legacycentral bbs

I hope this is the right place for this question. I could not find a better one.
History: I just converted to R134a on my 92 Legacy wagon (R12 is too hard to get). I cleaned up the system real well before putting the R134 in (shot brake cleaner through it twice, followed by compressed air to remove the remaining cleaner. My old compressor was sized, so I put a new compressor on and filled it with Ester Oil, then put in a new reciever/dryer and sealed the system. Next I used a hand vacume to evacuate the closed system as best I could, then I charged it with R134a. Then the weather turned cold, and I had no opurtunity to test the system. the first hot day, a relay inside started going whacko and the compressor only ran intermittantly, resulting is way reduced cooling. this was in "Max A/C". It seemed to happen less in regular "A/C" mode, but it was still there. I bled the system just a bit to perhaps purge whatever air was still in the system, and it ran much better, but the wildly clicking relay still happened after that (once). Is the A/C controlled by the ECU, and is this mad relay a result of the ECU "learning" about the new refrigerant? did I fix anything when I bled the system a bit this AM? is the mad relay perhaps going bad? if that last is true, which one is it likely to be? the one that has a temp probe in the fins of the evaporator, or the one facing the cabin area that has the temp probe coiled inside a cage? Is there any to "fool" the ECU into making the A/C run right? I tested the pressure several times and it always came back in the good charge range. . .

I believe you are seeing the compressor cycling due to an incorrect refrigerant charge. It could be high or low; in either case, the trinary pressure switch mounted on the receiver/drier will cut the compressor out often.

You can't use a pressure gauge to determine charge level. In any functioning A/C system the refrigerant is in both liquid and vapor form, which means if you measure the pressure you just see the vapor pressure of the refrigerant, whether there is just barely enough to maintain the gas-liquid equilibrium or so much that there's barely any gas at all.

You may want to have the system leak checked properly. Usually when evacuating the system, you shut the valves of the manifold gauges and watch to make sure system vacuum stays steady for at least several minutes. That reveals most leaks.

Fix the leak(s) if any, then have the system discharged and evacuated, and charge it again by weight.

What do you mean when you talk about bleeding the system? If you just mean you vented some of the refrigerant, that doesn't really do anything in terms of removing air in particular. It just removed some of the charge.

It is also possible that your condenser isn't functioning right or something, resulting in excessive pressures on cold days or something. A set of manifold gauges will let you measure high side pressures as well as low side pressures and could give some insight. Any A/C shop ought to be able to perform the diagnosis fairly quickly.

Thanks for the reply Vikash. The system was running "fine" on colder days, it was only when it got to a hot day that it acted up. I know that I hear the relays in the evaporator when the cooling decreases, and beyond that not much. when I tested using a guage at inital fill (I put in two cans), the guage was on the low side of fully charged. when I tested it yesterday on a hot day I was 3/4 into fully charged. that was with the A/C running both times. I'm not seeing any oil at any of the connection points, so I don't know how much of a leak there could be. . . I was thinking that perhaps I did not get all the air out before the inital fill and that was causing the problems (with it collecting in the reciever/dryer. When I bled the system (in the hopes of getting any air out) I just cracked an unused connection until it started blowing oil, then sealed it back up again. that this helped perhaps points to the system being overcharged? Add in the higher pressure after bleeding perhaps and the pointer is more clear? The guage does say that higher ambient temps will cause higher readings though, and the second reading was on a hotter day than the first, with the AC running well. . .

Really -- the pressure in the system doesn't tell you how much charge there is.

The pressure differences you saw on the gauge are due to the temperature differences affecting R-134a vapor pressure and due to the air in the system.

Maybe you have enough air to push system pressure up above the upper limit of the ternary switch. I'd find that easy to believe... I'd find it hard to believe that one of those Mityvac hand pumps could pull enough of a vacuum.

Wait -- were you able to hold a vacuum on the system until you started charging? How?

No, I pulled as much of the air as I could out with the hand pump, then charged. did not get to a stable vacuum because I kept hitting the release while I was pumping. I finally got frustrated and went ahead with the charge. The whacky relay was the key, however, as when I switched it for annother one, the clickking went away and I was able to get really cold air out of the vents, reliably. I mentioned this problem on USMB and several people there pointed me to a relay that is a temprature controlled switch with a temprature probe sticking into the fins of the evaporator. this switch is supposed to keep the system from freezing up (which would block flow, causing the system to fail quite dramaiticly). I believe the switch I had was on the way out and the reason I did not discover it before the change to R134a was that my R12 system was undercharged, causing the compressor ot work harder just to get any cold at all (the cold kept going in and out), and finally sieze (likely due to oil starvation as there was little to no R12 to carry oil to the compressor). Because of this, there were few times that I got air cold enough to even approach the temprature this unit thought was the cutoff temp, but was in fact several degrees higher. Now I have what appears to be reliable cold air out of the vents. I mean "make your arthritis act up" kinda cold. or even "gives your goosebumps goosebumps" kinda cold. this was on a fairly warm day, and I rode with Max A/C on all the way to work to try and make the system fail. It refused to. I think the relay (the black one on top of the evaporator) was the culprit. the two tests I ran (Idling the car for half a day to simulate sitting in traffic, with the system set to Max A/C, and driving to work with the system set to Max A/C) both failed to cause a failure situation. If and when I finally source a new relay instead of the used ones I had available, I will put that in properly. For now, as the replacement relay also has many years on it, I'm not going to the trouble. This is mainly because I do not want to have to do this again if I end up digging into the evaporator housing (which I would have to do to get the probe into the fins of the evaporator. . .)

Ah, that's cool that your initial instinct about the relay being bad panned out. I'm glad you got it fixed.