Capillary tube conversion for CA Machines.

This topic will undoubtedly become an FAQ as more CA owners find this forum.
Firstly, a great deal of thanks are due to jhigdon for pioneering the capillary conversion for CA machines.
My CA-1 and CA-2 conversions were based on his information with some modification. The following is based on my experience, so other member's conversion procedures are also welcome. But first an introduction:

Need for a metering device.
A simple explanation is that for refrigeration to occur, a metering device has to be placed in the high side line feeding the evaporator.  Refrigeration occurs when there is a sudden pressure drop from the high side condenser pressure, to the low side evaporator pressure. In its simplest form, a restrictor is placed in the high side line before the evaporator inlet. Various kinds of self regulating valves or restricting devices are used to fulfill this function.

Float valve.
This automatically regulating metering device is used in all GE pre-war refrigerators, including the CA Monitor Top. As liquid refrigerant enters the float chamber, the float rises, and refrigerant flows at a restricted rate via a needle and seat. The amount of restriction is dependent on how high or low the float is.

Problems with Float valves.
Experience has shown that most CA machines suffer from sticking or worn float valves. It appears that over 80 years, the Methyl Formate has eroded the valve seat.
While there are many CA units still working, a large proportion are less than optimum. It is quite possible to maintain refrigeration with a worn float valve, but evaporator temperature will rise, which causes long compressor run times. Eventually, cabinet temperature cannot be maintained, even with the compressor running continuously.

Symptoms of a faulty float valve.

• Long run times.
• Noisy evaporator (rushing sound) with evaporator not cold enough to frost properly.
  
• Hissing sound from float chamber when compressor stops.
• Fluctuating frost line.
• Correct frost line but not getting cold enough (evaporator should be getting into single digits in degrees F).
  
• Full refrigerant charge but no frost line.
• Noisy compressor (float valve blocked).
• Note that some of the above are also caused by a defective heater.

How to fix the problem.

Because of the precision machined parts, and uncertainty of replacing them as per original, it is easier to bypass the float valve altogether, and use a capillary tube as is standard with modern refrigerators. The narrow diameter capillary tube provides the necessary flow restriction. The float chamber is still used as a high side (liquid) accumulator.

Parts required.

• 40" of .028" capillary tube. If this is not available, use the next size up, or down. Consult a capillary tube conversion chart to determine the equivalent length. For example, 29" of .026" can be used instead.
  
• An inline strainer, or filter drier with the pellets removed. It is not known how the pellets react with methyl formate, so it is probably best to remove them. Moisture in a methyl formate system is not the problem it is with other refrigerants.
  
• 1/2" to 1/4" reducer - used to connect strainer to float valve.
• Scrap of 1/4" copper tubing, depending on strainer.
• 1/2" joiner, if removing the evaporator for cabinet top restoration and rewire.

An important note regarding Capillary length.
The 40" of .028 quoted is for CA machines running on 60Hz. i.e. model CA-2-B16. The 50Hz model, CA-2-B15 runs at 1425 rpm, as opposed to 1750 rpm of the 60Hz machine. Because of the reduction of refrigerant flow the capillary tube needs to be resized accordingly. The float valve automatically regulates the flow between different supply frequencies, but the capillary does not. For 50Hz, the capillary tube is 44" of .026. Charge level is between 2.72 and 2.75 lbs for the CA-2. Exact amount needs to be found by experiment.

For R123 filled machines.
Refer to these threads for the development of this:
monitortop.freeforums.net/thread/430/guenea-pig-ca?page=1
monitortop.freeforums.net/thread/517/ca-2-evaporator-wanted
monitortop.freeforums.net/thread/522/pics-r123-capillary-tube-converted
 120" of .031" capillary tube is the final recommended for R123 (for 60 cycle use).

Procedure.

• Unless one is familiar with handling methyl formate, and its flammabilty hazards, the cooling unit needs to be evacuated first. For those experienced, it is possible to do the conversion with the unit still charged. See the thread monitortop.freeforums.net/thread/192/working-on-unit-fully-charged
  
• With cooling unit on its side, so the 1/2" float valve seat tube (Everdur) is horizontal, cut off the float seat, about 1/2" inch from the bottom, and withdraw the seat. Note that the float valve seat is solid and slides out of the Everdur tube. Do not cut right through. The condition of seat and needle can be examined.
  
• Remove all cutting operation particles from the Everdur tube, and solder on the the 1/2" to 1/4" reducer.
  
• 15% silver solder used for all operations.
  
• If using a filter drier, drill hole, about 5mm, in centre of filter drier, and extract all pellets. Install brass screw, and solder in place.
• Install filter drier or strainer, checking direction. Coarse screen to float valve; fine screen to capillary.
• Cut capillary to length. A file can be used to nick the circumference, then the tube can be snapped. Clean internal bore with X-acto knife.
  
• Insert one end into filter, making sure it is far enough in so that solder cannot wick in and block the tube. Be sure not to insert so far that the screen is damaged.
• By applying a few psi of nitrogen to the charge valve, ensure flow is evident at the other end of the capillary tube. This proves there are no blockages.
• Finally, connect the other end of the capillary into the high side evaporator connection - again ensuring the end of the tube is inserted at least an inch to prevent solder wicking blockages.
• Conversion is now done, and cooling unit ready for pressure testing (40-60psi), vacuuming, and refilling.

Refilling.

Important things to note are firstly, less than the specified 2.25lbs or 2.75lbs will be required. The correct amount has to be found by experiment.

To do this, gradually put in small quantities, and run the compressor for a minute. I put in .45lbs in two lots, then fine tuned quantity. As frost line comes up, put in lesser quantities, and run the compressor longer - allow several minutes for the frost line to stabilise. It will depend on each conversion, and is critical to obtain the best performance; i.e. shortest run times and longest off times.

The heater should have been running, with the oil up to normal temperature before finalising the charge, and the cooling unit must be in the cabinet.

A good plan is to put in 80% charge to start with, then add very small amounts if necessary (such as .04 lbs) until the right side header tank is frosted to about 1/3 to 1/2 way up. It is necessary to run the fridge for a couple of days, in the cabinet, to observe performance, as it takes that long for conditions to stabilise again. Note that room temperature and insulation quality can have considerable effect on cycling times. As refrigerant is added, cycling time will improve but then starts to get worse. The optimum amount is just before it gets worse. A rough estimate is about 80% of the original charge will be required. Typical run times are 2.5 mins (CA-1) or 4.5 mins (CA-2), and off times 12-16 mins. Note that frost line falls slightly when the temperature has stabilised and the machine is simply maintaining cabinet temperature. It should always be above the bottom of the header tank, however.