Make a Solid State Cooling Block for 50 bucks

Published: May 13th, 2016   Last Modified: May 7th, 2018

Occasionally you need to do an enzymatic reaction at temperatures lower than ambient, and some of those times you need to do it in reasonable volumes. Your thermocycler is being used by some plucky graduate student, and it wouldn’t fit anyway in those wee little cups for PCR tubes. What to do? Shell out 1000-2000$ for something commercial that you’re only going to use occasionally? NAHHHHH!!!! Whip up your own with parts bought off ebay and stuff you have lying around! 50 bucks or less, easy.

Specs? You can select a temperature and it can cool down to -2.0C in 10 minutes or less, good enough. Hindsight will appear as a theme in this write up, because I didn’t take pictures as I built the thing, so the only reasonable option was to take pictures while tearing it apart. You can pretend I’m building it by starting at the bottom. Worked well for a while, but a few critical oversights made the performance deteriorate, and there’s always time to do it right the second time.

This is a teardown of one of my prototypes. I won’t go into the wiring or the electric theory too deep, you can figure it out pretty easily by looking it up online, and half the fun is trying it yourself and screwing up, alot.

Look at this beauty, something only a soviet could whip up. The base is taken from an old MJ Research thermocycler, it provided an integrated, removable heatsink with fan assembly. The Peltier elements (which provide the cooling) were salvaged from an old PCR machine, but could easily be bought from ebay for a few dollars each. Essentially, when you apply DC electricity to a peltier element, one side gets hot, one side gets cold. If you can shuttle out the heat from the hot side with a heatsink and a fan, then the cold side will get even colder! This pile of parts is powered by a computer 12V power supply. Yes, there’s some exposed wiring, don’t touch me safety trolls. Let’s turn it on and see what she can do.

 

The brains of the operation is a little thermostat module (green) bought off ebay for 6 bucks shipped. Buying parts from Hong Kong is cheap, just expect 2-4 weeks shipping. This cute little part can sense the temperature of the block with a temperature probe. When the temperature is above (or below) a certain set limit, a relay turns on the peltier element which starts cooling the block. As you can see I stuck it right in there where a glass thermometer could go. I also wanted to be able to see what temperature the samples were actually seeing, so I submerged a separate temperature probe in mineral oil inside a 1.5 mL tube.

 

2C (3C) in the tubes is all it could manage this time, as I said, the performance deteriorated quickly. As we continue this autopsy we can see that the cause was that the heating and cooling of the unit in operation cracked the thermal adhesive between the peltier elements, the aluminum block, and the heatsink. Bad thermal conductivity = worse performance = can’t get as cold.

 

Insulation to keep the cold in? Foam from freezer boxes and some hot glue did great, actually, if a bit ugly. Note the cute integrated heatsink and fan duct that blows air through the fins. Here’s the cooling block/peltier/heatsink sandwich with the foam removed.

You can see I sealed everything away from moisture with some silicone caulking. The autopsy revealed that using bathtub silicone was not the right choice, as the acetic acid released during curing actually corroded some of the copper wires inside. Hindsight. Let’s lever it open.

Pretty chowdery in there. I had 6 little peltier modules in series glued on both sides with Arctic silver thermal adhesive. As I said, the cooling/heating stress cracked the joints and this all came apart easy. Next time a flexible thermal pad would be a better choice.

Old equipment that people throw away are a boon of parts. Some grey beard in a mountain retreat individually tested and matched these to have the perfect thermal characteristics over the entire PCR tray. Reduced to cooling my projects. You can see a bit of corrosion on the copper wires. Doubt it’s moisture, probably acetic acid fumes from the curing silicone.

There you have it, broken down, ready to be re-used in other incarnations. Besides the thermostat module (6$), the secondary thermometer (5$), some silicone caulking, thermal adhesive and tools all other parts were salvaged. You could buy a cooling block for 20-30 bucks off ebay if you’re savvy, and peltier modules are cheap, too. Just start slapping peltier elements on things and see what happens, it’s fun.

Update:
I call this picture “Humble-er Beginnings”, early prototypes of this cooling block actually used water as a thermal conductor. Yes, that is cork and tape as an insulator. Sorry there isn’t a better picture, this was something like 4 years ago.

4 thoughts on “Make a Solid State Cooling Block for 50 bucks”

  1. I’m thinking of making one of these for my lab, with an H-bridge to switch the peltier from cooling to heating. Then it can be used as a cold box and a warm box at the same time.
    Good writeup, I’ll make sure to find some thermal adhesive when i make mine!

    A propos, your images are upside down 🙂

    1. Hey! Glad you enjoyed it. Yeah, an H-bridge is a great idea, controlled by an arduino or the like. You could make a neat incubator chamber that way, too. I think I’ll invest in some 3M thermal tape to see if that solves my cracking issues. PCR machines seem to have a non-adhesive thermal pad that’s just held on by clamping force, but that requires actually designing something to clamp the sandwich together :O

      Thanks for tipping me off, it looked fine on my computer but wordpress does funny things with uploading images…rotated them manually, tell me if that worked for you. Atleast the Aussies could look at them alright.

Leave a Reply

Your email address will not be published. Required fields are marked *