm going to be looking at a Gold Chipset Cooler to add into my existing Exos setup.

http://koolance.com/products/product.html?code=CHC-A05

I just want to know how much I can expect my water temps to go up on average and weather or not it would be better performance wise to just use an air solution???

Comments

  • TmodTmod
    edited June 2003
    Silver is a better thermal conductor than gold. I don't know why they plated it with gold instead of just leaving it copper. Here is the thermal properties of a few heat sink materials.

    Silver=428
    Copper=398
    Gold=315
    Aluminum=247

    The higher the number the better the thermal properties.
  • EnisadaEnisada Edmonton Member
    edited June 2003
    Hmm ic..but what about the rest of my questions???
  • TmodTmod
    edited June 2003
    I cannot give you a exact answer to that question.

    What I can tell you is that if the thermal properties of gold is less than copper and they plated the inside as well as the outside that would prevent the heat from transferring into the fluid as it would with copper or silver.

    Just look for a copper chipset cooler and go with that.
  • CCWCCW Suffolk, UK
    edited June 2003
    It is a copper cooler, it is not made of gold.

    "The all new gold chipset cooler features the same materials as the CPU-200G: solid copper, with 21k gold plating on top and internally, providing the highest performance possible from a 1 x 1 inch (2.6 x 2.6 cm) cooler!"

    The gold plating is on top therefore the copepr will actually be amknig contact with the GPU.

    Craig
  • TmodTmod
    edited June 2003
    I had never said it was made of gold. But I didn't notice that just the top part and internally were plated.

    I wonder if the inside is mirrored like platings usually are. Surface Tension of water likes to cling to mirrored finishes.

    I guess they put gold on the top and inside for corrosion protection. My .02 is they should have just left it all copper.
  • EnisadaEnisada Edmonton Member
    edited June 2003
    So would I get more performance from a water setup or an air setup????
  • TmodTmod
    edited June 2003
    I would say that the water will get everything cooler than air would.

    If you are going to run something like the gold plated cooler block then run some water wetter or something like that to remove some of the surface tension of the water.
  • EnisadaEnisada Edmonton Member
    edited June 2003
    K, thanks alot I will go ahead with the water setup and order the part for it today...
  • CCWCCW Suffolk, UK
    edited June 2003
    Water will get CPU temps to just over ambient room temp.

    Craig
  • Geeky1Geeky1 University of the Pacific (Stockton, CA, USA)
    edited June 2003
    tmod... I can tell you why they plated it with gold...

    copper is an excellent conductor of heat. Copper Oxide is... not. Copper and Silver both oxidize, reducing their thermal conductivity. Gold, on the other hand, does not oxidize...
  • DexterHolland911DexterHolland911 Hong Kong
    edited June 2003
    Enisada - Answering your original question, your water temps should not go up much at all. But then also there is no real reason for watercooling the chipset. Even if you overclock, the chipset doesn't really require watercooling.

    Guys - There is a difference between thermal conductivity and specific heat capacity. Tmod, you gave thermal conductivity statistics. Here are the specific heat capacities of some metals.

    Aluminum - 900
    Copper ~ 385
    Silver - 232
    Gold - 128

    The specific heat capacity of metal is how much heat energy it takes to heat 1 kg of that metal by 1 degree C. Therefore for a heat sink, you want a higher number ie it takes more energy to heat up the heatsink, cools the CPU better. Thermal conductivity is a measure of how much heat the metal is able to transfer through it (pass on) and for this you are right, the higher the number the more conductive the substance. This combination is why swiftech made/make their heatsinks with a copper base and aluminum fans, the copper transfers the heat better to the aluminum, which holds the heat better (due to the higher specific heat capacity). This is also why you see silver plated copper or aluminum heatsinks.
  • EnisadaEnisada Edmonton Member
    edited June 2003
    The GPU doesn't require watercooling for overclocking? For the high performance...
  • DexterHolland911DexterHolland911 Hong Kong
    edited June 2003
    If you are going to seriously overclock, it helps. I have my Ti4600 on watercooling and it adds maybe 30mhz to the core overclock.
  • stoopidstoopid Albany, NY New
    edited June 2003
    Volt modding the gpu or video memory is the best way to guarantee results from better cooling and overclocking.
  • DexterHolland911DexterHolland911 Hong Kong
    edited June 2003
    Better cooling goes hand in hand with volt modding.
  • TmodTmod
    edited June 2003
    So are you saying that the Aluminum/Copper construction will outcool a solid copper construction providing the same fan is used in a aircooled setup and both heatsinks are identical in size?
  • stoopidstoopid Albany, NY New
    edited June 2003
    Originally posted by DexterHolland911
    Better cooling goes hand in hand with volt modding.

    Precisely... ;)
  • DexterHolland911DexterHolland911 Hong Kong
    edited June 2003
    Tmod - Yes, assuming the Copper and Aluminum are fused properly, an Aluminum/Copper solution will outcool a solid Copper one, ceteris paribus :D. The problem is this sort of solution is hard to design, and costs more.
  • TmodTmod
    edited June 2003
    Hmmm that is strange as my SLK-800 will outcool my Swifty MC462 using the same fan. And the swifty is bigger.
  • DexterHolland911DexterHolland911 Hong Kong
    edited June 2003
    Real different. Look at the fins/pins layout. The SLK-800 has far far more fin area, and is designed better. Really they are completely different so it isn't fair to compare the two. If you tried to put Copper pins on the swifty it probably would not cool as well as it does currently with Aluminum pins.
  • Geeky1Geeky1 University of the Pacific (Stockton, CA, USA)
    edited June 2003
    Dexter's right... the only way to compare Cu, Al, and AlCu as materials for heatsinks is to use heatsinks that are absolutely identical in every way except for the material(s) they are made out of. Even then, there are a number of variables that could throw the results off one way or another- the fan (if it's not the same one in each test- not the same model, the same exact fan) the power going to the fan, the interface between the cooler and the heat source, the room temperature, the location of the probe, the accuracy of the calibration of the probe, the accuracy of the probe itself, etc. etc.
  • DexterHolland911DexterHolland911 Hong Kong
    edited June 2003
    Most of those errors should be minimal, ceteris paribus. The gap should/would be quite noticable, assuming the design was the same. Even the inefficiency of the connection of metals would be minimal and not enough to effect results.
Sign In or Register to comment.