use plastic washers.
not enough heat transfer taking place to really worry about anyways though.
Printable View
use plastic washers.
not enough heat transfer taking place to really worry about anyways though.
I agree on the heat transfer - it might be an issue if you are running a 50C differential. But I don't recommend plastic washers - the application of torque is not accurate when you are trying for 200 PSI. Fiber washers are common, and a couple of those will also give good insulation. They allow much more accurate torquing of the assembly.Quote:
Originally Posted by STEvil
I always wanted to use TEC's but I never wanted to do this to my boards...
I figured that would ruin them when i was done with he TEC stuff.
http://lh3.ggpht.com/kgeronilla/R_sk...0/TRUETEC9.jpg
Very impressive, i made something similar, albiet much more ghetto a couple years ago when i first started working on my air cooled TEC case cooler. Used a 360w tec on my 2400 barton mobile with a 1/4" copper cold plate and a TT big typhoon with 110cfm SS 120mm fan, bolted to another 1/4'' copper plate. I don't recall the specific voltage i ran the TEC at, but i know i was running the cpu OCed to 3ghz (which was the speed it was run at with just the big typhoon cooler alone) and cpu core temp was 10-15c or under short duration load. Couldn't get an accurate read of idle temp but i freaked out and shut it down when i saw ice forming since i hadn't greased the socket =D. Always something i had in the back of my mind to try again, especially now that i have a much better understanding of how TECs work, but have been focusing on the case cooler.
As for people comments on the cold plate, 1/8th minimum thickness for copper, 1/4'' inch for aluminum. The first design for the case cooler used a 1/4'' thick copper bar 12in long, 4in wide, with 3 cheap aluminum cpu heatsinks with 80mm fans bolted to it spaced equally apart, for the cold side the another 1/4'' thick copper plate, 3in x 4in, with a TT Blue orb bolted to it. with no heat load mounted in an empy computer case the cold side got down to -7c with a hotside of 29c at best with a 23c ambiet. Think that was with a draw of 11v @ 17amps, but i'll have to go dig up the test data to be sure. It was cold enough for some serious frost to form on the copper cold plate, and frosty condensation to form to the tips of the Blue orb fins. So in my expeirience, copper conducts heat perfectly fine across wide areas regardless of direct area contact. With the tec bolted in the direct center of the 12 inch copper bar, there was only a 5-10c difference at the most between the temps directly above the hot side and the temps of the outer heat sink areas.
I would certainly want to have one of these, if not for everyday use, just to play around with, but this would certainly have the small customer base of the more insane modders and overclockers =P. I've spent the past two years working on my air cooled tec case cooler motivated by the fact that almost everyone thinks it's impossible to accomplish, or at the very least too impractical to be appealing, especially after the lackluster results of the retail TEC cooled case that came out a few years ago, though even before i knew anything about how TECs worked i knew the design was horrible.
At any rate, i've just picked up working on said case cooler again, and have decided to split my focus between entire case cooling and ducted airflow to apply direct cooling to the more important componets. First run results of the ducted design on the cpu resulted in 12-15c max load temps on my Toledo core amd 4400x2 OCed to 2900mhz, @1.4v after running 100% for several hours and idle within a few degrees of 0c + or -. That's with a max heat load of 150w, using two 437w tecs running off a 24v, 30amp mean well TEC psu. Using the massive nickle plated copper melcor heatsinks with 110cfm SS fan on the hot side. Each of the TEC/Heatsink units are mounted in their own case with 80mm flexi tube ducting which i've insulated as well as rigged up an adaptor to have both ducts connect to a single intake. Still not completly satisfied with it, and am waiting for new tecs to arrive before i can do much more.
But again, very impressive, though im interested to see what kind of results you'd get with the tried and true solid HS soultions in place of HP.
Look here, something the same
http://forum.crazypc.ro/showthread.php?t=7760
great job !!
i would like that to have on my Skulltrail.
What size did you use exactly ? Thank`s.
when i saw that huge plate ... i was thinking what if there's a waterblock with the size cooling the hot tec ...
Well done :up:
Max voltages your cooling could hold?
Quote:
Originally Posted by scifikg
Would the Swiftech MCX775-V fit the bill?
Found one for sale: NewEgg Link
That's just awesome! Sweet temps without the worry of a WC'ed setup. This has definitely opened my eyes to TECs and aircooling combo! Well done!
what worries of a water cooling setup?