Shortly after the completion of the CA-2-B15 project, I received a query about a CA-1-B15, as a result of my postings on the Flickr group. The bizarre thing was that this fridge was located only half an hour away from where I live. A monitor top in Australia is rare enough, but to have another so close to home was the last thing I expected.

CA-1-B15 Introduction.
For American readers, the CA-1-B15 is the same as the CA-1-B16, except it is made for 50 cycle operation. This entails having a start relay and control able to function correctly on the lower frequency. It essentially means the contacts are beefed up to deal with increased current, when used on 50 cycles. The compressor motor is satisfactory on both frequencies as is. Of course, the cabinet light does not care about supply frequency. For models sold in Australia, a stepdown transformer was supplied to provide 110V from the local 210-250V mains.
The CA-1 cooling unit is used with 5 cu. ft. cabinets. It uses 2.25 lbs of methyl formate as refrigerant and has a cooling capacity of 430 btu. Compressor is 1/8 hp.
High side float regulation is used to control refrigerant flow into the "U" shaped flooded evaporator. The CA-2 unit is for 7 cu. ft. cabinets. It has the same compressor, but with the larger evaporator has a cooling capacity of 480 btu. There is also increased condenser area by means of tubing encircling the underneath of the cabinet top. Its charge is 2.75 lbs of methyl formate.

Faulty CA-1.
The owner claimed that the fridge seemed to have a leak. The compressor was working. Would I like to have a look and see about repairing it? Of course I couldn't say no to that, especially as it was so close. So, I visited the owner and was greeted by a 1934 CA-1-B15. Cosmetically, it looked not too bad. A quick look inside showed what looked like a crack in the join of  the low side tube. Perhaps it had leaked here? Out of curiosity, I started up the compressor which sounded as it should - quiet with no rattling. After about 20 seconds maybe, a loud hissing was evident inside the cabinet.
Opening the door resulted in a spray of methyl formate vapour. It was fairly strong but certainly not unpleasant. Nevertheless, I shut the door to avoid breathing any more in.
At this point, assuming it was just a cracked connection, I worked out a quote of what it would cost. Despite the price of methyl formate, and the time it would take, and given it would be a spare time job, they seemed keen to proceed. But, given the amount of methyl formate that came out, I decided to check further.

Broken High Side tube
I put my hand up the side of the evaporator and could feel something loose. I was able to lift the cooling unit above the cabinet top by placing my knee under the evaporator with my hand in between. This is when I realised something was amiss - there was oil all over the evaporator. Lifting the unit revealed the problem. The high side tube was torn off the evaporator. To make things worse, it was torn off right at the side where it goes into the header tank. No protruding stub that could be reconnected. This changed everything because I didn't like the chances of reliably connecting a new piece of tube to the stainless steel after that sort of damage. It would add to the complexity of the repair.
By now the owners could see the problem, and that I could never guarantee anything. I suggested that a later fridge using R12 than anyone could service would be a better option. I suggested they have a think about what's involved first before proceeding.

It's mine.
Then the lady of the house made me an offer I couldn't refuse. If I paid them what it cost them would I want it? So, for $120 I ended up with another monitor top, another CA.
The story of how they came to get it was that it was at an auction house in the city along with a 1940's gas stove they also bought. The idea was to create a kind of retro kitchen on the back patio that they could use for entertaining. However, as I was soon to discover, at some point the cooling unit had been mishandled. CA-2 deja vu all over again.

Just arrived home. The damage was now clear to see, as was the bodgie transformer installation.

One obvious difference between the CA-1 and CA-2 is how much easier it was to move. The 5 cu. ft. cabinet is lighter and smaller. It easily fitted into the back of the ute without having to take the legs off.
To get the cooling unit off, we used two lengths of wood under either side of the condenser and lifted it off the cabinet. The shorter evaporator and shorter cabinet made this possible, whereas it would not be so easy with the CA-2.

The damage.
Once home I could have a clear look at the damage, and here I could see a repair possible for the evaporator connection. The tubing was broken off flush with the header tank. There was no damage to the surround, so the header tank was still sealed. I could also see that the high side tube inside the header tank was a perfect fit for the tube which comes out of the high side float. All I would need to do is use this tubing and insert it, say half an inch, into the header tank tube and resolder. With that worry out of the way, I then noticed the float valve damage. The tube emerging from the float chamber, which contains the valve seat was bent to one side as well as suffering an indentation. I also noticed gouge marks on the low side evaporator tube.

High side tube cleanly torn off the evaporator.

Looking at the cooling unit lying on its side next to the cabinet, I started to see the likely scenario of events. It appeared that whoever moved the fridge in the past did not know the top lifts off. It seems that the idea was to place it on its left side in the back of a truck/trailer etc. However, as it was tilted over, the top started to lift. Hingeing on the left side of the cooling unit, the right side lifted causing the high side tube and float seat to catch on the cabinet lip. Further support for this theory is the presence of a dent in the condenser on the left side.

Float valve tube (made from "Everdur"- a low thermal conductivity alloy) was bent which could affect needle and seat alignment. Not visible here, but the tube was also partially crushed.

The transformer was interesting. This time it was mounted under the cabinet, but was modern. It had been attached by an assortment of self tapping screws. The connections were quite bodgie and not terribly safe. No cover was provided for the transformer or its connections. The wires were merely connected to the terminal blocks and left exposed. No cable clamping had been provided. Furthermore, the earth wire from the 240V mains lead had been cut off. This, plus the fact it was an auto transformer did not give a feeling of much safety. The 110V lead up to the light and compressor was original, and not surprisingly, in poor condition.

No attempt at covering the transformer or connections. Unbelievable that the earth wire was cut off, especially as this is an auto transformer which provides no isolation.

Restoration begins.
Having been through most of the learning curve from doing the CA-2 restoration, this one should be a lot quicker. In fact, it would be much the same procedure given the similar damage. However, I was concerned about the float valve. Given the "everdur" tube had been slightly crushed I feared the seat might be out of alignment with the needle. That, and the realisation that the needle and seat could be worn or sticking made me consider the capillary conversion option. It would be a shame to refill it with methyl formate only then to find out there were troubles.

First thing was to get the cabinet top apart, so removed the evaporator as per the CA-2. Inside was in much better condition. Although a rewire was in order of course, I would not have to repaint the inside.
Surprisingly, the oil heater was good. It measured 950 ohms. I rewired using 3 core flex like I did with the CA-2. Heatshrink tubing was used to recreate the correct colour code.

Cooling unit rewired. A repaint was not necessary, but the compressor to cabinet top seal was replaced with Silastic. While the rubber tape was largely good, sections were deteriorating.

The GE lead seal on the relay box was still there, so we can assume it had not been opened since manufacture. However, the relay contacts inside were surprisingly pitted. Not like those for the CA-2.

Evaporator Repair.
This went to plan without any problems. As before, I temporarily soldered Schrader connectors to the evaporator inlet and outlet for testing. I was particularly pleased to find no blockage. Now I knew how freely the nitrogen should flow through. The evaporator was left pressurised at 45psi for about a month and showed no sign of leakage.

Evaporator was kept under pressure for a month. It did not leak.

Preparing for the capillary conversion.
Once I'd made the decision to go through with this, I prepared the float valve. Bending the tube back was much harder than I thought - it must have been subjected to quite a lot of force. Then the seat was cut off. This was done 1.7" up the tubing because the dent in the tube might prevent the seat being extracted. The needle was in excellent condition. However, close inspection of the seat revealed some slot shaped wear at opposite sides. So, I had definitely made the right decision to change to a capillary.

Float seat removed from Evedur tube. This is necessary to disable the float valve.

A 1/2" to 1/4" reducer was required to connect the Everdur tube to the modified filter drier. I found that the ordinary 15% silver solder worked perfectly here. To minimise the risk of burning the cabinet top insulation, the reducer was soldered on before the cabinet top plate was replaced.

1/4" tube soldered to outlet of float valve. Due to availability, I used a 1/2" to 3/8" followed by a 3/8" to 1/4" reducer.

The filter drier needed modification as it is unknown how methyl formate would react with the moisture absorbing pellets within. A hole was drilled into the side and the pellets extracted. The hole was then sealed up with a brass screw and silver solder. The filter has a coarse strainer at one end, and a fine strainer at the other. This end is what feeds the capillary. The filter is necessary because of the small diameter capillary tubing. If any foreign particles blocked the tube it would be impossible to clear out - unlike a float valve.
Given the work done to the cooling unit, and that it had been run open to the atmosphere, such particles should be assumed to be present.

Pellets extracted from the filter drier and hole sealed up.

The cabinet top plate was cleaned up and cold galvanising paint applied to the rust spots. New stainless steel cup head bolts were obtained for the evaporator as the originals had the nuts rusted on, and had to be cut off. As with the CA-2-B15, new washers were made from neoprene. With the evaporator reattached, the top plate was ready to go back on. Prior to doing this, the 1/2" joiner was soldered to the low side compressor tube. It was found that the grommet for the float valve would not fit with the reducer in place so had to be left out, with the intention of sealing the gap with silicon or similar.
40" of .028" capillary tube was specified for the CA units, but due to availability, I had to use .026". There are charts to convert between different lengths and sizes which showed that for .026", I would need to use 29".
To cut the capillary tube, I used a file to nick a groove all around the circumference and then snapped it. An X-acto knife was then used to ensure the bore was clear. Using pliers or side cutters is unsatisfactory because the capillary will be crushed. The ends of the capillary need to be poked in an inch so that no silver solder finds its way to the opening.
Nitrogen was run through from the charge valve whilst soldering all the parts of the capillary conversion. Before the final connection, where the capillary enters the high side tube, I checked for a flow of nitrogen at the end of the capillary.

Capillary conversion complete.

Once the final connection was made, the cooling unit was vacuumed, then pressurised with 45psi of nitrogen and left for a week. No leaks were evident, so the unit was vacuumed again with the heater on, and the parts of the cooling unit heated by means of a heat gun. A second vacuum was performed until the pressure was down to 29.870". One problem evident is the charge valve adaptor is prone to leaking. I made a lead washer to seal against the charge valve which improved the sealing, compared to the supplied copper washers.

A digital vacrometer is far more accurate than the mechanical gauge and shows leakage at low levels which the analog gauge cannot.

Refilling.
On the day, further vacuuming was done for several hours. The pressure stabilised at 29.893". As with the CA-2, I used the plastic funnel tightly fitted over the charge valve to pour in the methyl formate.
First, I poured in about .45lbs, and ran the compressor for about a minute to get it circulated. Already, the bottom of the evaporator was cooling. After a short rest, I added another .45lbs, and the area of cold increased. In the end I had added 1.29lbs which got the frost line to the top of the header tanks. This is 57% of the normal charge. I also added a likely looking amount of  compressor oil (mineral type 4GS). This was also done via the plastic funnel. An interesting thing showed up here. The oil sits on top of the methyl formate and does not mix. It's also much more dense of course and the charge screw needed to be opened a little further for it to be sucked in.

Performance.
Immediately, there was a clear difference between this cooling unit and that of the CA-2 which still has the float valve. Not only did the evaporator frost much faster, it also got much colder.
Almost a month after filling, there was still no evidence of non condensable gases. This would indicate no leaks. There is however and interesting variation in frost level in that it sinks on the right side tank after the compressor has run for a few minutes. However, I assumed that to be lack of full charge. My reason for not putting in full charge at this stage is because the unit was not operating in the cabinet, so conditions are not as they normally would be. It could be possible to overfill. Once I had the cooling unit in the cabinet, then charge level can be optimised. Even at less than optimum charge, performance is far superior to the CA-2 with float valve.

Frosting commences at the end of the capillary tube.

Success of the capillary conversion is seen here. This amount of frost accumulated in only about 10 minutes.

Frost line with only 57% of recommended charge. The rag around the filter and capillary was an experiment to see if the cool air coming off the evaporator affected the frost line.

The Cabinet.
Despite the good appearance externally, several things made me suspicious about its real condition. The Textolite strip at the bottom of the door frame was warped. There were also a few pinholes in the cabinet, and a rust stain beneath the door opening. When I unscrewed the transformer, the screws were wet. First thing was to remove the Textolite strips. And here I struck problems. Most of the screws were extremely difficult to remove. I removed enough to see under the Textolite, and the horror was revealed. All the insulation was wet and had disintegrated. Obviously, the inner liner was going to have to come out. With a lot of persistance, most of the screws came out. The golden rule here is to use a very good fitting screwdriver, otherwise the heads will be chewed as they are quite shallow. I used a lot of CRC2-26 to try and free up the rust. By the end, three screws were too damaged and I had to drill their heads off. The thread is very close to 4 mm, but is obviously something else. It should be possible to retap and install similar 4 mm screws.
Once all the Textolite strips were off, it was very easy to lift out the liner and insulation. Fortunately, the liner had only minor surface rust. There was no way the insulation could be re used, even if it was dried out.

Years of moisture inside the cabinet had resulted in extensive rust and ruined insulation.

The cabinet rust was quite extensive, but looked much better after a considerable amount of wire brushing and scraping. The areas around where the legs screw in was covered in wax, so this escaped damage. Fortunately, the rust had not affected the cabinet strength. It goes to show, a good looking cabinet could be hiding something like this. And to think the previous owners were content with simply cleaning the outside of the cabinet...

Cabinet Restoration Begins.
I had heard about products like KBS and POR-15 as a method of treating the rust, from the Flickr group, and in the Model T Ford club. Given the severity of the rust, ordinary rust converter seemed like it would be inadequate. So, I studied both KBS and POR-15 and came to the conclusion they were much the same thing. One story was that the chemist who created POR-15 later created KBS.
I decided on the KBS treatment as it was Australian and I could easily get it.
The treatment consists of three stages; first a water based cleaner (interesting idea with all the rust!), secondly what appears to be some kind of rust converter, and thirdly, a thick sealer.

This is after the water based cleaner and rust converter was applied. There are actually a few spots of shiny metal!

After the second stage (rust converter?) and sealer, now ready for painting. You can see the spots of body filler used to repair the holes.

Normally, I am very sceptical of rust treatment, but the KBS really seemed to work. Much to my surprise, the water based cleaner made a huge difference to the inside of the cabinet. It looked a whole lot better, and there was even shiny metal visible in parts. After cleaning and drying, the next stage was applied...something by the name of "Rust Blast". This was then rinsed off leaving a coating over the rust.
Finally, the third stage was applied. This is a very thick flexible sealer that really sticks. I had to thin it before being able to brush it properly.
Once done, the improvement was like the difference between night and day. All the holes were filled with ordinary car body filler. This is important to prevent the entry of moisture.
Then the usual coat of primer and three coats of white appliance enamel were sprayed on. At this point, the outer cabinet was now complete.

Painting the cabinet.
As with the CA-2, I used aerosol cans to spray with. One coat of primer was used, followed by three of white appliance enamel. This made a huge improvement in appearance, and now the question of insulation needed to be dealt with.

Insulation.
As the original Thermocraft insulation was damaged, some alternative had to be found. I considered expanding spray foam, and while this would work, I didn't like the permanency of it. It would certainly seal out moisture and even provide a bit of extra strength to the cabinet, but the liner would be stuck in permanently unless isolated by a plastic sheet. Trying to keep to the original design, I decided to use styrofoam sheets, 60 mm thick as intended for under floor insulation. As it turned out, this was just the right thickness for the side pieces. The bottom piece had to be raised up with two layers of 10 mm thick foam pieces so that the inner liner was supported in the correct position.

Painted and new foam insulation installed.

The inner liner was in good condition. It needed only a few touch ups with black paint. However, the wiring had to be replaced. Dismantling the light socket resulted in breaking a screw, but fortunately I had a suitable replacement. I got the impression that the light socket, like the inlet socket, and door switch were never meant to be rewired. All the connections were sealed. Hence, the use of the junction box.
For the inlet socket and switch, I cut off the old fabric covered wire fairly close to the body and soldered on new wire, covering the join with heatshrink tubing. Unfortunately, the junction box screw was rusted in position so was not reusable without a lot of work. With the new wiring, it wasn't required anyway. The inlet socket and light socket were sealed with Silastic to prevent air entry.

Liner in position and light socket wires ready to be connected.

Liner needs to be supported not only by one 60 mm thickness of the foam, but by two 10 mm thicknesses also. In this position, the Textolite lines up perfectly.

Original insulation blocks were in good condition so were re used.

Cutting the foam.
In the end I was very impressed with the foam. It fitted well, and it was obvious how much better than the old Thermocraft (corrugated cardboard) it would be.
Do not use a saw to cut it! I initially did this but the mess was unbelievable. Tiny pieces of styrofoam went everywhere - and stuck to everything. I tried a hot wire cutter and this was not only mess free, but considerably faster and gave a better edge. My hot wire cutter consisted of a short piece of nichrome wire obtained from a burnt out jug element. It was connected to a 14V transformer which was in turn connected to a variac. This was set so the wire was just below red heat. The wire itself was just supported in alligator clips.
With the hot wire cutter, it was easy to make the unusual shapes required for the front of the cabinet that go beside the Textolite strips, particularly around the pedal mechanism.

Textolite.
I simply cleaned the strips with a scouring pad and water, leaving them in the sun to dry. New screws were purchased from Antique Appliances. The new screws were said to be 10-32 size. This is not quite the same as the original, which appears to be 10-30. Nevertheless they do go in, even though they're on the tight side. One nice feature with the replacements is that they have a much thicker head which makes it a lot easier to screw them in with the screwdriver slipping out.

Pedal mechanism.
The wire rods and stuffing box were cleaned, which involved grinding the rust off, and then sprayed with Galmet cold galvanising paint. The stuffing box is in very poor condition and ultimately I may replace it with a rubber bellows.

Cabinet top seal.
Here I could have reused the original No-ox-id cloth, but would have required a bit of work sealing the joins and tears. Besides, one problem with it as that the corners catch against the cooling unit when it is lowered in. So, I decided to try duct tape. I was quite impressed with this once completed, and the sharp defined corners wouldn't catch on the cooling unit. Besides that, there was no doubt a better air seal was provided. To get the width required several overlapping strips. The tape I used was 48 mm width.
The final part of the seal was a new neoprene strip to bear against the cooling unit.

Cabinet now complete with new top seal in place.

The door.
This was an easy restoration as all the Textolite screws came out without a huge amount of difficulty. The insulation was good, so it got reused. As with the cabinet, the door was primed and given three coats of white appliance enamel. The latch was lubricated with waterproof grease. The original gasket crumbled away so I made a new one from thin cardboard. Of course the door seal had to be replaced so a new one was ordered from Don at Rusty Metal Rescue. It is important to note that the hinges should be tightened on the door after they are screwed to the cabinet. This is so they are set to suit the thickness of the seal. I was caught out by this when the door was difficult to close after installing the new seal. The hinges and latch were polished with Brasso.

The Thermocraft insulation for the door was in good condition. It consists of layers of corrugated cardboard wrapped in brown paper.

Installing the cooling unit.
As with the CA-2-B15, I used an engine hoist. This makes lining up the cooling unit against the cabinet very easy before finally lowering it in. It would be extremely difficult otherwise, especially as there's not much room besides the fragile tubing. As expected, the duct tape seal also added to the ease of lowering it in. No corners to catch on the cooling unit.

Finally lowering the cooling unit in. The tubing must be kept clear of the inner liner edges. Unfortunately, many Monitor Tops get damaged at this point when the tubing catches and is ripped off or crushed.

After eight months of work, the X5 cabinet and CA-1-B15 cooling unit were back together.

All the hard work done. Now for the easy stuff!

Final work to be done.
A friend had provided me with several metres of black two core flex with 1.5 mm conductors. With this I could make up an authentic power cable that looked exactly like the original.

The feed through socket is for the cabinet light.

Wiring complete and like new.

As I got this fridge, there was a 15W 240V bulb in the cabinet. Obviously in the past the original had blown, and the owner not realising it was 120V, or not being sufficiently motivated to find a 120V replacement, had used an ordinary 240V bulb. It gives an dull orange light on 120V. While 120V bulbs have been sold in Australia, due to their rarity they have been expensive. Also, in more recent years, the available wattages have been limited. So, I simply bought a carton of 40x 15W 120V bulbs from the U.S. Even with postage, total cost was still less than they would have cost here.

Inside the cabinet with new 15W light bulb. The photographic developing tray is a temporary substitute for the glass chiller dish.

Finalising the refrigerant charge.
Now I was able to test run the machine. Even with only 57% of the refrigerant charge (1.29 lbs), the cabinet came down to temperature in about 1.5 hours. In the cabinet, out of any air draughts, the evaporator frost was more even. However, it was obvious that more needed to be added. While the fridge could be used as it was, the cycling times would no doubt not be as good as possible, and also with the level below the skimmer in the right side header tank, I was concerned about oil eventually building up in the evaporator.
The CA-1 is intended to be charged with 2.25 lbs of methly formate. However, when the float valve is replaced with a capillary tube, the optimum charge is less. This had to be found by experiment, and is why I did not put in the expected amount while the machine was outside the cabinet. Operating it this way does not give an accurate frost line.
With the cooling unit now installed, it was time to add more methyl formate. As before, I cooled the bottles just to minimise any internal pressure. The orange funnel was placed on the charge valve and more added. At 1.795 lbs all up (80% of full charge), it was looking very good, with frost level stabilising almost halfway up the right header tank. It seemed effortless for the cooling unit to bring the cabinet down to freezing point.
When adding refrigerant, adequate time must be given for the frost line to stabilise.

With 80% charge, and under load, the frost reached halfway up the header tank.

With 80% charge, the frost line had evened out and covered the full length of the right header tank. Performance was very good, with 2 min. 15 sec. run times and 12 min. off time. However, I felt that the frost line should be higher, about 3/4 of the way up the header tank, under load. So, I added another 10%, now making the full charge 2.035 lbs.
This did get the frost to the required height under load.

The new thinner layer of frost can just be seen about 1/4" above the original thicker lever where it was at 80%.

After stabilising, the frost line is still in the same place as when there was 80% charge.

Cycling times seemed quite a lot worse after I added the extra 10%. Now, the run time was 2 min. 30 sec. and the off time about 10 mins. However, upon turning the thermostat down three divisions, and removing the ice tray, it was vastly improved. Off times were now 16 mins. 19 sec. This was actually what I was looking for.
So, it seems the optimum charge is between 80 and 90% of the original specified, when a capillary conversion is done to the CA machines.
It appears that overcharging the CA does not show up by frost level, with the excess presumably remaining in the compressor dome. The rule of thumb seems to be therefore, to fill until the frost line evens out. A small amount may then be added to get the frost line 3/4 of the way up the right header tank when the machine is under load. Any excess will simply reduce off times, thus increasing average power consumption.

Restoration Complete.
A week's testing ensured the excellent cycling times were not some sort of fluke. Once the cabinet is down to temperature, under no load, the condenser runs cold, as does the float valve, with only slight warmth on top of the compressor dome.
I have been delighted with the performance. At no time has this fridge given any trouble. Nearly six months on from the first filling, it has still not needed an NCG purge.
There is no doubt that all CA machines should have a capillary conversion. So far, from my CA-1 and CA-2 experiences, and those of others who have restored these models, all needle and seat assemblies have been worn. It appears the methyl formate either erodes the seats by chemical reaction or just by flow, and in modern times prevents CA machines operating to their full potential.
Installing a capillary conversion eliminates this problem permanently.

Postscript - 18 Months after restoration.
Performance of the CA-1 has been nothing short of sensational. It has proven to be a reliable and dependable refrigerator with excellent cooling performance. Even at the height of summer with a few 36 degree C days, cabinet temperature is easily maintained. It took 18 months to require an NCG purge. This is an excellent record, given the machine had been open for six months and the oil would have absorbed some moisture.
In retrospect, the capillary tube should really be 44.5" long in view of the lower compressor RPM on 50 cycles. However, I was unaware of this at the time I did the conversion. In fact, to my knowledge, this is the second known CA unit to be converted to use a capillary tube. However, given the excellent performance, I am happy to leave it as is..
The performance of the cabinet insulation has been a welcome surprise. With the compressor switched off during defrost (23 hours before it automatically switched back on), the cabinet temperature increased by only 6 degrees F.