The first phase-change experiments
Page 1: First experiments

As soon as I decided that I wanted to build a phase-change system to cool the system components of my computer, I investigated the workings of phase-change systems and suitable refrigerants. A phase-change system is quite simple to build, if you can find a way to evacuate the system (pulling a vacuum into the system to remove all the air and moisture) and fill it with a suitable refrigerant. Since no real deep vacuum is required in the evacuation step, it is possible to find a solution for it. Finding a suitable refrigerant is a more complicated problem. In The Netherlands it is impossible to obtain common refrigerants like R134a or R502 without a license, so I had to search for alternatives. Obvious choices were propane (camping gas), which boils at approximately -42 °C under atmospheric pressure, and butane (lighter gas), which boils at -0.5 °C under atmospheric pressure. Both gases were not suitable since they are very flammable and possible explosive. Another choice was a spray sold at the local electronics shop, called Freeze K75, mady by Kontakt Chemie. This stuff is normally used during troubleshooting faulty electronic devices. It is able to cool things down to -50 °C. It is also non-flammable and affordable. Thus, it could be just the refrigerant I was looking for.


A spray can filled with Freeze K75 cold spray

To test it, I bought a very old R12 compressor unit from a large refrigerator for only a few bucks. The thing was old and rusty, but at least the compressor ran. And it contained a nice large condenser unit which I could use in further projects. I did not expect the unit to live long since it is not made to run with cold spray as the refrigerant. But then, I could find out why it broke, and what I could do about it.


An old salvaged R12 compressor unit from a refrigerator

For the first tests, I just made a simple evaporator by soldering the capillary tube into a copper bulb. I evacuated the system using an old membrane pump which just happened to have both an inlet and outlet connection. The vacuum it pulled was not high, but good enough for the first tests.


An old membrane pump misused as a vacuum pump