View Single Post
Unread 02-03-2001, 12:11 PM   #5
Rich W
Cooling Neophyte
 
Join Date: Feb 2000
Posts: 77
Default

In theory it would do better because you remove an interface (the waterblock-peltier interface).. but in practice, it would be difficult. If you could construct the box (indeed, you would need to) so that the peltier 'fit' inside it, and *NOT* so that the cap fit on the peltier. If you made it so that the cap fit on the peltier's hot side, then the edges of the PVC (or whatever the cap is made of) would cover part of the peltier's surface... and (I know from experience) that even having a VERY slight amount of a peltier's dimenion not being directly cooled will result in rapid overheating. If you construct the cap so that the peltier fits 'inside' the cap, and then is sealed somehow, it would work a lot better.... ie, no part of the hot side would be covered so as not to be cooled by the water.

My experience with protruding edges of peltiers has shown me that you can have something like 1/3 or 1/2 of a mm sticking out on a side (since there are no P/N junctions the close to the edge).. but anymore overhang than that and you run into problem REAL quickly. That is why a water jacket is better unless you are REALLY careful.... you can cool the entire face of the hot side, but you still have one interface to introduce inefficiencies.. You can minimize these by using copper water blocks, high turbulence designs, and silver thermal paste.

Hope it helps!
Rich W.
Rich W is offline   Reply With Quote