• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Peltier cooling.... opinions?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

a c i d.f l y

Member
Joined
Apr 1, 2003
Location
Austin, TX, USA
I'm guessing most who go the route of Peltier (possibly chained) supplementing with watercooling or similar methods to maintain the "hot side". What is the reliability of utilizing this type of cooling in a day-to-day scenario (power consumption mostly irrelevant)? I'm trying to research the most extreme cooling that could retain "daily driving" potential.

I only ever had a single 65w Peltier on a 120 MHz Pentium that I was able to ramp up to 166 MHz using a stock cooler. :sly: So to say my experience is dated is an understatement. What is your experience?

Anything beyond watercooling that I've read up on requires replinishing a substance or are limited to temporary situations or are fairly massive. Unless they're packaging portable multi-stage phase change units that I'm not aware of...:rofl:
 
The largest issue, beyond that of powering a pelt large enough to keep an OC'd PhII cool, is that of condensation.

In no-load situations a good water cooled pelt will be able to drag your CPU below zero*c, not only will water condense on it but it will freeze into ice.
Then the next time you put a solid load on the CPU the ice will melt and you'll have water. That water can cause issues!

For benching going subzero is absolutely worth it, the risk of hardware death is there but you can maintain a constant watch for water/ice buildup, and the extra few % of OC is critical.

On a 24/7 box it's a different story, an extra 5% OC isn't going to change all that much, and you have a larger chance of frying something.

I'm not going to say flat out "Don't do it!", but personally I don't think it is worth the effort, electrical costs, and risk of sudden watery death.



Plus if you really want subzero, you can buy a single stage phase change cooler (-20*c to -50*c depending on the unit and load), it will get you much colder and can still be used 24/7.
 
On a 24/7 box it's a different story, an extra 5% OC isn't going to change all that much, and you have a larger chance of frying something.

I'm not going to say flat out "Don't do it!", but personally I don't think it is worth the effort, electrical costs, and risk of sudden watery death.


Plus if you really want subzero, you can buy a single stage phase change cooler (-20*c to -50*c depending on the unit and load), it will get you much colder and can still be used 24/7.
+1; perfect response. Couldn't have said it better if I had tried. IMHO, not worth the risk for 24/7 use.
 
It can be done. It can be done safely even. However, your looking at 300W of pelt for one CPU, cost of a PSU to drive it, and the effort to seal it all up. Dielectric grease, high density silicone/neoprene foam, ect, ect. I have to say though that in a world where 30mhz isnt a game changer(anymore) its not particularly worth it. If you wanted to push your daily driver as hard as possible though it is an option, one that doesnt come well recommended.
 
Big thing with peltiers is that they are so inefficient at what they do.

Yes they are easy to throw on (and even destroy in the process). But the larger they are, and the more load your trying to deal with the more money you are gonna end up spending on other elements of your cooling setup. If you don't have a water cooling setup you have to build one, and that is still a good chunk of change.

I'd be more tempted lately to say go with a chilled water setup using a ac unit as a chiller. The cost is lower than most single stage setups. Plus if your a gamer it allows you to cool your video cards at the same time. Instead of just cooling your processor.

The big problem with any of these is condensation. Going from idle state to heavy load, and melting the ice that forms is quite a real problem. The longer the run times the higher the potential for problems, even with single stage setups.
 
Back