View Single Post
Wayne 962 Wayne 962 is offline
Author of "101 Projects"
Wayne 962's Avatar
Troubleshooting the AC / Heater unit in the Project 959... (10-28)

Okay, so before I decide to send off the AC control switch to be rebuilt, I wanted to make sure 100% that the unit was faulty. So, I decided to perform some extensive testing - the proper way. By meticulously going through each component, I can be assured that each one (and the wiring) is working and is not contributing to the problem (which is that the heater is on full heat all the time, and the compressor doesn't come on).

Good electrical diagnosis is not rocket science, but if you don't progress slowly, and methodically, and record all of your results, then you may have not done anything at all. A lot of bad mechanics are in a hurry, and just perform guesswork regarding the cars - they are more apt to replace parts (at your expense) rather than actually do the nitty-gritty work that is required to accurately troubleshoot an electrical problem. I hope that by documenting this process here, that some people can learn what is involved with the steps of troubleshooting electrical gremlins.

So, I started with the three temperature sensors in the car. These three give feedback to the climate control system and help the computer to decide how to meter the air. I took off the plugs on the back of the unit, looked up the sensor pins in the electrical diagrams and then tested each one.


Test Temperature Sensors:

Pins A1 & B6 = Blow-out or vent temp sensor. Read 11.92 Kohms which is within spec for room temp.

Pins A2 & B6 = Internal passenger compartment temp sensor. Read 11.60 kohms which is within spec for room temp.

Pins A9 & B6 = External temp senors. Read 10.98 kohms, which is about right.

All three temp sensors have similar ranges. I heated up the car the other day and re-read them again, and they changed resistance according to how the spec is listed in the factory manuals.

Conclusion: all three temperature sensors and their respective wiring are fully operational.


Test Heater Timing Valve:

The factory manual states that you should pull out harness R from the central electric system, and test pin R12 for power. Actually, it says test pin R12 for power, and then it says pull the harness out. I made this mistake - there's no power to the harness when it's unplugged. It took me 3 days to realize that the manual meant to stick the probe into the connector while it's still connected. Success - there's power running to that point.

Pins R12 & R13 = coil for the heater valve. Measurement of 14 ohms, which is correct, and confirms that the wiring to the heater timing valve coil is correct.

Next, I read the diagrams, and saw that the valve is actuated by bringing pin A11 to ground. First, I monitored the current and resistance through this pin to see if the control unit ever brought it down to ground. Never happened. When the pin is grounded, then it powers the valve, and stops the flow of hot water through the heater. When the A/C system is on, or the thermostat is set to maximum cold, then this pin should be at ground continuously. Also, when the heater knob is set in the middle range, the heater valve should cycle on and off every 3.5 seconds or so. This never happened either. When the pin was manually grounded, you could hear the timing valve in the engine compartment click on and off (the manual said to listen for this happening). So, the operation of the valve seemed to be working (although I have no idea if it's clogged), but the control switch is never sending the signal to shut off the flow of hot water.


Testing Flaps

I read somewhere that stuck flaps may also cause the heater to fail to the full heat position. The 959 has two flaps, one in the footwell, and one up for the defroster. Taking a 9V battery, I hotwired the operation of the motors while monitoring the resistance in the sensors that measure flap position. Switching polarity on the motor opens and closes the flap:

Footwell motor:
A5 and B8 show 234 ohms to 2.13 Kohms
B3 and B8 show 2.15 Kohms to 168 ohms

Defroster motor:
A6 to B7 show 2.34 Kohms to 500 ohms
B5 to B7 show 333 ohms to 2.175 kohms

Both seem to be operating roughly within spec.

I also connected the control unit and manually operated the footwell and defroster motor sliders while watching the actual flap mechanism (you can see the flaps from the trunk and from inside the dash). Both worked normally. Probably not a problem with the flaps.


Recirc flap & relay

Next I checked the recirculation relay and flap motor (this one has no sensors on it). Pressing the button on and off, you can hear the relay off on the right (under the glove box) and you can also hear the flap opening and closing properly.


A/C Switch & Relay

Likewise, clicking on the A/C compressor switch successfully operated the A/C relay and set the blower operation to low. I went back today to recheck the actual pin outs on the connector, and sure enough, pin A8 has +12V applied when the A/C switch is turned on. This seems to happen whether the thermostat is full on or full off, further indicating a problem with the control unit. Odd though, in my tests, I was not able to get the compressor clutch to operate. If pin A8 has +12V, then according to the electrical diagrams, the compressor clutch *should* operate, unless there is a problem with the hi-low pressure switches or the anti-icing cutoff switch. This currently has me slightly puzzled, but I don't have enough time right now to do more troubleshooting.


Defroster Switch

The defroster switch activates the relay on the control unit, sets maximum heating, turns on the compressor, and sets the flaps to go to the windshield. This seems to work properly, as the compressor had +12V, the flaps closed, and the blower went to max.


On Jeff's unit, I took it apart to see where the contact points are for the temperature control. As you might know from having an old stereo, sometimes the potentiometers of these old units fail, which results in a crackling noise when you adjust the volume. I thought the pot on this unit might have failed, so I ran tests on Jeff's unit and compared it to the 959 unit:

A B C = pins on pot

944 Unit:
Full CCW A-B 1.37 Kohms
Full CCW B-C 0.6 ohms
Full CCW A-C 1.37 Kohms
Full CW A-B 0.6 ohms
Full CW B-C 1.3 Kohms
Full CW A-C 1.3 Kohms
1/2 way A-B 2.74 Kohms
1/2 way B-C 2.7 Kohms
1/2 way A-C 1.37 Kohms

959 Unit:
Full CCW A-B 1.36 Kohms
Full CCW B-C 0.6 ohms
Full CCW A-C 1.36 Kohms
Full CW A-B 0.6 ohms
Full CW B-C 1.3 Kohms
Full CW A-C 1.3 Kohms
1/2 way A-B 2.8 Kohms
1/2 way B-C 2.4 Kohms
1/2 way A-C 1.35 Kohms

As you could see, the pot on the 959 unit tested almost identical to the 944 unit. Rats, when you have a moving part like this, it's somewhat common to find a failure.



With the exception of the A/C clutch not operating properly, the proper operation of the valves, flaps, and sensors of the system were confirmed. In addition, the signal to turn off the heater valve was not working properly when monitored as per the factory manuals. With this information, it becomes apparent that the A/C control unit is faulty with respect to the thermostat portion of the unit.

What to do now? There is a place in Italy that seems to have a bunch of experience in repairing these units (, and here is a site that recommended them after they performed a repair on the similar-looking 964 unit: I am going to email them tonight with a link to this thread, and see what they have to say. I think that based upon what I've read, the unit should be repairable. Unfortunately, comparing it to the 944 unit, there is one board that is significantly different. Also, the electrical diagrams for the 944 show the operation of a mixed-air flap (not exactly sure what that is, and why the 959 doesn't have one) which could cause problems.

In a pinch, I'm quite confident that I could adapt the 944 unit to the 959. What I would need to do is to obtain a servo motor with a pot that will emulate the mixed-air assembly on the 944 and give the 944 control unit the proper values it's looking for (maybe even use the motor and sensor off of a 944). Then, I would have to swap the pins from A4 (I believe) to A11, which is the heater valve control signal for the 959. In a pinch, I'm quite confident that I would be able to get that system to work. However, that's a lot of effort, and if the original unit can be repaired, well, that's really the proper way to go then.

The good news is that through all of this testing, I learned that grounding pin A11 will energize the water valve and stop the flow of hot water into the heater. So, I can hot-wire a lead from pin A11 to ground to permanently shut off the heat while I'm waiting for the unit to be repaired. The only hang up and concern that I have is that some solenoid valves have what are known as duty-cycles. That means that they are not supposed to be activated 100% of the time. I'm not sure if this valve has a duty cycle. I doubt it, but I would hate to burn out the valve accidentally ($600 if it's still available). I'm thinking that the valve does not have a duty cycle - the text of the factory manuals seems to indicate that the valve should be energized (closed) when the A/C system is on, which would be all the time on a long drive. If anyone has any input on this, I would love to hear it. I'm sure the 944 valves operate in the same manner.

Wayne R. Dempsey, Founder, Pelican Parts Inc., and Author of:
101 Projects for Your BMW 3-Series 101 Projects for Your Porsche 911 How to Rebuild & Modify Porsche 911 Engines 101 Projects for Your Porsche Boxster & Cayman 101 Projects for Your Porsche 996 / 997 SPEED READ: Porsche 911 (October 2018)

Check out our new site: Dempsey Motorsports
Old 10-28-2007, 08:20 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #1 (permalink)