MMM
Results 1 to 5 of 5

Thread: help with TEC theory

  1. #1
    Xtreme Member
    Join Date
    Oct 2007
    Posts
    407

    help with TEC theory

    Okay. I have become very confused about the theory behind TECs. The consensus here seems to be to forget about TECs because they are useless. I also vaguely remember someone, maybe Cathar, going into an in-depth analysis of TECs and determining that, yes, they are in fact useless. But I don't get it. Can someone explain to me what has caused their downfall? Why are they so useless?

    I have an old TEC here that I have had for years (since the time that they were more popular). I am going to have a water cooled loop with at least one PA120.3 radiator (cooled with Sanyo Denkis) and at least one DDC 3.2 pump with Petra top. My heating elements in the loop will consist of an overclocked E8400, a P35 northbridge, and a G92 8800 GTS. So it looks like, even with a single PA120.3 rad, I am going to have 'excess' cooling capacity that is going to be wasted. And if necessary I could buy another rad to throw in the loop to keep the TEC happy. I don't understand why I cannot just slap a TEC on the E8400 heat spreader and cool the hot side with my Storm or D-Tek water block to shave off a few more degrees.

    I thought a TEC simply increased the delta-T between two elements while throwing in some waste heat. However, from what I have been reading this simplistic analysis seems to be wrong. What I don't understand is why. Can someone please take pity on me and explain this? What am I missing?

  2. #2
    Xtreme Cruncher
    Join Date
    Oct 2005
    Location
    BFE
    Posts
    4,432
    tec has died for now because of they way they work and the amount of electric they use. the way a tec works is like a heat pump. Your E8400 produces about 108watts of heat. the best 50mm tec you will get is a 400/320w. the 400w is the amount of electric it will use to cool 320w of heat. now looking at the #'s you say well thats cool but your chip is stock and producing 108w of heat and then you have to add in the pump temp and all the other heat producers in there. remember the the hotter your water temp is the worse your tec will work. so you will need like two 120.3 rads to keep up with the water temp because of the gpu, north, cpu, and tec. and then you will need a psu to run the tec off of costing about 400$. It is just so expensive and you can make a chiller from an old fridge that will cool everything way down and get good oc with. on my little story here I didn't even add in an OC. just think of uping the voltage and what would happen to the temps.



  3. #3
    Xtreme Member
    Join Date
    May 2006
    Posts
    187
    Quote Originally Posted by gojirasan View Post
    I thought a TEC simply increased the delta-T between two elements while throwing in some waste heat. However, from what I have been reading this simplistic analysis seems to be wrong. What I don't understand is why. Can someone please take pity on me and explain this? What am I missing?
    Under the right conditions a TEC can create a delta T. I think that when people go to use them, the inefficiency is a harsh reality. Published efficiencies are often a boatload of crap. I tested a cheap Korean TEC at about 40% efficient. Not only do you get to pay for more electricity, but you get a whole bunch of extra heat pumped into your office too!

  4. #4
    -150c Club Member
    Join Date
    May 2005
    Location
    Northeast, USA
    Posts
    10,090
    Great during the winter though


    If you have a cooling question or concern feel free to contact me.

  5. #5
    Xtreme Cruncher
    Join Date
    Oct 2005
    Location
    BFE
    Posts
    4,432
    That is why me and nol only do tec during the winter



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
  •