Long and Covoluted intermittant A/C Problems with '88 300TE
 

I have spent months trying to diagnose this problem, and spent quite a bit of time searching the AC threads on this grate site... Each time I think I have found the source of the problem, it turns out not to be it. So, here is the LONG version of the story:

I purchased this 1988 300TE (w124.090, engine 103.983) to be my primary commuter car when I lived in Kansas. It worked great, and I had no complaints. Then I moved. Two hours down the road, the AC cuts out. The cold dry air coming out the vents changed over to cool moist air, and after a little while, to hot dry air.

The first diagnosis was low freon. I could not find anyone in NW Iowa (where I was many miles from a major metro area) who was willing/able to actually do the recharge on this car in the time I had in the area. A few days later, I headed north to visit family in Northern MN, and the AC worked the whole trip (Cloudy day, relatively cool dry air). I got the system recharged up there - but I found out later they managed to botch the job, and didn't get the fittings properly sealed, about half my R134a leaked out within about two weeks - but by that point I was in Virginia, and had to pay someone else to re-do the fitting and recharge. On my way out of MN, it was a hot humid day, and the AC system cut out after about 20 minutes on the road. Further drive testing determined that at full blast, the AC would only run for about 20 minutes (from a cold engine), but on a minimum cool setting, it would run indefininately (at least at first - before the refrigerant started to get low a week or so later). By the time I was commuting in Virginia, the AC would only run for about 5-10 minutes (much hotter and more humid than on the Prarie or Plains). I took it into a shop in Virginia, and they very quickly diagnosed the leak - it was almost visible to the naked eye (even I could see the yellow dye staining everything around the low pressure fitting). They drained the system, replaced the fitting, and recharged it. That's where things start to get interesting.........

I picked up the car and drove it to another place in the county to register it (I'd only lived in the couny for about a week), and sure enough, about 20 minutes after I left the shop, the AC stopped blowing cold dry and started blowing cool moist. I turned right around and went back to the shop. They kept the car all day the next day - running the car on idle most of the day trying to force the AC to fail. It didn't. Then they got in the car to drive it, and it cut out within a block. Temps both days were high, and the humidity similar. It was the edge of a long weekend, and they woudn't be able to do much for the weekend anyway, so I was going to take the car and drop it off again the next week - except it wouldn't start. It seems the battery was dead. A quick jump-start and check of the electrical system with a voltometer, and it becomes immediately apparent that the Alternator is only putting out about 12.2V even at high RPM (e.g. not enough to run/recharge everything properly). A replaced alternator later, everything but the AC is running much better (I had been having a few other minor annoyances that were evidently electrical in nature).

So, I take the car home, and pull the wiring diagrams off the shop manuals, and start trying to trace a fault in the control lines (I'm an Engineer myself, so while this vehicle is unfamiliar, the concepts are not). I cannot find one. All the electrical lines seem to be working properly - the compressor is getting shut off after about 20 minutes of operation, and I'm stumped as to why. As I look more and more into the operation of the AC system, I suddenly realize I'm low on engine coolant - so I top it off, and immeidately take a test drive. The engine is now getting cooled better, so the engine coolant switch doesn't cut off the compressor as often, and for about an hour I can recognize the compressor cutting out at high engine temp, and turning back on once the engine cools back down. No problem. It seems to be fixed. It was a HOT and dry day.

The next day it's cool and rainy, and the AC cuts out after about 20 minutes.
:confused:

So, I start checking all sorts of things, even asked a family member who is a retired GM mechanic (whom I had not spoken with in over a dozen years). I explain the whole problem to him, (as I have above), and he thinks it's either a) the dryer crystals are shot, and I need a new dryer, or b) my condensation drain may be plugged. The Humidity (both inside and out) is what he keyed off of. If I have a plugged condensation drain, then the water will back up and reduce the air-flow over the system, and not permit the proper heat-exchange, thus freezing the system and forcing the compressor to shut down. He suggested I put a temperature probe in the vent to see what temp the air coming out is. I did so the next morning, and every day since. The temp runs down to about 43.0°F at minimum. I have seen it as low as 41.7°F. The AC seems to operate between about 40°F and 50°F, but once it gets much above 50°F it has lost compression, and rapidly climbs to the ambient outside air temp, as well as becoming massively humid.

As soon as he mentioned a possible water build up, I remembered that I had heard on occasion what sounded like water sloshing around in my dashboard - but I have found no mositure inside that could be associated with it (I have found some moisture near the hatch in the back, but that is following rain and is coming in through a crack in my antenna base). I also have not seen any water dripping beneath the car, at all. Ever. He suggested I take a coat hanger or other wire and push it up through the drain hose to see if I can clear the blockage.

I did some checking, and found what looked just like a section of bycycle tire inner tube (wide and flat) coming down from about the midline of the car, directly beneath the windshield, and routing around the top of the transmission to the driver's side. It does not appear to be a straight-shot from the hose opening to the top, and I am really nervous about the idea of probing around in there with a coat hangar. I did attempt to probe it with a 1/4" plastic/nylon hose (clear, so I can see immediately what's in it, but soft enough so that I couldn't do any real damage to anything with it). I was unsuccessful.

Searching on this site seems to tell me that indeed I have a blocked condensation drain, but mine does not appear to have the insulation on it that I have seen described elsewhere, nor do I have a clear path forward from here. My read on it is this: I have a clogged condensation drain, and I need to get that water out of there, regardless of if it fixes the AC problems or not. I am an Engineer, but I do electronics and software, not mechanical, so this is kinda outside my realm of expertise. I do understand the basic concepts, and with a good set of instructions I can probably do the repair myself, but I'm missing a disk from my shop manuals on the 124 series (Gotta love moving and having everything in boxes), so I do not have any idea where to start on going in from the pasenger compartment side, nor do I really feel good about poking with a coat-hangar from the undersdie. I do understand that I may have a wrong diagnosis, but I do need to get that water out regardless - unless I am supposed to have a water-sloshing sound in my dashboard in this particular model year (1988 300TE).

It has been both hot and warm, but consistantly humid every day since the system worked for as long as I tested it; and has consistantly cut out at between 18 and 25 minutes after first getting the vent temps down into the 40s. I did have one day where after the system cut out, the humidity in the car built up to over 95% (I have a humidity meter on my probe thermometer), so I get both temp and humidity).

If you've gotten this far, thank you for your patience in reading my ramblings, and thanks in advance for any advice you can give me. I will stand ready to answer any questions, and I throw myself upon the expertise of this site.

CTM

I've not heard of the condensate drains becoming blocked on a 124 chassis car. There are two of them, they exit out the bottom of the HVAC box, they are large diameter, and proceed straight through the transmission tunnel without and turns. See the attached photo of my '87 TD wagon receiving a new evaporator. The holes through which the condensate drains exit are circled. For reference, the gold box just rearward of the holes (the airbag trigger) is directly below the radio.

At any rate, it's easy enough to check. Idle the car in place, windows open, on a humid day. You should accumulate two puddles of water under the car, one on either side of the transmission, just aft of the front wheels.

As for the cutting out after use, I suggest you check the clutch gap on the compressor. It should be 0.50mm +-.15mm. The gap increases over time as the clutch material wears. I find that once it reaches ~.80mm or so clutch engagement becomes inconsistent. And yes, one sign of excess clutch gap is when the compressor works fine from cold, but becomes sporadic after it warms up. Any decent auto a/c shop will have a kit with a selection of new spacers to reset the gap to spec.

Another possibility is the presence of grease/oil on the compressor clutch facing. This situation can cause enough slip when engaging the clutch that the safety cutout it triggered. Try cleaning the clutch facing using Simple Green, brake cleaner, etc.

Best of luck,

- JimY

I have checked. No puddles.

two things come to mind [1] have you tried shutting down the engine when the compressor shuts off and restarting the car to see if the a/c restarts.if it does,then you may have slightly loose serpentine belt or belt tensioner issues.[2]what about the evap temp sensor,when they shut down the issue is that they won't turn back on[supposed to turn off at about 34-36deg evap temp and come back on at about 40deg]based on your description #2 could explain all your problems.

I have tried shutting down the engine and restarting. It has worked once or twice, but only gets the AC rolling for about a minute or two. Is there anything I can check at a hands on level (i.e. no special equipment) to see if the belt is tight enough? I can see the belts running without apparently slipping both when the AC is working and when it isn't.

I will have to double-check on the evap temp sensor, and get back to you.

if it restarts then it's probably the compressor speed sensor shutting down the relay.quick and dirty test is to see if you can lift the belt by hand and take it completely off the upper pulley in the middle.by the way you can take out the a/c relay from behind the batt and use wire to join terminals 15 and 87 and plug back in.then drive to assure yourself that nothing wrong mechanically only electric.but in this state the a/c compr will always run so remember not to shut down blower or may freeze up coil.

Ok, did a quick check on the belt, and it has about 1"-1.5" of play in it across the top of the block. Wouldn't a halfway decent mechanic have discovered the belt was loose when they replaced the alternator two weeks ago? Is that too much play, or just right?

I'll try the relay bit this evening, just to test it out.

CTM

yes it should have been drawn to your attention by any one of the several techs that looked at your car unless you were pleading poor to them.

So is the water sloshing sound a red-herring, or is that an unrleated problem I need to look into seperately?

The ordinary mechanic not familiar with MB might not have known about the speed sensor and its function.

However, I assume you had the AC problem both before and after the alternator was replaced.

The tensioner is troublesome and they might not have been able to adjust the belt properly.

For the drains, get a piece of wire (like a coat hangar wire) and GENTLY push it up in the drains. Gently because the hoses are nothing but a spring wrapped in foam.

Ok... did some more checking this evening. Crawling under the car, I found the drain hoses - quite deteriorated. I know that one was at least partially obstructed, but doing some gentle probing I cleared any evident blockages. When I idled the car, I started getting condensation drip (quite a bit, actually) at 10 minutes, but the system still stopped functioning at about 20 minutes. I actually watched the compressor stop, and it never came back on. I stopped the engine, and restarted it. The AC compressor did not engage. I stopped the engine, and let it sit for a few minutes, and when I restarted, the AC compressor still did not start.

I did not get any additional time to check by jumpering the relay as suggested, due to a pending thunderstorm (don't have a garage here like I did in Kansas). Nor did I get to do more than more fully examine the belt. The belt has about 3/4" of deflection midway between the coolant pump and belt pulley. I can feel the tensioner resisting my push/pull. I did locate the tension adjustment control, but didn't have a chance to try increasing it.

During the test run, I never saw any indication of the compressor slipping, it simply ran until it was shut off. The engine temp never really got high (I kept it in idle, and had my wife occasionally throttle the engine. During this test run, the ambient air temp was about 90°F with about 45% humidity. Vent temps barely got down into the 50s, but humidity in the cabin got down to about 35%. Once the compressor shut off, vent temps spiked quickly into the 70s and 80s, and humidity in the cabin climbed above 60%.

Unless someone tells me not to, I will jumper the relay in the morning, and re-test. I don't get the feeling it's a loose belt, but I will deferr to the expertise on this forum.

Another round of testing, and still no success.

Ambient Air Temp: 85°
Ambient Humidity: 45%

Engine starts, AC runs. Five minutes after startup, engine temp is running slightly above 80°C, engine oil pressue is not quite pegged, idle RPM is about 900. I can see the compressor is engaged, vent temps are down to 63°F (from an original 122°F sitting in the sun all day - original internal humidity was 29%), and internal humidity is at about 30%. The beginnings of a condensation puddle is forming on the passenger side of the transmission. All is well.

I run the car in a high idle (RPM 2500).

Ten minutes after startup, engine temp is just below the unlabled tic-mark between 80°C and 120°C (I take this to be about 100°C). When I return the engine to idle, RPMs drop to below 1000 and the oil pressue drops to below pegged. AC Compressor is still engaged, and there is a fair bit of condensation forming along the low-pressue lines - as well as in two puddles beneath the car. Vent temps are down to 58°F, and humidity is staying around 30%.

Fifteen minutes after startup, once again following a high-idle run, engine temp is not significantly changed, engine idle is down to about 800, and oil pressue is down to about the middle of the guage (between the 1 and the 2). Vent temps are down to 54°F and humidity is down to 27%.

Eighteen minutes after startup (right on time) the compressor shuts off. Engine temp is unchanged, idle oil pressue is slightly higher (just under the 2), and RPM has returned to 900. Vent temps spike quickly into the 70s, and humidity is above 48%.

An immediate engine shut-off and restart does not re-engage the compressor.

I shut the engine off, and remove the Klima Relay, and jumper terminals 15 and 87, and reinstall it.

Start the engine, and the compressor does not re-engage. Vent temps are running above 80°F and humidity is running above 60%. Engine was off for about 5 minutes during this jumpering.

Engine temp has fallen to about 90°C, and RPMs are running about 900 at idle, with oil pressue at the 2 mark.

I have removed the jumper, and will re-try the AC system in a few minutes - giving it a 30 minute break for a re-start.

I can clearly see in photos I took during this procedure (I will post in a bit) the compressor clutch engaged and dis-engaged. This indicates to me (I may be in error) that it is not a slipping clutch. I never saw the RPMs fall below 600, though they did get down to about 800 at one point. If it was just the RPM cutoff, the jumpering should have allowed the compressor to operate regardless. It was not even attempting to run.

Any other suggestions?

You could have a high pressure cutoff- this could be caused by either a clogged expansion valve or a deteriorated hose that has collapsed upon itself, shutting off the flow of refrigerant. It could also be a drier plugged with leak sealer. The thing to do is to jump out (very temporarily) the high pressure cutoff switch when the failure occurs and see if the compressor starts. If it doesn't start, ohm out the clutch coil while it is hot, it could be an intermittent coil. You should be able to locate the lack of voltage to the clutch fairly quickly, I'd think...

All the hoses and lines appear to be intact. Would it not be better to get an AC Mechanic to do a pressue guage test (i.e. run the system to failure while high and low are hooked up to guages) to see what's going on?

CTM

Your KLIMA relay has pins that are labeled "15" and "87"?

Are you sure you removed the KLIMA relay? According to my ETM jumping pins 5 & 7 would bypass the KLIMA relay and manually activate the A/C compressor.

I think 5 & 7 are the socket numbers where the Klima relay plugs in, 15 & 87 are the pin numbers on the Klima relay itself. If memory serves (usually doesn't in my case...) socket 5 is +12v and socket 7 is the compressor coil.

Got a volt/ohm meter? Check the continuity through the compressor coil to ground, both when cold and the a/c is working, and also when hot and it stops working. I'm starting to think the coil is going open when hot.

Of course, these symptoms also fit an excess compressor clutch gap, which you haven't checked yet. :D

I don't think it's the compressor-engine speed comparator circuit because the compressor never comes back on after a restart. If it cuts out once it'll do it again, but frequently runs for a minute or two first.

To be thorough I'd also check to see if the CC head unit is calling for compressor operation when it cuts out. I forget which pin this is on the Klima socket; it's on the wiring diagram. Should see +12v on this pin when a/c is called. Easy to test by switching between a/c and econ.

- JimY

Pin #10 (on the KLIMA socket) is the activation signal from the PBU, I can't from the diagram if its a +12V signal or a switched ground.

I know the ping 10 activation signal is ground on a diesel 124 - BTDT. I've been told it's +12v on a gas 124. The two cars use different part numbers for the CC head unit, so I tend to believe this information.

- JimY

Ok, another round of testing/feedback results:

Ambient Air Temp 66°F
Ambient Humidity: 88%

AC runs as normal, bringing vent temps down to about 45°F and cabin humidity down to about 25% before cutting out at about 20 minutes. Stop and go traffic (more stop than go).

Ambient Air Temp: 85°F
Ambient Humidity: 36%

AC runs not quite as cool, vent temps only reach down to about 54°F before cutting out. Humidity stays between 30% and 40% until system failure at 17 minutes. Mostly stopped traffic. AC cut out just as I got off the surface roads and onto the high speed road.

Ambient Air Temp: 65°F
Ambient Humidity: 45%

AC runs not quite as cool at first, but gets well and truly cold (43°F) after about 15 minutes. Humidity never gets below 40%, and the AC runs until the 36 minute mark, then cuts out. Stop and go traffic. This is the first time since I've been doing detailed monitoring that it has run for more than 20 minutes and still cut out (I have had it run without failure for an hour or more, the only reason it stopped was I turned off the engine when I reached my destination).

I have unburried my multimeter, and will be running checks on all the relay connections this evening.

http://catalog.worldpac.com/mercedesshop/sophio/wizard.jsp?partner=mercedesshop&clientid=catalog.mercedesshop&baseurl=http://catalog.peachparts.com/&cookieid=23I0RAMEE23W0MC6KZ&year=1988&make=MB&model=300-E-002&category=R&part=A%2FC+Pressure+Switch

Is the aux fan coming on low speed??
IF NO , check this part...you may have too much high side pressure from no fan, cutting out compressor.
Or, if this sw is cutting the comp out, you could also have an over-charge.
You should have aux fan operation with either condition.
Do you have fan?

I do still have the aux fan. It (appears to anyway) operate occasionally, especially when the engine is hot, and in hot ambient conditions. I have not yet checked voltage switching on the existing pressure switch, that is a task for this evening.

My 87 had no low speed fan due to burned out resistor. Made a huge difference in both AC cooling and engine temp. I don't want to ever run under conditions that trigger the high speed function of the fan :D

Ok
Just be aware that the aux fan can be triggered by both engine coolant temp [ high fan] and the high pressure sw at the drier [ low fan] . There are 2 switches at the drier .The one I posted for you is a dual cut-out sw [ as you can see in the part discription] . That cuts out if the refrigerant charge is too low so that the comp will not come on w/o the system having a charge, but it also cuts-out if the high pressure gets too high ..so, this is one of the saftey switches to protect the compressor..the other comp cut-out one is a coolant temp sw that cuts the comp out if the engine coolant gets too high, and then there is the speed one in case the belt/clutch slips .
There is another sw at the drier [ the one on the side with the pigtail wire harness] that controls the low fan. If you jumper that switch w/key ON, you should see low fan.
This sw has nothing to do with the comp circuit, it only triggers the fan, BUT, if that fan does not come on when the pressures start to get high, then the sw on top of the drier will reach cut-in pressure due to no airflow from fan, resulting in a comp cut-out. So, you can see that the aux fan is needed to keep the high side pressure in check so it never reachedes the comp cut-out spec of the top sw.
Best to use gauges to see what the high side pressures actually are , but if the comp cuts-out and you see that the reason for it is b/c of the sw I posted has opened , you then know you have too much high side pressure..from an over-charge or a NO FAN condition
Just two easy things to check that would fit your condition/diagnosis.
Verify fan circuit w/jumper test..and check high pressure comp cut-out switch w/ohm meter/ continuity test when condition exist.

Exactly...and that is what will happen if the low fan circuit is not operating correctly..
Common, misdiagnosed problem...alway check low fan circuit on a/c and engine over-heat problems ..specially when it take 2 seconds to verify the low fan circuit with a simple paper clip.

Have you tried jumping out the high pressure cutoff switch yet?? Have you ohmed out the clutch coil when the compressor fails yet?? You already know that it fails, why don't you check these items?? As Arthur says, jumping out the high pressure switch only takes a few seconds, and ohming the coil only takes a few more...

As Ricrard states, all/any of his conditions can cause your problem.. the main reason I am leaning in the aux fan / overcharge possible is b/c you do not seem to have the problem until you loose airflow from a non-moving vehicle/traffric/idle condition .. that points me towards conderser airflow vs. blockage drier/lines/etc. .your comp. cut-out seems to be the RESULTS of high pressure run away due to lack of airflow...but, the test wll tell the tale.
If the fan circuit verification test passes, then you may also want to look for debries between the cond and radiator...but fan jumper sw test is still top of my list..you mention seeing fan at hot engine, but I think you are mis-interputing that as FAN OK, not realizing that coolant high fan has NOTHING to do with a/c LOW FAN operation..two completely different circuits that just share the same fan motor. LOW FAN jumper test must be done to verify the electrical intergrity of that circuit. If no low fan with that test, you could have a problem as simple as a fuse or burnt wire at the resistor.

Ahh, but that requires the multimeter which I have not had unpacked before today. That is in fact my plan for this evening.

The failure symptoms can be quite deceiving on some of these A/C problems - I had a problem on my '82 380SL where the thing would blow warm to hot air even with the AC set to max cool (knob at cold end detent), but once the car was moving it would cool down pretty well. The problem turned out to be that the auxiliary water pump wasn't working! The auxiliary water pump is necessary in order for the monovalve to operate correctly, ie, it works on water pressure, like the solenoid valve in your washing machine works... a small primary valve is operated by the solenoid, which moves the larger diaphram using the water pressure... in my case, I eliminated both the auxiliary water pump and the monovalve and replaced them with a vacuum operated water valve operated by an electrically operated vacuum solenoid, similar to the other solenoids in the system that operate the flap valves. Now it works great all the time- PLUS, the car runs much cooler, as it isn't trying to both heat air and cool air at the same time.
I'm sure you will make progress once you perform the two simple tests...:)

Ok, didn't get anywhere NEAR as much checked out as I wanted to before the severe thunderstorms rolled in last evening, but I did check the pressue switch (S31/1 - Pressue Switch, refrigerant compressor according to the wiring diagram, but it is the component that was linked above) - which is showing a complete circuit both in operation and at failure (no change between states). However, S32 (Pressue Switch, Auxillary fan) changed from a complete circuit at operation to open circuit at fialure. This is NOT the picture linked above, but the wiring diagram says it is the one that runs the Aux Fan. The Aux fan ceased operating at that point as well (having previously run the entire time the engine had been running).

I did not get a chance to measure coil resistance (or even locate where to do that) before the thunderstorm came up. Maybe today...

It's normal for the electric fan to turn off shortly after the compressor stops. The fan switch is triggered by high side pressure. No compressor, no high side pressure. The fan should not shut off at exactly the same moment the compressor stops, a period of time (few seconds to a few minutes, depending on about a million factors) is required for the high side pressure to drift down.

Easiest location to check the clutch coil resistance is by pulling the Klima relay. I think you want to check continuity & resistance between pin 7 & ground - but double check that pin number on your wiring diagram.

The coil should show about 5 ohms resistance.

- JimY

As I understand it, you are saying that S31 does not change states when the failure occurs, but shortly after the failure occurs the aux fan shuts down. Now the component to check is S25/5, the coolant temperature switch. Check the output "b" line, the brown/violet wire to see if it has changed states. It feeds a signal to the compressor cutout control unit if the engine temperature has exceeded 115 degrees centigrade that shuts off the compressor. If it is ok, your compressor cutout control unit is faulty or your ETR switch is signaling that the evaporator temperature is to cold, which will also turn off the compressor. You can monitor the voltage at S31 to check the ETR switch output. Another, but remote possibility is that the signal from the idle control unit is going bad. The compressor is also cutout at engine speeds above 5750 rpm and turned back on at speeds below 5550 rpm. You can monitor this signal at the compressor cutout control unit, pin 4.

Right. S31/1 never changes state from engine off and key out of the ignition all the way through the test and well beyond failure.

S32 changes state immediately at failure (I'll have to test again to see if it switches before or after the compressor switches off), and does not change state back.

If the weather cooperates, I'll test some more this evening - though from the look of the sky, that is not likely.

So, That takes the aux fan out of the diagnosis..
..also look at the wires at the comp closely..sometimes the speed sensor/coil wires get funky ...
How did you confirm s31 ???
and... Are you still workimg w/R12??

I attached to the leads coming out of S31/1 and S32 (they are adjacent connections) and monitored their continuity during the test.

The system was converted to R134a some time ago, I only wish I were running with R12.

OK .. the reason I asked was b/c once the comp shuts down , the high side pressure drops and could , in fact , close the high side sw due to the drop in pressure ..one wants to monitor the sw while the comp cuts out to make sure the sw has not reached that cut-out point at the time of occurance..it is possible to have the sw cut the comp and then you go and do a cont test on the sw and find good cont. only b/c the high side has dropped pressure in that time frame and you assume the sw never reached cut-out .
Using 134a with the orig pressure sw can run into such a problem b/c of the higher pressure on 134a ...your condition seems to be right at that high pressure , high heat , same running time , so that all has to be considered .. I agree the problem seems to be in collant temp or ETR cut-outs , but just want to make sure you were aware of the possibility of testing the pressure sw after comp cut-out could fool you..
Your refrig charge has to be decent b/c you do have aux fan trigger w/comp is running and it goes off when high side drops..if s31 were to reach cut-out , overcharge would still be a viable diagnosis..
One more item I would check would be the fuses .. intermittent cut-oot problems can be as simple as a poor connection at the relay feed fuse..you have the schematic, so check out what ones are in that circuit and give them a twist in their holder ..Benz fuses are notorious for making intermittent contact at the bullit ends and a simple twist will cure it..not saying that is your problem, just something that does happen ..

Ok, measured the Coil Resistance today - it was about 8.5 ohms.

Some additional poking around turned up a broken component that I cannot identify (pic follows). It is located by components of the AC system, but I do not know if it is a part of it or not. It appears to be a ceramic coated resistor, but the ceramic coating has broken, and there is evident rust on the inside. In both pics, the bottom of the picture is towards the front of the car. Is this part of the AC system or something else, and how quickly should I replace it?

The broken thingy is the resistor for the low speed operation of the electric radiator fan. Low speed is triggered by the high pressure switch on the a/c drier - the red or green one. When the high side pressure in the a/c system reaches the setpoint, the electric fan is triggered on low speed. This mostly happens when stopped or driving at low speeds. You can test the low speed operation of the fan by disconnecting the two spade lugs from the red/green switch on the drier and shorting them together. Fan should run, assuming ignition is in the run position. This really oughta be working. Without it the high side pressure stresses the a/c compressor. You also stay more comfortable - more air moving across the condensor means more cooling inside the car.

The coil resistance sounds too high. It would only be drawing ~1.6 amps, which is about half what I recall. I suggest you run the a/c to failure then retest the coil resistance. I wouldn't be surprised if it has gone infinite at that point.

That resistance was at failure. I plan on replacing the resistor (would that be a dealer-specific part, or would I likely be able to pick it up from the numerous auto-parts stores in this area?) today, and doing a logic-circuit test (read the voltage in real-time on as many of the control lines as possible) to determine what is failing when.

troubleshooting
 

i have been following this thread and trying to develop a convenient way to troubleshoot.
here is my question,has anyone tried to eliminate both the compressor speed signal and the engine speed signal for the purpose of determining if the compressor would run in the auto mode?
has anyone got access to the internal drawings of the kilma relay?
in most cases ac signals are rectified and converted to a pure dc value for the comparator circuits. it sure would be convenient to eliminate this possibility when troubleshooting the system.
by the way some one asked about the signal level at n22 pin 1 from s31/1 and it is high when compressor is off and pulled to ground when on.
also i am working on a 88 300te,86 300e and both have the same problem,at times wont cycle back on when in the auto mode but will run continously in the manual mode.
larry perkins lou ky

Seems like a whole lot of diagnosis work here....has anything been replaced? I would work under the assumption, since the AC does work - all be it sporadically, that it is probably a broken solder connection in the CCU in the dash (the push button unit). My 87 560SL was doing the same thing your car is doing. Compressor would come on and then shut down, the fan would come on and then stop for no reason. It was maddening!
Replaced the CCU with a rebuilt one from Benz Barn our of Georgia for $135.....everything has worked great since - even got the vents to work again (was so afraid it was a heater box thing!).
Even if that part does not solve your problem it will take a big known problem component out of the diagnosis and get you some new shiney buttons!:D
Good luck!

I have not had any time in the last week or so to do any more diagnostics - though I did take some down time during my Rescue Shift to replace my aux-fan series resistor (which was quite well mangled. If I get a chance, I'll post pics.

The only change I've noticed in the last few days is that I do not even get 15 minutes of cooling now (down from the 20-25 I was getting a few weeks ago) - this tells me I'm probably chasing a freon leak. I've not been able to see any visible dye tracer, but I'll be hunting it with a black-light when I get a chance this week.

ctm,arthur called mine
 

i tell u that fellow is good,got mine to my steam cleaner,washed that sucker real good around the a/c compressor,ran a 1/4 tank of fuel out yesterday and the ac cycled at any position i put it.
i really didnt want to post so fast as i will be driving this week in 94f temp but it sure looks like the signal from the ac speed sensor is getting clipped down to a level that wont operate the klima relay,with the crud becomming more conductive as the engine heats up(shorts across the output wires)
iin reference to my earlier subject on how to troubleshoot the system and bypass the compressor speed sensor,i wanted to be able to keep the relay in place for the elimination process. i know about the output contacts and all that.
ctm i bet a car wash will do the trick also if you dont have a steam cleaner. let us know.
larry perkins lou ky

Let me see if I understand you correctly: You are saying that I should take my engine and get it steam-cleaned, with a specific focus on and around the a/c compressor, and that should fix the problem?

Considering the information you have given earlier in this thread, what makes you think you have a freon leak? Has the weather been warmer in the last few days? If it has, it would indicate you are getting to a high pressure cutout sooner, which could be the result of an overcharge! The way to tell if you have a low freon charge is to connect a freon gage set and verify pressures. I'm beginning to think you should just take it to a qualified AC tech and get it fixed, as both your reasoning and diagnostic techniques don't seem to be getting you to a fix. Summer may be over before you arrive at the problem.

It hasn't been warmer, it has infact been cooler, and yet I've been getting shorter and shorter running time.

I am, at this point, planning on taking it to an AC tech to get it fixed, but I'm waiting for cash-flow from my move and job-change to catch up (gotta love burocratic paperwork). Should be in a couple weeks, assuming nothing else strange comes up (Like having to replace major components of my wife's car because a previous mechanic couldn't do a simple job correctly!). My car is driveable, the A/C is (at the moment) a luxury item that I do not NEED to get back and forth to work with. Once everything that I NEED to have functioning taken care of, I'll start paying for the things I'd LIKE to have. That said, at my own pace, I expect summer will be over before I arrive at the source of the problem. So be it.

I did not read thru all this so forgive my basic question if this has been answered.

Do you have bubbles in the sight glass?

I don't think so. The sight-glass is quite dirty, and difficult to clean enough to see anything.

Clean it with the eraser on the end of a pencil.

Ok. I cleaned the sight-glass, replaced the Aux Fan Series Resistor, cleaned the electrical connections, and did a spray-cleaning of my engine (with extra attention to the AC compressor).

The AC now works indefiniately at MAX cool and MAX fan in HOT weather.

Any other condition causes it to fail at between 10 and 20 minutes.

In 95°F ambient temps, with the fan on MAX and the temp on MAX COOL the vent temps never get much below 50°F, but it runs for as long as I need it (or even longer than I can stand - I don't need the cabin to be in the 50's when I'm dressed for a nearly 100°F day). IF I change any setting, the compressor shuts off very quickly. If I start with the fan on the minimum blow setting, the vent temps get down into the 40's, and the compressor shuts off after about 15-20 minutes.

In 75°F ambient temps, with the fan on MAX and the temp on MAX COOL the vent temps get down into the low 40s, and the compressor shuts off not long after getting the vent temps below 43°F (typically about 15-20 minutes).

In temps below 70°F, with the fan on MAX and the temp on MAX COOL the vent temps very quickly get into the low 40's, and the compressor shuts off not long thereafter - sometimes in as little as 5-8 minutes.

My assesment:
I have corrected several problems that were contributing to the problem, and I may be down to the last problem, or handful of problems.
The key issue I am seeing now is one that I believe to be thermodynamic - if I keep the system running at full blast in hot weather it never chills below a critical point somewhere within the system - thus it never shuts off the compressor. If I do not keep enough fresh HOT air blowing over the system (and thus keeping the system above this critical temp) the system will chill below the threshhold, and the compressor will shut off.

I take this to mean that I may have moisture inside my system (freezing and blocking a line, thus driving pressures to trigger a safety cutoff), or something similar.

Or am I really out in left field? I'm taking care of my forward Flex-Coupling this month, so professional AC repairs are not in the budget until next month.

have you checked or bypassed the etr switch?