Make a cold beverage using thermoelectric cooling

Originally published at:

1 Like

Peltier coolers are some of the most inefficient systems out there. I remember working on a system which used an early CCD array which needed to be Peltier cooled. Unfortunately some clever person put the heatsink inside the sealed cavity with the CCD, so that the CCD actually ended up hotter than without the cooler. (Yes, the solution - heatpipe- was obvious once the problem was identified, but still).


And it’s beautiful as well.

1 Like

10/10 for skillz. I’ll stick with ice.


I prefer to cool my drinks one atom at a time with lasers.


But don’t you have to get close to absolute zero before you start?

Mine’s a Bose-Einstein condensate with a little parasol and a couple of neutrons on sticks.


The efficiency is miserable; but there isn’t a whole lot of competition if your situation, due to size or vibration concerns or whatever, precludes using normal mechanical systems.

The fact that they are so stubbornly lousy, though, makes me wonder if team microfluidics could actually cram an entire refrigerant loop and compressor into the same space as a standard peltier module.


If you include the entire Peltier system with fan and heatsink, conventional technology is almost there:

Even these tiny ones (1.4cc pump!) have much more capacity than is needed for a single drink cooler.


These two statements aren’t really correlated; even if Peltier coolers were 100% efficient they don’t actually eliminate heat - they just move it, same as any other cooler. And because (like every other cooler in existence) they aren’t 100% efficient, they generate some heat themselves - so of course it ended up hotter. Not the fault of the Peltier cooler at all.

That being said: I remember in the mid-90s, when Peltier-effect CPU coolers were all the rage, being pretty disappointed with how much heat they were actually able to move. The hot side always seemed to get a lot hotter than the cold side got colder.

The real attraction of Peltier-effect coolers is that they’re (comparatively) tiny and silent. If those are the design characteristics that matter to you, a Peltier device is the right choice; otherwise, run away!

What, no Amazon link?

1 Like

I see I didn’t make myself clear.
The CCD was in a metal cavity containing other electronics. However, the heatsink had not been thermally connected to the wall of the cavity (it was not practical). The designer had reasonably supposed that the cavity as a whole would get a few degrees warmer but the CCD would be cooled by 20C or so, being smaller. In fact, the air in the cavity heated up by over 20C. If the cavity had been bigger the idea would have worked. As it was, a heatpipe from the back of the Peltier to the wall of the cavity solved the problem.
It’s probably unwise to write “of course” when my original description was insufficiently full.


That is impressively smaller than the usual ones; and they say that a complete system(with suitably chosen evaporator and condenser) can be fit into just over 1.6L.

Certainly looks more attractive when you need to move fair amounts of heat(as in their electronics case example); for smaller items the relative simplicity of TECs has its charm. They are mediocre at what they do under pretty much any circumstances; but they don’t stop being lubricated and potentially die if tilted more than 35 degrees from vertical; or have quite as many steps involving careful cleaning and inert gas atmospheres.

Yes, well, I’m not planning to use one in space.

This topic was automatically closed after 5 days. New replies are no longer allowed.