Breadboard Bits - Chillbox

The front view as it stands today. Time and harsh storage have not been kind to it. The finish is marred, the bottom hinge has failed, and it shows some signs of water damage. But here it stands.
One day many years ago, I discovered something interesting in my junk bin. I discovered that I owned a Peltier element! After managing to attach leads to it, and powering it up, sure enough, it created a temperature gradient. It didn't tend to get especially cold (it had obviously been damaged due to its origin in a box of bolts and other assorted hardware), but it did work.

So inspiration struck as I looked around for raw materials. Sure enough, my eyes began falling on things that I had available: mat board, aluminum flashing, packing foam. Pentium heatsinks, fans, prototyping board, electronic components, assorted connectors. I would make a refridgerator for a single pop can!

The inside of Chillbox as it stands today. The bottom hinge has been damaged, although it was a rather poor attempt at a hinge to begin with.
I started by building the inner shell out of aluminum flashing. I shaped the edges using my anvil and a rubber mallet. This part turned out really well; thin aluminum is a dream to shape with the mallet method. I drilled small holes (although, in retrospect, a punch would have worked better, if I had one) using a Dremel and pop-riveted the box together. I then glued a set of foam pads to each wall, leaving holes on the back to mount the LM335 temperature sensor and the peltier assembly itself.

The final exterior shell was made of mat board, a type of heavy cardboard used in picture framing. I ran wires for a power LED up to the front between the inner and outer shells, and I installed the temperature sensor under the back exterior wall.

Chillbox's control logic and peltier assembly. Also shown are the connector to attach to the power adapter, and some thermal compound that has managed to catch the flash.
The peltier assembly I made by wedging the peltier between a pentium heatsink and the back wall, using thermal compound on both faces. The heatsink was fixed in place with a set of bolts and springs tensioning a retaining clip that I fashioned of folded aluminum. The heatsink fan was strapped over the heatsink by twist tie.

The main control logic is a simple comparator controlling a relay. The relay turns on and off the fan and peltier to achieve the desired temperature.

Of course, the peltier being damaged meant that it never really got cold enough to engage the cutout. Also, the comparator circuit had no hysteresis, so it could possibly oscillate.

One piece that's been lost with time is the brushed aluminum nameplate that was to go on the front. The metal finish went flawlessly, but the actual lettering transfer failed repeatedly, and I never could get a satisfactory imprint.

Overall, it was a fun project, and rediscovering it has been a trip down memory lane.