dprantl

This is not exactly true. If there is enough capacity to bounce either of the evaporators off their freeze-switches, you will be getting extra cooling in the cabin. Even on a 100 deg F humid day, I can hear the rear evaporator solenoid switch after a while when the rear fan speed is set to 3. I suspect even if they were not freezing, two evaporators and blowers at 45 deg F cools down faster than one at 38 deg F.

Dan
'91 928GT S/C 475hp/460lb.ft

AO

Quick update:

The AC performed okay during SITM. One thing I noticed was that it seemed to slowly get less cool, to the point where it would blow almost ambient.

I would have turn it to warm for a while then kick it back on. THen it would work again.

Any thoughts on the cause (i suspect anti-freeze switch) and how to fix it?

dprantl

If your freeze switch was a problem, you would get reduced air flow out of the vents as the evaporator got blocked up with ice. It sounds like you had no airflow problem, but just an air temp problem. My first suspect would be the little relay in the A/C control head. Maybe yours is working for a while until it gets too hot and then stops working. When you put the head unit to warmer, it no longer commands the relay to energize the compressor clutch, thereby cooling the relay down. Then when you set it back to cold, the process repeats.

Dan
'91 928GT S/C 475hp/460lb.ft

AO

Thanks Dan. Makes sense. I beleive it still has the original relay. Maybe I'll do the swap and see if it makes a difference.

dprantl

When the air starts getting warm, check if the compressor clutch is still turning. If not, then you can diagnose to see where the +12V stops on the way to the clutch coil.

Dan
'91 928GT S/C 475hp/460lb.ft

dr bob

Dan--

The clutch circuit is not affected by the temp slider position. Movingthe slider adds heat from the heater core and HCV if the lever is moved far enough to the right. If the temp sensor loop is intact, iy would have to be all the way to the right on a warm day. I'll go with the freeze switch as a first guess, knowing it can be easily tested with ice water, an ohm meter and a thermometer.

WallyP

You can do a quick check on the freeze switch by just sticking a jumper on it and driving the car...

If the A/C is now colder, the freeze switch needs adjusting (some of them are adjustable, some aren't) or replacing.

dr bob

I think the original symptom is "too cold, makes a block of ice where the evaporator lives, partially blocks air flow, gets warm". That points to a switch that is stuck closed. If cycling the compressor off manually for a while helps, it's likely that the ice is melting some and airflow is restored when it's cycled back on again. The freeze switch may take some adjustment to get it just right.

Jim Devine

Found this:

http://www.robinair.com/acsolutions/...m/acvacuum.php

Pretty easy to understand & shows how important it is to do good prep.

Hope this helps!

dcrasta

Wow thanks for that link. Just saved me from buying a cheapo vacuum pump. I will have to research this more.

AO

Wow! Great link. That explains a lot. Maybe I do need to evacuate, vacuum it down and then re-fill. Very cool. Thanks.

dr bob

That Robinaire document is a great reference. They discuss the effects of moisture in the system primarily as it relates to corrosion and subsequent system hardware damage. While extremely important, there's a second benefit to low-pressure (high-vacuum) system preparation, and that has to do with the removal of air from the system prior to introducing refrigerant.

The refrigerant carries heat around the system by first absorbing it in the evaporator. The liquid refrigerant is passed through a smart valve that considers pressure/temperature and decides how much liquid refrigerant to meter into the evaporator. The evaporator is a heat-exchanger with hot cabin air passing over the outside of the fins/tubes. As the liquid enters the evaporator, heat from the cabin air is used to boil the refrigerant into vapor. The expansion valve meters refrigerant so that the pressure drops going into the evaporator, lowering the boiling point of the refrigerant. Heat from the cabin is therefore used to boil the liquid, so cooler air comes out and hotter refrigerant as vapor is extracted from the evaporator by the compressor pump.

Now lets add some air to the system. Air passes through the system as molecules mixed with the refrigerant. As it goes through the compressor, it is compressed right along with the refrigerant. As the blend goes to the condenser, ambinet air cools both the air moleculees and the refrigerant, except that the air molecules don't turn to liquid as the refrigerant does. They stay as bubbles in the liquid. So cooling capacity in the condenser is wasted cooling but not condensing those air molecules. The foamy mix is then passed on to the expansion valve, the one that meters liquid into the evaporator. The valve opens based on pressures expected fro the refrigerant only, not the combined pressure of the refrigerant plus air. It's confused by the air but isn't smart enough to know why the pressures are higher, so it throttle the refrigerant to keep the evaporator pressures closer to where it thinks they should be. The foamy mix passes through, with air taking valuable valve capacity. Once in the evaporator, the liquid flashes OK, but it's at a higher temperature because of the partial pressure from the air. Not as much heat is used to boil the mixture, so not as much is extracted from the cabin air. The warm air inside the evaporator doesn't carry much heat out, and in fact it contributes some heat to the refrigerant as they both expand. Then it's on to the compressor again for another round.

Air in the system is also a contributor of oxygen to the corrosion process. The acids formed when moisture in the system is broken down need oxygen to do their damage. Some comes from the moisture in the system, more from the air. So air is a double-edged dagger when it comes to having a healthy system.


Perhaps you can understand why I recommend leaving the vacuum pump on the system at least overnight. The maximum moisture removal happens at the highest temperatures, so a hot afternoon is better than a cold night for AC work. But the extended time helps a lot for getting as much moisture and air out as possible. The lower the pressure in the system during evacuation, the better, since it both boils more moisture, but also removes more molecules of air. As mentioned previously, I found a three-stage lab-quality vacuum pump that significantly exceeds the accepted commercial standards. It makes a marked difference in system performance. I cringe when I see single-stage air-ejector type pumps marketed for AC use, since they sledom do better than 27" or so. A minimum would be a two-stage pump, in good condition with fresh vacuum pump oil in it. The oil in the pump is hygroscopic on purpose, and ultimately will never be able to drw a system down below the vapor pressure of the water entrained in the pump oil. So fresh oil is a must.

AO

Okay, okay... I'm convinced I need to evacuate, vacuum down, and re-charge the system. Just need to locate the pump. Craigslist here I come!

Jim Devine

These were made by YOKOGAWA and labeled yokogawa, Johnson controls
and possibly more. Model h10 very accurate (and expensive)
As is always the case, when you really need one, there isn't a cheap deal. They are costly, but if you keep checking ebay, one will fall thru the cracks. They
have 2 models either batt powered or 110v plug in. Since the car is stationary
either will work.

Notice the spec sheet- capable of detecting less than 1/10 oz/ yr leak!

http://cgproducts.johnsoncontrols.co...DF/1927315.PDF

dr bob

Andrew--

Used pumps are OK if you have a clue to their prior use patterns. More than a few used on the market are worn out, so be ready to take a high-vacuum gauge with you to evaluate prior to a cash moving. For a hobbyist, an expensive pump is a lot; for a group of hobbyists it's a lot more affordable. Anybody near Andrew with a good vacuum pump? Anybody else near Andrew -need- a good vacuum pump?