Peltier Torque Question

[rgathright]

Well, I got my Danger Den Maze 4 blocks and TEC in yesterday. I wanted to share something disturbing that I found last night.

While I was applying my $$ Artic Silver compound (everyone knows to create just a tacky like layer on the TEC, any more than that is excessive, I use a tygon tube to distribute the material), I tightened down the two blocks to the TEC with pliers. I had maybe two good turns past finger tight. Then I used my CPU PS to test it out. A few minutes later the blocks were cold/hot, but not excessive hot/cold. Well, since this is my second block I decided to get crazy and tighten the living c$#@#P out of the TEC. 15 Minutes later, I had finished and added a few lock nuts to keep it in place.

I decided to plug in the TEC and within 30 seconds the cold side had condesation and about a minute later frost.

Does anyone know how much torque we can put on these TEC's? It make sense, the more pressure, the closer the TEC gets to the metal. :shrug:

[killernoodle]

A ton of pressure, up to 300lbs I think (not on the bolts, on the tec as a whole)

[rgathright]

The metal bars holding my two Maze 4's together are bending.
I guess I should just use the other bolt holes as well. Thanks for the info.

[MadHacker]

Originally Posted by killernoodle

A ton of pressure, up to 300lbs I think (not on the bolts, on the tec as a whole)

is that 300psi or total 300lbs?

[rgathright]

Ok, here is what I did... I found 6 holes on the two water blocks and drilled them out (I tried to use the threads but the copper was to soft to get them started.) On the plexiglass, I jused large plastic washers to spread the pressure out across the block. Then to keep them in place, I used plastic cement (testors) on the washer and bolt.

[!nc2ca]

when I cranked the ^(&^*# out of my bolts, let it sit, then take it apart again, I could see that my cold plate and waterblock were obviously flexing, because there was littile arctic silver on the edges, and lots more in the middle. My block and coldplate are both 1/4" thick, 50x80mm. I figured that this is still better than not clamping tightly, so i clamped it together and mounted it. Im not getting the temps I want though.

13c idle, 23 load (ish)
25c water
226 @ 11 volts (its a crappy old computer psu with only my blower on the 5v to load it)
2500M @ 2610, 1.9v

anyone else haveing problems with flexing?

[jaydee]

Originally Posted by !nc2ca

when I cranked the ^(&^*# out of my bolts, let it sit, then take it apart again, I could see that my cold plate and waterblock were obviously flexing, because there was littile arctic silver on the edges, and lots more in the middle. My block and coldplate are both 1/4" thick, 50x80mm. I figured that this is still better than not clamping tightly, so i clamped it together and mounted it. Im not getting the temps I want though.

13c idle, 23 load (ish)
25c water
226 @ 11 volts (its a crappy old computer psu with only my blower on the 5v to load it)
2500M @ 2610, 1.9v

anyone else haveing problems with flexing?

Your block design is flawed for TEC use. You need two more bolts on each side top and bottom. Get the bolts as close to the TEC as possible to help eliminate the bending. A 1/2" cold plate will help aswell.

[jaydee]

Originally Posted by MadHacker

is that 300psi or total 300lbs?

Should be PSI I belive. I hear anywhere between 150-300PSI. Best to yap at the TEC manufacture for the right tourque.

[rgathright]

!nc2ca, WHOA! The machine work on that copper is great! How can I get one of those? :shrug: Tell us some more about it!

[!nc2ca]

jaydee116,
I never anticipated that that heavy copper could flex, hehe, but It would have been a pain to add bolts on the top and bottom because the raised bit of the socket gets in the way. its hard for me to cut out part of the bottom coldplate. I guess i could buy some 3/8" copper to make a new coldplate that has holes drilled in the top and bottom too, and i could put a cross bar over the top of the block. It would be ugly though. i wonder if i could solder extentions on the top and bottom of my block.

lol rgathright,
i wasnt going to until i got my temps down, but basically its just a lumpy channel (thanks jaydee) block. I printed out a plan, punched holes, drilled them, then connected them with a reinforced dremel cutoff wheel on a lathe. the holes are drilled slightly deeper than the channels. I kept a logbook here

[Butcher]

Tried cheap white heatsink compound rather than AS? AS doesn't handle low temps very well; some have had better results with standard heatsink compound.