Original regulator on freezer became faulty. It was very simple made with a capillary tube. The tube became leaky for an unknown reason, thus removing the capillary effect. This caused the freezer to run until the thermal protection inside the compressor decided it was time to switch off. This happens after the freezer reaches the limit for the coolant used, making the freezer go down to just a bit below -30c. As this is not optimal in regards to the electrical bill, something had to be done. And something was done, 1-2 years later...
There were several options.
1. Buy a new capillary tube with switch.
2. Buy an electrical regulator.
3. Build from the ground and up an electrical regulator.
Option 1: Have the fault that I do not like mekanical solutions.
Option 2: Costly, about 50 Euro or so.
Option 3: Cheaper, only expense was the components.
I did some research on this, main problem was finding a relay I could trust to make and break the highly inductive load of the compressor. I do not fancy sparks flying around and risk the relay melting down and set a fire :-)
After talking with an expert on household equipment at work, he gave me a ready-to-use control with a potmeter, relay, sensor electronics and 2 sensors. Only downside was there was no documentation to find on it nor did he have any :-(
I found that by only attaching one sensor to it, it would switch the compressor on at around +5c and off again around +4.6c. I had an el-cheep-o indoor/outdoor sensor wrapped in tinfoil close to the sensor of the controller. The now spare sensor was wrapped into this too, and I measured the resistance.
The sensor seams to be a 5Kohm NTC sensor type. NTC I am sure of, resistance rises with lower temp :) and it was about 5Kohm at 23c, the 5Kohm is the rated resistance at 25c.
After a quick and dirty calculation, I found that about 18-20K resistance in parallel with the sensor, would lead the controller to think that there was only +5C in the freezer, where as the actual temp. would be -18c.
An 18K resistor was soldered onto the controller's sensor input. Fantastic. The compressor runs until -18.5c, switches off, switches back on at -17.4c. Perfect!
This worked for a couple of cycles, I compared myself with GOD, no other, and ate some late supper. Now, the regulator seams off again... el-cheep-o sensor says -8c and compressor is not running...
A solution must be found... and pictures uploaded :)
Edit: Reason might be, that the empty freezer is too quick to change temperature and the compressor thus has to stop and start often, this is not possible for it to do, due to back-pressure in the pipes. When foodz are loaded into the freezer, the time it takes to warm it all up is much longer and the compressor will not run as often (but for longer times, of course).
Edit: Freezer is now filled up with transformers, heatsinks, harddrives and a mofo big soldering iron (at least 2kg), to give some mass for the freezer to cool down.
Also the sensor has been duck taped to the inside wall, making it possible to take out the drawers without worrying about the sensor.
--Jan Goofy 19:23, 11 July 2009 (UTC)