Thermally conductive elastic material?

ed

ith

Maybe you could use a liquid interface. I once saw a working television completely submerged in an aquarium full of silicone liquid.

-- Joe

ed

ith

Sounds like the Sil pad material that is used for electrcially insulating components from the heat sink might work.

But, I'm confused as to why there is so much movement. I do something similar and have stainless steel screws connecting one aluminum block to the other. That keeps everything ridgid. One one side the screws pass through plastic shoulder washers, and then some wave spring washers to keep the compression roughly constant.

George H.

...

...

We haven't got much space for extra screws and other hardware (the top disc is a mere 20 mm across), and the surface must also be liquid-proof, so we're a bit reluctant to drill holes in it. But I think we'll try and construct one this way anyway -- because our current solution isn't robust enough by far.

Also thanks for the Sil-Pad reference; it these remind me of those very thick thermal pads used in laptop computers, so I'll search a bit more for those as well.

Richard Rasker

IIRC there's really thick sil-pad like material that goes by the name of "sil-astic" or something like that (don't trust my spelling).

Putting the thing together to have a positive contact with top and bottom, but without stressing the Peltier element would be a Good Thing, but I think it'd take some pretty darn clever mechanical engineering, and probably a few false steps where you _think_ you've gotten it right but you're still breaking Peltier's right and left.

er

ws

sc

How thick is it? Blind tapped holes in the top disk? Three threads is all you need... 0.1" for a 4-40 screw. (Sorry I'm a Yankee and still do mechanical things in inches.) If you need it thinner you could mill out a pocket for the TEC to fit into and leave it thicker at the edges for the screw threads.

George H.

ed

gh

ith

Back in 1993 we used graphite cloth as a compressible thermal contact medium between our Peltier junction and the surrounding metal-work and heatsink. It performed a good deal better than silicone pads, which really aren't much good - we were never able to get them to deliver the sort of low thermal resistances that the manufacturers claimed

Farnell now stocks something similar made by T-Global - their T62 tape. We got ours from Warth. TT-Global's 0.5mm T62 tape might be thick enough for your job

-- Bill Sloman, Nijmegen

Drill and tap blind holes in the back of the top disk, and secure it with screws from the heat sink side.

Sil-Pads are crap.

Hope This Helps! Rich

A large range silicone material is available with various thicknesses and hardness.

If you don't care about electrical isolation there are also graphite based materials with better thermal conductivity but they will be at the hard end of the scale and I'm not sure on how elastic their deformation is.

Look at the products from

There are also various silicone sealant and potting compounds some of which have optimised thermal conductivity and obviously no initial compression. Look at

None of the sil-pad sorts of things conduct heat very well. Their datasheets usually assume some huge contact pressure, and usually lie by about 2:1 to boot.

John

But _much_ more thermally conductive than air.

OK, great, I'll order some of their stuff first thing in the morning. I think I'll put this to the test, as well as the construction suggested by George. I hope this material does the trick, because it would save us some mechanical and electrical redesigning.

Thanks once again to all who took the time to reply,

Best regards,

Richard Rasker

In order to get mechanical compliance in the big-heatsink-to-test- plate assembly, consider a fluid coupling; maybe one of those PC pump and copper-block systems, to allow the test plate to glue to the Peltier device, but the back of the Peltier cooler just has a small copper block with some Tygon tubing carrying cooling water.

The test plate's mechanical mounting now does not go through the delicate thermoelectric cooler at all. Put a web of aircraft-composite type honeycomb under the plate, and use a second layer (which might stay hot) for your mechanical supports.

If time constants allow, you could cement the copper block to the test plate and put the Peltier and its heatsink at the other end.

Try Bergquist gap-pads. We use BER161 and 163 from Digi-Key (that's a Digi-Key part number, it is Bergquist's 1500 series. They have others with different hardnesses, different conductivities, etc.) Those numbers we use may be thicker than you need in your application.

Jon

Richard, I've worked on exactly the same problem (laser optic crystals mounted on peltier). We used Indium metal foil. It's routinely used for this purpose.

hydronics.