Used a different meter.... Something fishy here Max cold 283.9k ohms 18C 90.7k ohms 19C 40.2k ohms 20C 23.36kk ohms 21C 14.92k ohms 22C 9.95k ohms 23C 6.64k ohms 24C 4.27k ohms 25C 2.490k ohms 26C 1.094k ohms Max heat 1.2 ohms
Thank you, those match my own measurements exactly, within the tolerance of the resistors! If you now measure the voltage at Pin 5 with the control panel connected, you’re in for a surprise! I’ll explain tomorrow!!
Another owner reported these values: {QUOTE} With panel powered up the voltage on pin 5 (Yellow) varies as the temperature control is rotated :- Min 4.95V 64 4.62V 68 3.61V 72 2.58V 76 1.57V 80 0.63V Max 24mV
Thank you. So far, we've discovered the temperature control is nothing more than a switch which provides an increasing resistance to ground as the temperature is reduced. At max temperature, the resistance to ground is 0. How does the ECU use this varying resistance? Also, why use resistors with non-standard values, weird ones like 2370 ohms? Quick lesson (apologies to those who know), on the left of the following diagram is the formula for the output voltage of a potential divider where there are two resistors connected across a supply Vin and an output is taken at the junction of the two resistors. Which is what we have here... The rotary switch and its resistors take the place of R2 and by looking at the circuitry in the ECU, R1 is constant at 10000 ohms and Vin the output of a voltage regulator which supplies a constant 5v from the car battery supply which of course can vary. Image Unavailable, Please Login Whoops, that should read Vin * R2/(R1+R2) Plug in the marked values of the 10 resistors being used and we get the following results. The actual column is the voltages measured in my car, very close to the design values. Image Unavailable, Please Login So, the design objective is revealed: to provide a voltage which increases linearly by 0.5v steps with decreasing temperature. There will be a converter chip on board to convert the analog voltage in the range 0-5v into a number in the range 0-10. The voltages for the direction and fan controls are similar in total range with larger steps because there are only 6 positions, not 11. Spacing out the voltage steps gives improved noise immunity - always a problem in an automotive environment and the results are immune to varying battery voltage, at least down to a level where you don't care because the car won't start. The same converter will be used for other analog sensors in the system, most likely temperature sensors, all to give the software running in the ECU the information it needs.. This wraps up this investigation of the HVAC control panel. As far as I can tell, it's working perfectly, so the fault of no heat in my car lies in the ECU or the motorised valve. The dealer wanted to replace the panel for $3000, just as well I told them not to! Thanks for reading, hope you've found it interesting!
Or the pump. These fail a lot on the V12s, not so much on the 355’s. I recall that the WSM says this is used when the engine is at idle, but I’m not aware of any engine rpm input into the HVAC ECU. I’d say it runs all the time. Do you have hot water at the valve inlet? I believe the valve is opened by a stepper motor, so it may not be so easy to simulate the voltage for test purposes. Will you be changing the chips in the ECU yourself and do you know which is which? I have a list somewhere.
Yes, I need to check the pump and understand the interface from the ECU to the rest of the car. When the motorised valve failed, I took it apart and found the valve itself had seized and the motor had destroyed one of the plastic gears in the geartrain drive it. It's a DC motor with a positional feedback potentionmeter so the ECU can keep track of where the valve it set. There are 6 driver chips which are basically high-powered op-amps to drive the motorised valve, the recirculation flap motor, the vent direction motors and I assume the pump as well,. I bought 10 of the chips on eBay - TCA2465G - which are meant to be short-circuit proof but they do seem to fail. They are 20 pin surface mount packages which is no problem providing the board underneath is in good condition. The chips came from China, marked Siemens and I have no reason to think they are not ok but I will test them before installing them on the board. Given the cost of the ECU - even if you can get one - there's a need to minimise work on it. It's not helped by the board being sprayed with conformal coating to keep the damp out.
I had to replace my valve/motor assembly. Ricambi has a retrofit kit of an aftermarket valve with a new bracket. Works great. I also had to replace the ecu driver.
