With the engine off clean the compressor clutch with carb cleaner. Allow to dry. Test A/C system. My 91 300TE had a slight leak at the front compressor seal which deposited oil on the clutch and caused the the clutch to slip just enough for the rpm sensor to detect and disable the system until the next start cycle. I had to repeat the cleaning every year until I ultimately replaced the compressor before selling the car to a neighbor. Mark

Hi,
I'm in South FL. The car has 280K. I don't know if the compressor is original or not. I know the evaporator has been replaced. Possibly the compressor. I can't figure out how to tension the belt. There is supposed to be some type of pointer, but it's not apparant to me. I tightened the new belt till it had about 1/4" deflection. Seems pretty tight to me. The new belt didn't change anything. Is there a way to bypass the speed sensor to see if that's it? I haven't timed the cycle, but it seems regular as clockwork from start to stop. That seems like it would be a good clue for anyone who understands these systems. I think if I could time it with a watch and it would be very close each time. I'd obviously like for it not to be the compressor. I certainly would hate to replace it and it not be it!

You really need the FSM wiring diagram to understand the system, but I'll do my best, leaving out much of the description since the system IS turning on the compressor to start with.

The Climate Control Unit (CCU) monitors about a half dozen temperature sensors and sends a signal THROUGH the low pressure switch to the Klima (Deutsche for climate) relay behind the battery. The Klima relay monitors the speed sensor and the signal that comes through the low pressure switch and decides when to energize the contacts which engage the compressor clutch.

The monitoring of the RPM sensor is done as magic inside the Klima relay. One can bypass the speed sensoring function by using a standard ice cube relay in place of the Klima relay, but it takes some innovative electric fabrication work due to it's location. I have never done it, but you could gut the Klima relay and use it as a mount for the ice cube relay. That way you could, on the bench, do the soldering in place of the relay or a relay socket onto the Klima relay base, then plug it into its original location.

Before doing all this as a shot in the dark, make sure that you are getting a signal from the low pressure switch to the Klima relay and still have that signal after the clutch has kicked out. This is where you need the FSM wiring diagram to know the pinout for such testing, and of course how to wire up the ice cube.

Hope this mess helps.

Actually it does help, some. I have Alldata and the Factory manuals I believe. The main thing is that they are for mechanics, real mechanics, so they assume you know a lot. AC stuff is new to me because you just don't fool with it a lot. I'm thinking that a lot of the wiring harness troubleshooting is already done because the Compressor starts everytime you start the engine.

There is an ohms spec for that RPM speed sensor, but I guess what you're saying is it's not the sensor because it starts up each time?

I just checked the compressor price and it's nowhere near as expensive as i thought it would be. I've been fooling with Porsches too mIt sounds like uch!

Listen to Larry - he has the right advice here. Before messing with the expensive, complicated stuff, check the simple.

First of all, you root cause very much appears to be that you are triggering the compressor clutch cutout. The speed sensor on the compressor is used to compare the compressor RPM to the engine RPM. Too much difference for too long, and the compressor is disabled until the car is restarted. If you can get the compressor going immediately by shutting off and restarting the engine, this is almost certainly your root cause.

Next, of course, is the question of why the cutout is tripping.

Start with the easy stuff. Clean the clutch surfaces. A little bit of slipping when the clutch engages is sufficient to trigger the cutout. Try giving the clutch a little more bite by cleaning off any residual oil/grease/gunk. Just spray your favorite degreaser into the air gap (engine off, please). Simple green, brake cleaner, Mr. Bubble - whatever. Wait five minutes, rinse a little bit with water, give it a try.

Second, check the air gap between the front clutch plate and the driven pulley. You're an old Porsche guy? Good - you must have a set of feeler guages laying around. As the clutch wears, the air gap increases, and the opportunity for slippage increases. The gap is specified at ~0.5mm. My anecdotal observation is that once it increases to ~0.8mm, clutch engagement will become iffy. If the gap is too large, you can purchase a kit to reset it from a good auto a/c specialty supplier.

Let us know if the clutch gap is too large and we can advise on procedure to reset. It's $5, 1 hour, and no need to disassemble & discharge/recharge the system.

- JimY

Thanx! I degunked this pig yesterday. I'll concentrate on the clutch and measure the gap. This thing cuts out regular as clockwork. I never runs more than a few minutes. But conversely, it always runs for a while. I can always drive about 1/2 mile before it stops. It does it at idle or driving in about the same time interval. What if the cutout time was the exact same number of seconds each time? because that's what it seems like to me. How would that factor in?

There's nothing related to clutch cycling in the climate control system that is time based. Generally what happens is that clutch cyling is triggered normally and the compressor triggers the cutout function when it is switched on again. There is more resistance to starting the compressor when the high side pressure is up than when the car has been sitting.

Does the interior fan slow down as the cabin approaches the set temperature, then it cuts out? Or does it cutout while blowing gale force? The former would be expected and point toward the clutch slipping. The latter case is more inidicative of a control problem where the compressor is intentionally turned off due to a fault elsewhere.

In the last year or two I recall a discussion about a very difficult to solve cutout on an older 124. It was similar to your situation in that it was a very consistent time from when the car was started until the a/c shut off. None of the usual suspects were to blame. If memory serves (iffy proposition with me...) it ended up being the evaporator temperature sensor. After a few minute of cooling the sensor went non-linear, causing the CC head unit to think the evaporator was freezing up. The a/c would not come on until the car had sat for a good long time - then it would do the exact same thing.

My problem is: start the car and it runs the same amount of time each start. I can drive 200 miles and it will not come back on till the car is restarted and then it will run the same few minutes and stop. Very repeatable pattern. That's why I was wondering about the speed sensor. Why is it the same time length each start to stop cycle. It also restarts every time. Everything is like clockwork.

I expect that the time to trip is related to the time to build high side pressure.

Hmm, ok, thanx!

Les,

I went back and reread Jim's last post where he mentioned the sensor. It's not too difficult to pull the panel on the CCU, remove it and then check the temp sensors with an ohmmeter. If you have Alldata, I would expect that you could get a pinout and resistance values from there without too much trouble.

Several years back, I made a test box for the X/11 diagnostic connector near the battery. It is SUPPOSED to trap any errors with the temp sensors and blink a code. The test box consists of a switch and an LED. Somewhere deep in the anals of this site is a schematic for building it. You don't really need to build it if you have a jumper wire and an LED with some leads you can accomplish the same thing. As I recall the poster that posted the schematic was named sort of a common name like Jim Smith or something.

Do a search in the Tech Help Forum for maybe test box, or X/11.

All that said, I am suispecting a bad temp sensor in my 300E right now in spite of the fact that the diagnostics show no codes set. I'm looking for a little bit of time to check the sensors with an ohmmeter.

Good luck,

FYI, here is the thread I referred to: http://www.peachparts.com/shopforum/showthread.php?p=1531158

Unfortunately, there is no follow up with the final resolution. However, there is lots of good discussion on diagnostics and possible root causes.

Thanx for the link! I'll pour over that one. I couldn't find the one about the LED and jumper wire to get trouble codes. For the long haul with the car, that would be interesting I'm sure. I'm still looking for it.

Search for a post entitled "Getting Troublecodes on 300E." In the second post of that thread there is a link labeled "this page" This shows how to build the test set and may have more information.

I didn't take enough time to see if all the codes are in the write up. If not and you do get a trouble code, post it here. I have the CD and will try to find time and remember to look up the code for you.

Try this link http://home.comcast.net/~chwilka/Error_Code%20_reader_reset_tool.htm

I went back and read that article a little. It focuses on the CIS trouble codes, but use the same procedure, except CCU codes are pin 7.

Also, that shows using only a diode and doesn't show the cathode and anode. When I did it, I made a box that had the switch and the diode and it also has a lead that picks up B+ from the battery. I expect you should be able to make the one in the article work once you determine the polarity for the LED.

Hope this helps,
Larry