peltier

Cooling Processors quietly

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
SungHyun7
Posts: 149
Joined: Sat Nov 09, 2002 1:45 pm
Contact:

peltier

Post by SungHyun7 » Mon Nov 11, 2002 11:05 am

hey guys...

can i just put a peltier pad and slk-800 on top of it with no fan?

that should work right?

i'm running p3 1ghz 133mhz bus. fc-pga2

ChiefWeasel
Posts: 262
Joined: Sun Aug 11, 2002 3:26 pm
Location: Worcester, UK

Post by ChiefWeasel » Mon Nov 11, 2002 11:48 am

Hmmm... Peltiers are just gonna produce extra heat, so i doubt its gonna be the miracle solution Peltiers at first appear to be.

From my time on overclocking forums it is general consensus that air cooled Peltiers are dubious to say the least. Most require water cooling to be effective. Also you would probably need another PSU to power your Peltier, which would create more noise than a fan on your PIII would!

Sorry to stamp on another idea! I just wanna save you time and money! By all means have a play around, i went and bought a peltier cos they looked so cool! But i dont think they are gonna benefit us quiet PC enthusiasts.

Rusty075
SPCR Reviewer
Posts: 4000
Joined: Sun Aug 11, 2002 3:26 pm
Location: Phoenix, AZ
Contact:

Post by Rusty075 » Mon Nov 11, 2002 10:09 pm

Peltiers are heat pumps, and even very efficient ones at that. They move the heat from one place to another, while adding extra heat to it. Combine that waste heat with extra heat your PSU will be making to power the thing, and you'll end up with the need for more fans, not less.

And a problem arises when you run peltiers without active cooling. They can get so cold on the cold side that you get ice buildup. Then when you turn the machine off, the ice melts. Right onto your motherboard. not good, not good at all.

crisspy
Friend of SPCR
Posts: 228
Joined: Sun Sep 29, 2002 9:05 pm
Location: Powell River, BC, Canada

Post by crisspy » Tue Nov 12, 2002 9:49 am

I seem to remember peltiers making about the same amount of heat as they move. So 70W(cpu) + 70W(peltier) = 140W total output. Plus, if CPU > Peltier, then heat does a very sudden skyrocket. I think the only people who bother are extreme overclockers looking for subzero temps that allow faster MHz. Liquid CO2 or nitrogen is the next option.

SungHyun7
Posts: 149
Joined: Sat Nov 09, 2002 1:45 pm
Contact:

Post by SungHyun7 » Tue Nov 12, 2002 10:13 am

cool.

this is off topic but why doesn't usa use the heatpump method for the conventional air conditioning? it should save them huge electricity over the nh3 compression/decompression, i would think. if i remember correctly, doens't japan use heatpump for their a/c?

Fastman
Friend of SPCR
Posts: 47
Joined: Mon Oct 28, 2002 3:43 pm
Location: Vancouver, BC, CANADA

Post by Fastman » Tue Nov 12, 2002 12:50 pm

SungHyun7 wrote: this is off topic but why doesn't usa use the heatpump method for the conventional air conditioning? it should save them huge electricity over the nh3 compression/decompression, i would think. if i remember correctly, doens't japan use heatpump for their a/c?
Freon (and its non-CFC alternatives) and even the old ammonia based heat pumps / air conditioners / refrigerators are more efficient than thermoelectric junctions.
Heat pumps used for heating and cooling buildings have an efficiency of 400-600%. This means for every kw of electricity they move 4 -6 kW of heat.

The quoted efficiency of Peltiers is around 100% based on the 70Watts heat moved for 70 Watts electric consumption.

Fastman

SungHyun7
Posts: 149
Joined: Sat Nov 09, 2002 1:45 pm
Contact:

Post by SungHyun7 » Sat Nov 16, 2002 5:49 pm

well okay can i ask the same question with a slightly more info?

my point is this:
are we using ALL the capacity of a heatsink? what i mean is, if we put on another 70w of heat dissipation onto the heatsink, can it handle the extra 70w? I think such heatsink like SLK-800 or larger will be able to handle much more heat dissipation.

HEATSINK will stay hot, but CPU will stay cool because more heat onto the heatsink will make the heatsink hotter, which in turn make the heatsink dissipate MORE heat (besides, who cares if heatsink approaches 90*C as long as CPU stays at around sub 10*C). say CPU generates 70w, heatsink must be able to dissipate 70w to keep it cool. say CPU + peltier generates 140w, if heatsink can dissipate all 140w (at hotter temperature), we can keep the CPU at a MUCH lower temperature. you must understand that hotter material will dissipate more heat to the environment than the cooler material.

i'm pointing this out because I think peltier solution can be applied in NUMEROUS ways to our silent cooling by enabling heatsink dissipate MORE heat to the environment. if heatsinks can handle the additional heat created by the peltier, we can keep our components very cool.

only question now becomes, are the heatsinks competant to move double the heat (of course at a higher temperature)? we know it must eventually. and if peltier doesn't conduct the heat back to the cpu, we'll be all set. Can any one with a peltier pad try this? and tell us the result?

quokked
Posts: 333
Joined: Sun Aug 11, 2002 3:26 pm
Location: Melbourne, Australia
Contact:

Post by quokked » Sat Nov 16, 2002 9:13 pm

ummmm pelitiers don't quite work like that.... they work on a temperature difference between the hot side and the cold side, pretty much they work like a heat pump.... the slk-800 wouldn't be able to disspate THAT much heat going to it from both the peliter and the CPU....

The thermal rating of the SLK-800 is 0.46C/W from dansdata http://www.dansdata.com/coolercomp.htm and this is with a delta EHE fan as well (LOUD)

So this would mean that if u would put a peliter on it u'd have 140W x .46 = 64.4°c this would be the temperatures that would be coming out of that SLK-800 Approx :S and then u need to take into account the heat inside the case already and pump the heat out...
Oh yeah if u run the peliter too hot you'll bake your CPU over easy (There's many mishaps on the web that are peliter related) :)
for more info :) http://www.dansdata.com/pelt.htm

seishino
Posts: 150
Joined: Fri Aug 23, 2002 10:41 pm
Location: Boston, MA
Contact:

Peltier idea

Post by seishino » Fri Dec 13, 2002 7:48 am

Call me idealistic, but would it be possible to use a Peltier attached between the case and a long heatpipe to pull heat from the CPU to the outside of the case, where you could mount a rather extensive series of fins? Would the distance between the peltier and the CPU prevent the nasty ice buildup (near the motherboard) while still creating a good channel for heat dissappation?

OK, so there is the slight problem that we would have to put the door in a different place. But anyone see a reason why this couldn't work?

-C

Post Reply