| Construction of the second R134a system (Phase-Change II) |
| Page 1: Phase-Change I post-mortem |
|
The first phase-change system died due to a refrigerant leak in the evaporator coil. This was a good
opportunity to design a second, better system. What was wrong with the previous system?
- Too slow refrigerant speed in the evaporator coil
The evaporator coil used in the previous system was made of 12mm O.D. (10mm I.D.) copper tubing.
This tubing is too large for the amount of evaporating refrigerant. This prevents a decent return
of lubricant to the compressor (the lubricant for the compressor dissolves in the refrigerant, and
circulates with the refrigerant). Compare it with a droplet of water in a hose. If you blow through
the hose gently, the droplet won't move. If you blow harder, the droplet moves through the hose.
For the second phase-change system, I chose to use 1/4" tubing in the evaporator.
- Badly sized capillary tube
Since I took the capillary tube from another system, it's length and internal diameter did not
even come close to optimal values. Later calculations showed that I needed approximately 50 meters
of this tubing. I used only 2 meters.....
For the second phase-change system, the size of the capillary tube should be considered carefully.
- Automotive coolant
This type of glycol-based coolant proved to be far from optimal for the job. I used one of the best coolants
available, with a freezing point of -37 °C. I added water to improve the heat transfer properties of the
coolant. But still, when the temperature of the coolant decreases, the viscosity of the coolant increases too
much to be useable without an enormous performance loss.
For the second phase-change system, another coolant was nescessary.
- Low inner diameter tubing
The 10mm I.D. tubing used, together with a highly viscous coolant restricts the coolant flowrate too much.
Larger tubing is required to keep flowrates decent.
For the second phase-change system, 1/2" tubing will be used.
- Condensation prevention
It is almost impossible to insulate everything in touch with the cold circulating liquid. A better method would
be to prevent condensation altogether. This can be done by submerging the mainboard into oil, or removing
all the moisture in the air inside the computer case.
In the second phase-change system, the mainboard will be mounted into a dry atmosphere.
- Low compressor capacity
If you design refrigerators for living, you can probably get the mentioned capacity out of the compressor.
If you are building a DIY liquid chiller without the experience and tools of a professional, it is
hard to come even close to the capacity mentioned in the datasheet. To make up for the capacity loss
due to inefficient design, an oversized compressor should be used.
In the second phase-change system, a more powerful compressor is needed.
All the parts had to be obtained second-hand. Here in The Netherlands you are not allowed to buy and use stuff like
compressors, refrigerant, etc. unless you have the required (STEK) license. This limited the choice of components.
While the search for the new compressor (I blew my old one) and capillary tube were in progress, I investigated
different suitable liquids for the liquid loop. A direct-die system was still out of the question since I want to
cool various other parts of the PC, like the videocard, RAM and Northbridge. Not only the processor.
|
| |
 |
|
