Summer finally showed up, and I decided to try to fix the Chicken Wagon's air conditioning. As this car is our backup car, it would be nice to have it comfortable too!
When I got the car, the PO said "no air". Fine, this is pretty much to be expected. I noted that the belt to the compressor was missing, and the electrical connection to the clutch was unplugged. I did spin the clutch by hand and found that it spun freely, as did the hub (the compressor itself). That is, nothing was overtly wrong. However, with the broken vacuum pots, and especially with the broken solder joints in the ACC pushbutton assembly, it wouldn't have been operating properly in any event. There was only residual pressure left in the system, according to my gauges.
So my (wishful) thinking was, that the broken ACC panel was actually responsible for its non-functioning status. The ACC system had exhibited some rather spastic behavior before its repair (with much clicking of the vacuum switches behind the dash), and I theorized that they had pulled the belt to make it stop cycling the compressor too. Because the retaining clips to the oil cooler lines had been removed and placed in the glove box, and because the connector to the clutch shared mounting hardware with this, I rationalized that it had only been unplugged because they had worked on the oil lines. And I theorized that the charge itself had just gradually leaked away.
To save money during this testing phase, and to avoid any question of environmental harm, I used propane from a torch to put a small amount of pressure in the system. When the pressure got to be enough, the compressor engaged, and sounded good. Wahoo!
Sidebar: Propane is a good refrigerant, but it has approximately the same temperature/pressure curve as R22, not R12. If you try to use propane as a replacement for R12 you will have pressures that are much too high, and will probably blow a safety valve or a hose, with a small attendant risk of a fire. (Propane has a rather narrow range of flammability when mixed with air, too much or too little and nothing happens. It's actually much safer than gasoline in that respect.) However, in small amounts it works well to test out the system. Essentially you run with a severe undercharge so that pressures remain safe. There will be only slight cooling, because there is not enough mass in the working fluid to transport much heat. Along with this, there is probably too little working fluid under these conditions to transport the lubricating oil, so you only want to run for short periods of time to avoid destroying the compressor.Also, propane in bulk cylinders (hamburger gas) is not always dry enough for use in an A/C system. That found in Coleman torch cylinders is reputed to be drier than most, so it is recommended. You don't want to introduce moisture into your system.
See here for more information.
Unfortunately, according to my gauges, the system was not reacting properly. I was seeing high- and low-side pressures that were nearly the same, around 60–70 PSI. The primary cause of this would be bad reed valves in the compressor. Nuts! So I went to the U-Pull place and snooped around in the Chevy pen until I found an R4 compressor that seemed in good shape. The car I pulled it from still had residual pressure on it, so it wasn't open to the atmosphere, and was unlikely to be contaminated. Investment to date: $30.
I install the compressor, a knuckle-busting and oily fun fest. Vacuumed the system for an hour, and tried again.
No change! What the ....? So my next bright theory is that the expansion valve is stuck open, and the compressor is unable to build pressure on the high side because it's all running right around to the low side. (Either that or both compressors are bad exactly the same way, and this seems unlikely. Especially since both seem to suck and blow on the appropriate ports when you spin them by hand.) But this theory seems a bit far-fetched. I start to wonder if fluid is circulating or not. But to test this I need to block the system.
On the 300D, the low-side fitting is right next to a joint in the piping. So, I opened this joint and stuck rubber stoppers in each end. There, the system's blocked. I start the engine again. I pull the suction stopper and hear a nice thub-thub-thub sound from the compressor. I put my hand over the hole and it sucks it right down. This seems perfectly good. I repeat this a couple of times....BLAM! Whooooooosh!
Remember the other rubber stopper? I'd put it there to keep bugs, dirt, and moisture-laden air out of the system. Every time I let air into the compressor it built up as pressure on the high side which, as I had it temporarily configured, was everything up to the rubber stopper. It held pretty good for awhile, then shot out. The engine compartment and myself were covered in refrigerant oil. What a mess! I later found the stopper in the driveway more than four car lengths away. I poured some more refrigerant oil into the open hose to replace what was lost.
This bit of idiocy aside, things seem fine. But why is my suction gauge not reading correctly? The fitting is cinched down tightly, but it's not reacting to the obvious suction that my hand is feeling. I examine the fitting, and it looks good. This is not a cheap gauge set, it's a Snap-On, and the hoses are Robinair. (Bought used, but in good shape and of obvious quality. I've used them before, and they've worked fine.) Now that I know what to look for, it's apparent that my low-side fitting is making very poor contact with the Schrader valve. It appears that the stem is inset into the fitting more than most are. I pulled the valve stem, and it's OK. I put it back, and promptly broke it off. (Time out while I get the Easy-outs and pull the broken stem out. Good thing I can reach the hole for this fitting with my pinkie from the opened hose joint, so I don't have to worry too much about letting chunks loose in the system. What an idiot!) I replace the stem with another one from a spare high-quality brass fitting that I salvaged somewhere. Hmm, it still sits in pretty far. So I got out the file.
Now this is nervous-making. Filing on a fitting on an A/C pipe assembly. The fitting is not removable, so I can't mess up without really causing myself trouble. But it goes well, and I take off enough of the end of the fitting to make the stem sit flush, as it does on most other fittings I've seen. I reshape the end so it's tapered like before. I remove all the filings carefully. That ought to do it!
But it doesn't. This is really getting weird. I got out the dial calipers and measured the barrel diameter of the fitting: 0.350". Except that every other fitting I've got is more like 0.335" in diameter! The stupid barrel is too large and isn't allowing the hose connector to seat down enough to fully depress the Schrader pin. Out comes the file again. I take off enough around the outside to let the hose mate properly.
Success! I re-assemble the system, do a quick vacuuming, and try the partial charge again. More success! I get very distinct high- and low-side pressures that are acting right.
In retrospect, what was happening was that the gauges would appear to work correctly so long as the pressure was low in the system. But as soon as some pressure built up inside, it would push the Schrader valve closed thus isolating the system from the gauge. And, as you do all filling from the low side, as soon as you got a little refrigerant in there the process stopped, leaving you with a severe undercharge and a non-moving low-side gauge. The high side gauge was reading correctly, but because of the undercharge, it was low, and by chance about the same as was left on the gauge on the low side. A total red herring.
The car did not need a new compressor. It's a good thing it only cost me $15 and my time. I would really have been upset had I paid the going rate for this (racket). At least I kept the old one, I guess I can use it as a spare.
One of the pitfalls of converting an A/C system away from R12 is that there are a number of things that are dialed in for R12's exact characteristics. If you use a refrigerant with a different temperature/pressure curve, the system ends up operating inefficiently. One of the beauties of this hydrocarbon blend is that you can get a very close match to R12's curve. The system then operates well.
A usable charge of this hydrocarbon blend is less than 50% of an R12 charge. In part this is because hydrocarbons are more efficient refrigerants than hydrochlorofluorocarbons. Also, most R12 systems are actually overcharged to accommodate slow leaks. If you're doing your own servicing, a lesser charge is acceptable, because it's pretty easy to re-do. And, having a lesser charge reduces the fire risk in case of a sudden large leak.
What I find works is to feed in one can of the 70/30 mix, and then augment this with straight Propane until the temperatures and pressures are right. (With the system set for maximum cooling, and the engine at 2000 RPM.) What I do is to monitor the temperature of the pipe next to the high-side fitting using an infrared thermometer. As you slowly leak in Propane, you watch the R12 temperature scale on the high-side gauge and the actual temperature of the fitting. When they are close to each other ipso facto you are done. (This seems to be at around the 1/3–1/2 of a propane torch bottle on the cars I've done this on.) You also watch the absolute high-side pressure to ensure it's not too high, and you watch the low-side pressure to make sure it's still in the 20–30 PSI range. You have to make changes slowly enough to let temperatures and pressures reach equilibrium so that you get accurate readings.
It is my belief that the Propane is doing most of the cooling. The Isobutane is there to keep the pressures down. It provides enough circulating mass to transfer both heat and lubricating oil, and has a temperature range that keeps it from being a liquid in a harmful place. The proportion of the final charge that is the Propane is an 'undercharge', because if it weren't the pressures would be too high. But it's still enough to provide sufficient cooling, and the Isobutane is filling out the ranks. But because the system is still in a sense 'undercharged', it is very sensitive to leaks. A typical R12 system is actually 'overcharged' in that it is designed to be able to lose some portion of the charge (by inevitable slow leakage) without impacting performance. This blend doesn't have that luxury, so once some charge is lost the performance will suffer. Fortunately it is cheap and easy to correct! Because the components of the blend can leak at different rates you don't know what proportions are left once you've lost some charge, so you should evacuate the system and start from scratch rather than try to top it off.
I understand that this blend will work very well until ambient temperatures get above 100 °F. Because Propane has a critical point lower than R12, once its pressure gets too high it will stop condensing altogether. Then the only cooling is provided by the Isobutane, which has a higher critical temperature, and cooling performance is minimal. This can happen when it's really hot outside. If you live in such a place, stick with R12. (And you probably wouldn't be happy with R134a's inherently lesser performance if you were to convert.)
(OK, that's low. The above lengthy narrative is actually simplified, and I ended up fully charging the system four times (!), so I actually spent close to $50. Were I omniscient, I could have gotten away with spending merely $20. But any way you slice it, this is still pretty good. Notice that I did not replace the receiver/drier. One reason was to stay cheap. [Hey, it's a Chicken Wagon!] Another is that moisture in a HC-charged system is far less destructive than in an HCFC/HFC-charged system. In the latter, loose moisture ends up as hydrochloric and/or hydrofluoric acid, which then seeks adventure outside the system, usually by chewing through a wall.)