CPU Heat -> Power

Posted:
in Future Apple Hardware edited January 2014
Maybe this is a stupid question but if Apple wants to reduce CPU heat and increase battery life why don't they connect a thermocoupling to the heatsink to convert the wasted heat into electricity?

Comments

  • Reply 1 of 10
    mr. dirkmr. dirk Posts: 187member
    Quote:
    Originally Posted by ngmapple


    Maybe this is a stupid question but if Apple wants to reduce CPU heat and increase battery life why don't they connect a thermocoupling to the heatsink to convert the wasted heat into electricity?



    Size?
  • Reply 2 of 10
    tubgirltubgirl Posts: 177member
    thermocouples are not very effective i think. (?)

    you'd probably be better off spending that weight and volume on a bigger battery.
  • Reply 3 of 10
    thttht Posts: 3,168member
    Well, how about a "steam" engine implementation? Use the CPU heat to heat a working fluid that powers turbines to generate electricity?



    Using a thermo-electric material would be a great idea, but tubgirl's got it. They aren't very good at converting heat to electricity yet. First use would be in automobiles though. Much higher temp difference there.
  • Reply 4 of 10
    slugheadslughead Posts: 1,169member
    Yeah a huge problem would be the fact that computers should not get above 150°F.. Steam forms at 212.



    Heat -> electricity is probably never going to be efficient because of that whole 'thermodynamics' mumbo jumbo about 'entropy'
  • Reply 5 of 10
    I think Apple would only consider an inexpensive and compact solid state solution to recover leaked power, thus the reason I'd see them interested in using a thermo-coupling.
  • Reply 6 of 10
    thttht Posts: 3,168member
    Quote:
    Originally Posted by slughead


    Yeah a huge problem would be the fact that computers should not get above 150°F.. Steam forms at 212.



    Yes, but I mentioned "steam" to relay a concept. The working fluid would obviously have to be something that would have a phase change at the temperatures and pressure a CPU, GPU, and core logic would produce. Or it could simpy be a single phase system (gas only).



    Quote:

    Heat -> electricity is probably never going to be efficient because of that whole 'thermodynamics' mumbo jumbo about 'entropy'



    Yeah, heat engines are only 15 to 20% efficient. But, I wouldn't knock anything that can increase battery life. It all helps in the end. It's the cost and weight that will determine if it is viable or not.
  • Reply 7 of 10
    marcukmarcuk Posts: 4,442member
    I reckon that instead of using standard transistors, we should use LED transistors and convert the light back into electricity and let the thing power itself.
  • Reply 8 of 10
    meelashmeelash Posts: 1,045member
    linky: http://www.media.mit.edu/resenv/pubs...uman%20heat%22



    Key quote: "today?s standard thermopiles are 0.2% to 0.8% efficient for

    temperature differences of five to 20 deg. C"



    Also: http://www.wirelessnewsfactor.com/perl/story/13874.html
  • Reply 9 of 10
    lupalupa Posts: 202member
    Why not just use plastic solar cells, these work within the infrared range anyways so they could absorb heat. They are fairly easy to produce unlike traditional solar cells, and although they don't have the best return they probably wouldn't hurt.
  • Reply 10 of 10
    tubgirltubgirl Posts: 177member
    yes, solar cells can be significantly more effective, but a even a hot cpu (~70° C) doesn't radiate very much energy. (maybe someone clever could do the math?)



    and then you have to address the issue of where to place the cells inside the case?
Sign In or Register to comment.