What's wrong with my car's air conditioner?

[Skammer]

I've had my '98 Jeep Cherokee in the shop three times since last summer for the airconditioner, and each time they say that they think they've fixed it, but it still gives me problems.

Here are the symptoms: Sometimes, it works great. Usually when the engine is still cold or the temperature outside is not very high.

Then, it will stop working. The fan continues to blow, and the air is somewhat cool (but not cold). The most notable difference is that the air is no longer dry; it's like cool, humid air, which quickly warms and makes the cabin even more uncomfortable than just opening the windows. The air also has a disinguishable smell -- related to the high humidity -- so I can tell right away when it's gone bad, even though it takes a few minutes to get uncomfortable.

Sometimes on long rides the A/C will kick in again and work for a few minutes, but it is intermittant at best.

My mechanics have replaced a relay; they have checked and filled the coolant and some other things I have forgotten, but this week we've started to have warm weather again and I'm still having the same problems. I'm going to take it to a different mechanic, what should I tell him (or her) to look for?

[Ins&Outs&What-have-yous]

As far as I can tell from your post, the compressor is working fine, but it only receives power intermittently. In '98 Cherokees (and most passenger vehicles for that matter), the A/C compressor is powered by the engine via the serpentine belt. It is possible that the belt is slipping along the compressor pulley especially after the engine heats up. If this is the case, then the serpentine belt needs to be adjusted or replaced. Do you ever hear the belt squeal when the A/C stops working?

[Hampshire]

Damnit! I think I need an MRI. I read the thread title as
What's wrong with my cat's hair conditioner?

I hate when that happens.

[kaylasdad99]

Originally Posted by ArchitectChore

As far as I can tell from your post, the compressor is working fine, but it only receives power intermittently. In '98 Cherokees (and most passenger vehicles for that matter), the A/C compressor is powered by the engine via the serpentine belt. It is possible that the belt is slipping along the compressor pulley especially after the engine heats up. If this is the case, then the serpentine belt needs to be adjusted or replaced. Do you ever hear the belt squeal when the A/C stops working?

Wouldn't the belt be driving the pulley whenever the engine is running, whether the AC is on or not? I was under the impression that a solenoid-operated clutch needs to be powered and engaged to make the compressor turn with the pulley.

OP: is there a snap or clunk noise when the AC kicks in or out? Maybe a grounding wire is intermittently losing contact with the engine block as the engine heats up. Or possibly even a wire crimp is thermally expanding and losing contact with the actual wire. That's not something that's likely to be caught when the mechanic tightens a bolt on a connector.

[gotpasswords]

A second opinion's a very good idea.

Off the top of my head:
The compressor may be short-cycling because of improper refrigerant levels, the compressor clutch could be failing, or the engine computer could be sending bad signals to the clutch.

[Ins&Outs&What-have-yous]

Originally Posted by kaylasdad99

Wouldn't the belt be driving the pulley whenever the engine is running, whether the AC is on or not?

Sure, the belt continues to rotate the pulley but the pulley should be idle while the A/C is off, right? On my '88 Cadillac a few years ago, the worn A/C belt (not a serpentine on that car) stopped squealing when I turned the A/C off. The problem stopped when I replaced all the belts.

I only mentioned this as a possiblilty on the Jeep. I realize that Skammer's A/C interruption could be due to electromechanical problems unrelated to the belt, but I don't much about troubleshooting circuit gaps in autos.

[Skammer]

When it's very hot outside, and I start up the car with the A/C on, I'll hear the belt squeal unless I turn the A/C off. But normally I don't hear anything, including pops or clicks. The only sign the A/C has quit is the subtle change in the feel/smell of the air coming out of the vents.

I think (but I'll double check) that the compressor was what they replaced last year; and they also refilled the refrigerant. That's not to say it isn't one of those, but it shouldn't be.

Thanks for all the responses so far.

[Kevbo]

You probably need to be using a shop that specializes in automotive AC.

[flex727]

Does the problem occur more frequently when it is humid out? Sounds to me like your A/C is freezing over. Try leaving the fan on but turning off the the compressor next time the problem occurs. Wait about 5 min and see if the air changes some, then turn the compressor back on and see if it works. There's supposed to be a switch that cycles the compressor on and off to avoid freezeup - perhaps it's not working.

[danceswithcats]

Originally Posted by flex727

Does the problem occur more frequently when it is humid out? Sounds to me like your A/C is freezing over. Try leaving the fan on but turning off the the compressor next time the problem occurs. Wait about 5 min and see if the air changes some, then turn the compressor back on and see if it works. There's supposed to be a switch that cycles the compressor on and off to avoid freezeup - perhaps it's not working.

That's the direction of my thoughts. With an electrical schematic of the AC system, the cutout and limit switches controlling the compressor coil circuit can be isolated. Perhaps Gary T or another pro will offer additional insight.

[Darryl Lict]

Originally Posted by Kevbo

You probably need to be using a shop that specializes in automotive AC.

I agree. My mechanic didn't even attempt to diagnose it and sent me to the air conditioning guy who fixed it right up.

[Rick]

Originally Posted by flex727

Does the problem occur more frequently when it is humid out? Sounds to me like your A/C is freezing over. Try leaving the fan on but turning off the the compressor next time the problem occurs. Wait about 5 min and see if the air changes some, then turn the compressor back on and see if it works. There's supposed to be a switch that cycles the compressor on and off to avoid freezeup - perhaps it's not working.

Based on the symptoms listed in the OP it is NOT a frozen evaporator. if the evaporator gets too cold, water from the air will freeze when it hits the evaporator, and form a sheet of ice across the surface. Needless to say, air cannot flow through ice. The OP mentions warm humid air flowing from the vents, therefore, the evap is not freezing.
Which relay was replaced?
You said they replaced the "coolant" Coolant is the antifreeze and water that is placed in the radiator to keep the engine cool. Refrigerant is what is inside the A/C system to keep you cool. Some times it can also be referred to as Freon (although this is technically incorrect these days). Did they replace the engine coolant, or the refrigerant in the A/C system?
I suspect that it is an issue with the compressor clutch, or its power supply.
Do you know what an A/C compressor looks like? Can you find yours on the engine? If the answers to this are yes, then I can ask you some specific questions that will help nail this down. If the answer is no, then go to a shop that knows auto A/C (it really isn't that hard a subject) because without a little more info all we are doing here is guessing.
When you go to the shop, take copies of your previous bills so the A/C shop can see what was done.

[zenith]

Chrysler uses a coolant pressure switch that interrupts the electricity to the compressor clutch when system pressure is low. It should be in a metal line mounted on or near the firewall of your Jeep and should look like an oil pressure "idiot light" switch ,i.e. kinda dome shaped with a 2-wire connector coming out of the top of the dome.

It works just the opposite of the oil switch, interrupting a circuit at low pressure rather than completing it but it looks very similar nonetheless.

Simple refrigerant pressure gauges that look like tire guages but calibrated differently are available for<$10 at most auto parts stores. It is used in a fitting that looks like a tire valve located in the same line as the pressure switch.

Make a jumper wire. Remove the connector from the pressure switch, jump the two sockets in the connector. Start the engine and turn the AC to "high".

If the AC line sweats or frosts and your suction pressure is approx. 30 psi, then the switch is bad. If pressure is < 30 , pull out the jumper and let the compressor shut down before it damages itself. Low pressure means not enough refrigerant. Thje causes are either a leak or the system improperly charged the last time it was worked on.

The switch should be replaced by someone with a machine that properly depressurizes/recharges the refrigerant because removing it "opens up" the system.

[Rick]

There are a few errors in your post zenith.
First off as I said before the A/C system does not use coolant.

Secondly sweating or frost is largely a function of humidity. If you are in LA in July (or Phoenix, or any number of other places with low humidity) you will see little or no frost. The presence or absence of sweating is at best a very poor indicator of system performance.
Low pressure does not always mean low refrigerant, it means a low temperature at the evaporator. Temperature and pressure are directly related. R-12 was almost exactly 1:1, pressure equaled temp, so if your low side pressure was 28PSI you evaporator temperature was 28F. R-134 is very close but not exactly 1:1 between temperature and pressure. Anyway with a jumper in place, the lowest pressure you will see is a function of air flow across the condenser, humidity, and ambient temp. I have seen 4 PSI on a completely full system with the switch jumpered and the cooling fan running.
R-134 fitting do not look anything like a tire fitting, and I don't think I have ever seen a pencil type gauge for that type of system. here is a picture of a R134 service valve (3rd one down). It looks more like an air line quick disconnect than it does a tire valve. Also the schrader valve is recessed on a R134 fitting, and all the connectors I have seen for R134 have a screw wheel to extend or retract the schrader depressor like this one
Not at all like a tire valve.

[Jinx]

Beyond what has been mentioned, if low freon is one culprit, I understand there is a freon with a dye to help mechanics find leaks int he A/C system (when observed under UV light). - Jinx

[Rick]

Yes this is true, UV dye is routinely added to A/C systems to help find leaks.
However the symptoms described in the OP do not necessarily support a diagnosis of low refrigerant. Low refrigerant would give little or poor cooling all the time. The cools and then blows warm humid air described would make me want to look at the electrical supply to the compressor or an issue with the compressor clutch.

[Eliphalet]

One other thing to try...

I had a Mercury Sable that had similar symptoms. I took it in several times, and finally one mechanic decided to clean out the area between the condensor and the compressor (IIRC). Apparently, leaves and other crap collect between the two, and limit the cooling capacity of the condensor.

$30 and I was out the door.

Just a thought, anyway.

Eli

[Kevbo]

Many possible causes have been mentioned in this thread, but without putting pressure gauges on the system, and monitoring temps and pressures, it is impossible hard to narrow it down to a single cause.

Air conditioning isn't rocket science, but it absolutely requires understanding the entire system and how all the parts interact. You can't usually see anything broken, With gauges you can see the symptoms, but you have to deduce the source of the problem.

Too many "general" mechanics know only enough about AC to be dangerous. It is slightly better these days, as an EPA license is required to work with refrigerants, but most of the license test is about how to comply with law, and very little to do with actually diagnosing or fixing the systems....and there are shady outfits that are equivalent to "diploma mills" so many folks that have the licenses did nothing but pay a fee, and take an open book exam.

And that is why you want a specialist for this work.

Originally Posted by Eliphalet

..finally one mechanic decided to clean out the area between the condensor and the compressor (IIRC).

Change that to between the fan and the evaporator, and it is something I have seen. This last fall, in fact, the leaves packed up on the speed control resistor for the fan and caught fire, filling the interior with smoke, and blowing the thermal fuse on the SCR.

[Gary T]

Originally Posted by Eliphalet

One other thing to try...

I had a Mercury Sable that had similar symptoms. I took it in several times, and finally one mechanic decided to clean out the area between the condensor and the compressor (IIRC). Apparently, leaves and other crap collect between the two, and limit the cooling capacity of the condensor.

That would have been between the condenser and the radiator. Debris there would restrict airflow across both.

Regarding the OP, as many have said the described symptoms don't point clearly to just one most likely cause, but something like the above is a reasonable possibility. Some systems have a high-pressure cut-out that could be activated by the heat/pressure buildup in a condenser that wasn't getting air across it. This would be consistent with the problem ocurring more when the weather is hot and/or the A/C has been running a while.

As some have mentioned, diagnosing this requires more info than we're going to get here. I'd want to see the ambient temperature, the temp of air coming out the A/C ducts, the system pressures, the operation of the condenser fan, the operation of the compressor clutch, and probably a few other things that don't pop to mind but would present themselves if I were standing there by the vehicle. I agree that an A/C specialist, or a competent high-quality shop, is the way to go.

[Rick]

A/C systems work on change of state between a gas and a liquid (condenser) and then from a liquid to a gas (evaporator) anything that interferes with either of these changes has an effect on A/C performance. I have seen dog hair on condensers, dirt packed into condensers, and accident damage to condensers all cause problems. With evaporators I have seen plastic bags, and pieces of paper also cause problems. however none of these cause the symptoms cited in the OP. Going from blowing cold to warm humid air is not caused by a dirty / plugged condenser or evaporator.
kevbo While I agree that a complete diagnosis of an A/C system requires use of pressure gauges, and proper knowledge, the symptoms in this car point more to an electrical issue with the clutch control circuit. Quite likely a voltage drop causing the clutch to disengage when the temps get high. There is a slight possibility that moisture in the refrigerant would cause the symptoms listed, but that in my experience is pretty rare.
Maybe we can get the OP to stop back by and answer some questions.

[zenith]

Originally Posted by Rick

There are a few errors in your post zenith.
First off as I said before the A/C system does not use coolant.

Secondly sweating or frost is largely a function of humidity. If you are in LA in July (or Phoenix, or any number of other places with low humidity) you will see little or no frost. The presence or absence of sweating is at best a very poor indicator of system performance.
Low pressure does not always mean low refrigerant, it means a low temperature at the evaporator. Temperature and pressure are directly related. R-12 was almost exactly 1:1, pressure equaled temp, so if your low side pressure was 28PSI you evaporator temperature was 28F. R-134 is very close but not exactly 1:1 between temperature and pressure. Anyway with a jumper in place, the lowest pressure you will see is a function of air flow across the condenser, humidity, and ambient temp. I have seen 4 PSI on a completely full system with the switch jumpered and the cooling fan running.
R-134 fitting do not look anything like a tire fitting, and I don't think I have ever seen a pencil type gauge for that type of system. here is a picture of a R134 service valve (3rd one down). It looks more like an air line quick disconnect than it does a tire valve. Also the schrader valve is recessed on a R134 fitting, and all the connectors I have seen for R134 have a screw wheel to extend or retract the schrader depressor like this one
Not at all like a tire valve.

Sorry, I misread the OP. In a 1998 Jeep, the refrigerant used would be R134. My reply referenced pressures and fittings used prior to Chrysler's 1993 changeover from R12. I actually found a bad REFRIGERANT pressure switch in an R12 system in a 1991 Chrysler product using the method I detailed, but R134 IS a different animal and does use different fittings and thus my old tire-valve type guage won't work. Also, you're right that very little sweat/frost will form at low humidity. As a resident of a place that's nearly 100% RH nearly 100% of the summer,I forget that some people from arid climes actually think that 95 degrees@ 10% humidity is hot while I think it feels no worse than 70 feels back home.

Bad editing left one mistaken reference to "coolant" when I meant "refrigerant".

I need to more closely RTFQ and more carefuly edit replies in the future.