How do you decide which 4 pins of the "ice-cube" relay get installed into which of the 12 pins on the KLIMA relay?

You run power to the coil.

You run power to one of the load pins.

You run the signal from the PBU off pin #10 to the opposite side of the coil. This provides the ground signal for the coil.

You run the the output to the compressor, pin #7 to the the second load pin.

It's all been done before and it works fine for the purposes intended.

The article is a good reference, however, I'm familiar with the system to the level he has written in the article.

A faulty CCU is also a possibility. As are poor connections just about anywhere in the circuit. I would think it possible for the evap. temperature sensor to drift out of range, and only permit compressor operation when it (the evap sensor) is at an extremely high temperature. Once cooled off enough to shut the compressor off, it could take an extra long time to heat up again. Just speculation at this point.

I've got a similar problem, but haven't really looked into it yet. In my case, the CCU would not ground the compressor request line at all, ever. I 'solved' it by installing a toggle switch to ground the compressor request line. When I get the time, I'll explore further.

__________________
1989 300 SEL that mostly works, but needs TLC

Got it - I thought you had tried to plug an ice cube relay into the physical space occupied by the KLIMA relay. When you say "I replaced the KLIMA relay with an ice cube relay" I suspect I was not the only one that had assumed this.....

If you are losing the ground signal to trip the relay it should not be too hard to figure out if you are losing the signal via the CCU or one of the switches that will interput the ground signal.

Sorry for the confusion. Yes, I made up a set of jumper wires to fit the plug sockets and also fit the pins on the ice cube relay. The ice cube relay is hanging out about three inches above the sockets, with the wires pointing downward and engaging the pins.

I'm losing the signal via the PBU. The only switch that will interrupt the ground signal is the low pressure switch and I jumped it. Same results.

Arthur thinks the evap temp sensor might be out of spec. If it is, the PBU will drop the compressor and go into default. Won't come back without a shutdown.

I need to find the specs for resistance of the evap temp sensor to enable some determination if it is the culprit.

I have an old PBU to try. The old one would always run the compressor, even in economy. At least we can rule out the PBU if the same problem recurs.

Quick and dirty test:
1) Measure the resistance of the evap sensor before starting the A/C.
2) Buy a resistor of the same value, and plug it into the harness in place of the evap sensor.
3) Run the A/C. If it continues to run past your 4 minute failure limit, you'll know it was the evap sensor shutting things down. If not, then it probably has nothing to do with the evap sensor.

__________________
1989 300 SEL that mostly works, but needs TLC

If the problem is the fact that the evap sensor is out of limits, how would you know that the value of the resistance with the a/c off would be within specs??

It might work, but, if it doesn't, you haven't confirmed anything, correct?

Have you verified that you are losing the ground signal? It could be that you have a weak AC compressor clutch that is giving up after a couple of minutes of operation.

I have verified this. The AC compressor clutch works fine and will remain engaged indefinitely with an external ground signal to the ice cube relay coil.

Furthermore, during the same test, I watched the ground signal on pin #10. Remained for four minutes..............................

I'm counting on the fact that the A/C works when first started. Therefore, the evap sensor resistance at start-up allows compressor engagement.

Therefore, a resistor of the same value would simulate evap. conditions at startup. If the compressor still shuts off after 4 minutes, the evap sensor is not the problem. If, however, the compressor continues to run, it's likely that the evap sensor is at fault.

__________________
1989 300 SEL that mostly works, but needs TLC

If it continues to run, I agree with you.

If it shuts off, it may be the resistor value is not to spec. I think it's possible that there is a delay in the PBU before it drops the compressor out. The fact that it is not instantaneous does not necessarily mean theat it is within spec for the first four minutes. But, I'm just grasping here.

What leads me to believe this is the four minute time frame is the same, no matter if the evaporator is at 80 degrees or if it was just run and it is at 50 degrees. The time does not seem to be related to evap temp. You would think that the colder evap would cause a compressor shutdown earlier on the second and subsequent runs. While this is possible (because I have not timed it carefully enough), it doesn't appear to be the case.

It's not likely. The entire purpose of the evap sensor is to allow the system to maintain an evap core temperature above freezing. Were it allowed to get colder, the condensate would freeze and block the airflow through the condensor. A delay would be of no value that I can see.

I see. So cycling the key allows another (+/-)4 minutes? That would seem to indicate that the shut-down has nothing to do with temperature. Not the Evap temperature, and not the temperature of the components in the PBU. The PBU also has inputs from an outside air temperature sensor, an in-car temperature sensor, the heater core temperature sensor, and the coolant temperature gauge sensor. All of these, except the coolant temp sensor, share ground point x48p. Out of that list, I'd think it likely that the outside air temperature and the coolant temperature sensors might cause the compressor to shut off. However, again I see no value to a delay circuit.

Before I'd condemn the PBU, I'd clean up the power and ground connections, as well as the sensor ground x48p.

__________________
1989 300 SEL that mostly works, but needs TLC

Arthur has been assisting me off the forum.

I've measured a resistance of 2.2K on pins #4 abd #10 on the right side plug to the PBU.

If I have the correct pins (no certainty due to the poor quality of the schematic), then the resistance is below the minimum value of 2.5K where Arthur believes the PBU will drop the compressor and default.

Any additional thoughts??

That's a (calculator) 12% difference. Is that a resistance @ certain temperature? Easy way to check would be to plug in a 2.5K (or greater, since that's the minimum value) resistor and see what happens.

__________________
1989 300 SEL that mostly works, but needs TLC

I decided to try my old PBU (engages compressor in economy after 10 minutes or so). Lo and behold, the problem is eliminated. The compressor runs for eight minutes, steady, through a couple of high pressure aux. fan cycles.

I'm going to leave it with the old PBU for awhile and see how it does.

Thanks for all your assistance.