Need Help With Climate Control

[crackership]

Hi All,

I've been chasing a "full heat all the time" problem for a few days now since getting my car running (badly) again (and trying to go on tuning drives in 80F weather with the nuclear reactor that is the 944 heater on full blast).

At first, I figured it was that little clip thingy, but I stuffed my face under the dash to discover I'd already replaced it with Tom's design some time in the past - I only vaguely remember this, but I guess I did it.. I must be getting old

Turns out the motor that moves that flap just doesn't respond to turning the hot/cold knob. The fan knob works, the defrost and AC buttons work - I'm not sure if the sliders work (not really sure how to tell) but the hot/cold knob doesn't move that motor.

I pulled the climate control module out and, again, thanks to Tom's amazing writeup on the subject, I was able to confirm that the motor does work by applying power and ground to pins 10 and 11 on the center plug - OK, great! The motor works! So why doesn't it work when I twist the knob?

Thanks in advance!!

-Dave

[Tom]

It's 'probably' the CC unit itself, but before reaching that conclusion, have you tested all three temp sensors per the charts I posted? If the cc thinks it's freezing inside, that 'might' explain your issue. Also possible the pot behind the knob is bad, but probably unlikely. Sounds like you 'get' the charts, but if helpful I also did a companion video here. If the temp sensors are good, my guess would be some of the board connectors inside the CC are corroded apart -- pretty common from what I can tell. If you want me to see if I can fix it, I'm game. I collected all or most of the IC's inside the unit when I was obsessed with it for a while -- though chances are it's a simpler corrosion issue. But definitely check the temp sensors first, and probably worth confirming the solenoids work and that the heater valve holds vacuum....

[crackership]

Thanks, Tom!

I just went through your procedure and things seem ok ish, though I'd say some are a little odd. Here's what I get:

A12 -> A7 = 2.4k ohm
A12 -> B6 = 420 ohm
A1 -> Ground = 9.1k ohm (It's about 80F in my garage)
A2 -> Ground = 9.3k ohm
A9 -> Ground = 17k ohm (this one seems a little wonky)
A3 -> Ground = 43 ohm (outside the range of your pdf, but about what you got in the video)
A4 -> Ground = 41 ohm (again, outside the range of your pdf, but about what you got in the video).

With all that said, I think I've discovered a problem (very possibly the problem). When pulling out the CC module, I noticed the familiar smell of burned electronics.. I pulled the CC unit apart and discovered that this little yellow capacitor appears to be having a pretty bad day:

Screen Shot 2022-08-25 at 10.03.46 PM.png (1.3 MiB) Viewed 11961 times

Screen Shot 2022-08-25 at 10.05.30 PM.png (961.66 KiB) Viewed 11961 times

Screen Shot 2022-08-25 at 10.06.17 PM.png (1.29 MiB) Viewed 11961 times

Do you think this is fixable or is it dead?

[Tom]

That should buff right out.

I'll get to the blown capacitor in the next post, but in the meantime...

What was the temperature outside when you tested those temp sensors? 17k equates to low 50's F. Could be a wiring/connection issue, but if that one is bad, you can replace just the little thermistor that is soldered to the end of the sensor harness wires. NTC thermistors are spec'd by their resistance at 25C and a "beta" value -- which approximates their resistance curve as the temp changes. The three temp sensors in the 944 HVAC system are rated at 10k @ 25c with a beta of about 3900k over the relevant temp range, as best I can approximate. Digikey 570-1200-ND (mfg. ACC004) is a good replacement:

https://www.digikey.com/en/products/det ... 04/5967514

hvac temp sensor.gif (301.31 KiB) Viewed 11950 times

[crackership]

What was the temperature outside when you tested those temp sensors? 17k equates to low 50's F. Could be a wiring/connection issue, but if that one is bad, you can replace just the little thermistor that is soldered to the end of the sensor harness wires. NTC thermistors are spec'd by their resistance at 25C and a "beta" value -- which approximates their resistance curve as the temp changes. The three temp sensors in the 944 HVAC system are rated at 10k @ 25c with a beta of about 3900k over the relevant temp range, as best I can approximate. Digikey 570-1200-ND (mfg. ACC004) is a good replacement:

Thanks, Tom! The car was in my garage (~80F) with all the doors and hood open at the time I tested all three sensors, so I would've expected them all to read about the same. I'll grab a (some?) thermistors and swap that one out. Do you think it's wise to swap all three since (I believe) the CC works by sensing a differential between them and this would mean all three need to operate pretty much identical? Also, where are these little guys in the car?

[Tom]

What was the temperature outside when you tested those temp sensors? 17k equates to low 50's F. Could be a wiring/connection issue, but if that one is bad, you can replace just the little thermistor that is soldered to the end of the sensor harness wires. NTC thermistors are spec'd by their resistance at 25C and a "beta" value -- which approximates their resistance curve as the temp changes. The three temp sensors in the 944 HVAC system are rated at 10k @ 25c with a beta of about 3900k over the relevant temp range, as best I can approximate. Digikey 570-1200-ND (mfg. ACC004) is a good replacement:

Thanks, Tom! The car was in my garage (~80F) with all the doors and hood open at the time I tested all three sensors, so I would've expected them all to read about the same. I'll grab a (some?) thermistors and swap that one out. Do you think it's wise to swap all three since (I believe) the CC works by sensing a differential between them and this would mean all three need to operate pretty much identical? Also, where are these little guys in the car?

If you get the thermistor I mentioned above, it should match the others (so I'd leave them alone). The outside temp sensor is under the blower motor cover (black plastic at the base of the windshield) and is attached to the side of the blower motor intake screen. It has a different housing than the one I posted, but the thermistor inside has the same specs. It's built into a little connector that plugs into a JPT connector, so probably worth pulling the sensor and cleaning the connectors with sandpaper, etc. and testing at the sensor itself, just in case the extra resistance is coming from the connector/wiring. It never moves or gets bounced around, but the connector style is the same as the injectors and speed/ref sensors, so prone to failure under the connector on the harness side. I'd probably ohm it out from under the hood to A9 too, in case there's extra resistance in the harness (it goes through another connector within the harness). Last I checked that sensor is NLA, but if you have a hard time getting the thermistor out, let me know and I may be able to help.

[Tom]

Edit in case I buried the lead with a TL;DR post.... The CC unit is probably not working right and why you are stuck on heat. Most people would toss it and find a new one. My comments below are for people interested in trying to diagnose and repair a bad climate control unit....


Now, re the blown capacitor. I believe that is C1 on the circuit board, a 25V 47uF electrolytic capacitor. It would be easy enough to swap out that capacitor, but something most likely caused it to burn up -- meaning a new one would likely suffer the same fate if you don't sort out the cause. Check your fuses too to see if whatever blew that capacitor also took out a fuse.

The CC unit has two unique voltage rails -- one is ~6.5 volts and the other is closer to 11 (battery voltage minus 2). Both are created with zener diodes and transistors, rather than a dedicated chip. The 6.5v rail helps create references voltage for the chip (4201A) that control the heater motor. The 11 volt rail is the supply voltage for all the 4201A chips that control the servo motors. The capacitor that blew on your unit hangs off the 11 volt rail (to ground), so there's a good chance none of your motors work via the CC unit -- have you checked?

I would check to see if that voltage rail is dead by probing the pcb pad shown below. If you still have voltage there, then it's possible something downstream of the voltage regulator circuit blew the capacitor, such as a blow 4201A chip. If there is no voltage, then the rail is probably dead, and more likely a blown zener diode or transistor (which should be available). If it's a blown 4201A chip, I have them. When mine blew, I discovered they were obsolete and had to buy the remaining stock (like 150+ of them for like $20) from some vintage chip dealer in the UK -- which at the time were the only ones anywhere I could find anywhere. Not sure how electronicsy you are and if this is any help or not? Like I said, I'd be happy to take a shot at fixing it if helpful. You can also get used ones on eBay, but you never know what you're going to get....

hvac 11v rail check.gif (289.64 KiB) Viewed 11933 times

[crackership]

Edit in case I buried the lead with a TL;DR post.... The CC unit is probably not working right and why you are stuck on heat. Most people would toss it and find a new one. My comments below are for people interested in trying to diagnose and repair a bad climate control unit....


Now, re the blown capacitor. I believe that is C1 on the circuit board, a 25V 47uF electrolytic capacitor. It would be easy enough to swap out that capacitor, but something most likely caused it to burn up -- meaning a new one would likely suffer the same fate if you don't sort out the cause. Check your fuses too to see if whatever blew that capacitor also took out a fuse.

The CC unit has two unique voltage rails -- one is ~6.5 volts and the other is closer to 11 (battery voltage minus 2). Both are created with zener diodes and transistors, rather than a dedicated chip. The 6.5v rail helps create references voltage for the chip (4201A) that control the heater motor. The 11 volt rail is the supply voltage for all the 4201A chips that control the servo motors. The capacitor that blew on your unit hangs off the 11 volt rail (to ground), so there's a good chance none of your motors work via the CC unit -- have you checked?

I would check to see if that voltage rail is dead by probing the pcb pad shown below. If you still have voltage there, then it's possible something downstream of the voltage regulator circuit blew the capacitor, such as a blow 4201A chip. If there is no voltage, then the rail is probably dead, and more likely a blown zener diode or transistor (which should be available). If it's a blown 4201A chip, I have them. When mine blew, I discovered they were obsolete and had to buy the remaining stock (like 150+ of them for like $20) from some vintage chip dealer in the UK -- which at the time were the only ones anywhere I could find anywhere. Not sure how electronicsy you are and if this is any help or not? Like I said, I'd be happy to take a shot at fixing it if helpful. You can also get used ones on eBay, but you never know what you're going to get....


hvac 11v rail check.gif

Thanks so much, Tom! So, I actually have a degree in electronics engineering. However, that was what feels like a billion years ago and I have done very little of it since then - I've been in software for my whole career - so I "get it", but I'm super rusty when it comes to application of that ancient knowledge. If I could figure out how to get the three PCBs apart and off of the clear front plate, I think I'd feel a lot more like I can handle this

I have tested the motors by applying voltage to the pins on the harness and they do work, but none of them seem to work via the CC unit controls - is that what you were asking?

I'm definitely leery of eBay units and the few NOS ones I've found are $$$$$$. I am a bit worried that the PCBs on mine just look like they're in terrible shape, but I'll check for voltage on the rail and go from there.

I mean, if you have time and the inclination to take a whack at it, I'm happy to send it to you, I just feel bad asking for the help because somewhere in the dusty corners of my mind is all the knowledge to handle this myself..

[Tom]

Thanks so much, Tom! So, I actually have a degree in electronics engineering. However, that was what feels like a billion years ago and I have done very little of it since then - I've been in software for my whole career - so I "get it", but I'm super rusty when it comes to application of that ancient knowledge. If I could figure out how to get the three PCBs apart and off of the clear front plate, I think I'd feel a lot more like I can handle this

I have tested the motors by applying voltage to the pins on the harness and they do work, but none of them seem to work via the CC unit controls - is that what you were asking?

I'm definitely leery of eBay units and the few NOS ones I've found are $$$$$$. I am a bit worried that the PCBs on mine just look like they're in terrible shape, but I'll check for voltage on the rail and go from there.

I mean, if you have time and the inclination to take a whack at it, I'm happy to send it to you, I just feel bad asking for the help because somewhere in the dusty corners of my mind is all the knowledge to handle this myself..

I suddenly feel conspicuous giving electronics troubleshooting advice to an EE. However, I feel my finance and law degrees have prepared me to give this advice, so I'll charge on.... Totally up to you if you want to dig in. If you'd rather send to me, might be best to remove the clear plastic first, so you can break it yourself and I won't need to feel guilty about it. On my car, mine is cracked several times over, but I use clear packing tape to hold it together and that's been working for nearly 2 decades now....

Separating the boards is a total pain. They are held together with a series of bare electrical leads/pins that carry signals and serve to connect the board together physically. The leads are spindly and have a tendency to corrode and fail. There are a couple dozen leads on each edge, so quite tedious to unsolder, but worth the effort to replace. I've disassembled maybe 6 or 7 of these, and literally everyone had connector leads that were either already reduced to dust/crud, or were 99% eaten away. On two different units I worked on, replacing those pins alone fixed the whole thing. I had some wire "bus bar" handy so used that -- I think it was 24 gauge, but don't quote me on that. Use the biggest solid wire you can get in the holes. Here are the little pins that hold the boards together....

cc-pins.jpg (1.3 MiB) Viewed 11911 times

I posted some of this on RL before the great schism of 2021, but to keep the discussion alive in case you dig in yourself, here is the theory of operation -- as determined by an accountant turned lawyer

This is what I’ve come up with so far in terms of how the motor/flap controls work. I’m no engineer, so any/all comments and corrections from the better-informed are encouraged. The key seems to be the chip they used to control the motors/solenoids, an obsolete TLE4201A chip by Siemens (aka 4201A1, which works fine as well). It’s hard to find the chip, but I was able to find a batch from a place that specializes in obsolete chips. (See datasheet attached.) There are actually four of these chips in the unit, one for each of the three motors, and one for the two solenoids. The description below is based mostly on the defrost and footwell flaps to make the concept easier to follow.

The 4201A puts out close to its supply voltage on Pins 1 and 9 and, depending on its inputs, is able to reverse polarity of the output voltage between those two pins. That’s how it opens and closes the flaps. It can power its output in either polarity, or put out no power at all, depending on its inputs. The circuit uses a voltage divider to set up a narrow range of voltage on pins 3 and 8, roughly 3.34v on pin 3, and 3.16v on pin 8 (depending on the exact level of the ~6.5 voltage rail in the unit). The output gives power in one polarity above 3.34v, power in the opposite polarity below 3.16v, and no power in between those voltages.

The control input for the chip is on pins 2 and 7, which comes from the dash slider pots for the defrost and footwell flap, and from the temp sensor circuit for the heater flap; along with the feedback potentiometer inside each motor. (This input can be over-ridden by defrost buttons and the like, but I’m sticking to how the basic system works for now). If the control voltage on 2 and 7 is above 3.34v, the 4201A sends power out in the polarity needed to open the flap. If the control voltage is below 3.16v, the chip sends power out in the polarity needed to close the flap. In between 3.34 and 3.16v, the chip puts out no power. The sliders in the dash and the motor pots work like a see-saw, always keeping the motor moving until the two pots create a combined/blended voltage between 3.34 and 3.16. If you move the defrost slider all the way to the left, for example, it contributes very little voltage to the control signal, so the control voltage will be below 3.16v and the motor will turn unless/until the motor shuts the flap. When the motor is all the way in the closed position, its pot is contributing its max voltage to the control signal, getting it back above 3.16 where the 4201A turns off the power to the motor. I modeled the signals in iCircuits and think it’s very clever. Move the slider 40% to the right, and the motor needs to turn about 40% before the control voltage gets above 3.16v to shut off the motor. Same results at any point on the slider. This may be engineering 101 to those in the know, but it was new to me and seems clever. In any event, I thought it was cool and might help somebody down the road….

tam schematic cc.gif (172.68 KiB) Viewed 11911 times

4201A datasheet (1).pdf

(147.69 KiB) Downloaded 333 times

[crackership]

Thanks so much, Tom! So, I actually have a degree in electronics engineering. However, that was what feels like a billion years ago and I have done very little of it since then - I've been in software for my whole career - so I "get it", but I'm super rusty when it comes to application of that ancient knowledge. If I could figure out how to get the three PCBs apart and off of the clear front plate, I think I'd feel a lot more like I can handle this

I mean, if you have time and the inclination to take a whack at it, I'm happy to send it to you, I just feel bad asking for the help because somewhere in the dusty corners of my mind is all the knowledge to handle this myself..

I suddenly feel conspicuous giving electronics troubleshooting advice to an EE. However, I feel my finance and law degrees have prepared me to give this advice, so I'll charge on.... Totally up to you if you want to dig in. If you'd rather send to me, might be best to remove the clear plastic first, so you can break it yourself and I won't need to feel guilty about it. On my car, mine is cracked several times over, but I use clear packing tape to hold it together and that's been working for nearly 2 decades now....

Definitely don't feel too conspicuous - I went to DeVry In total fairness, I actually think I got a great education there, but I haven't used it in so long that I honestly remember very little. Plus, they teach you how to design circuits, not how to troubleshoot broken ones . Your practical experience with this CC unit is worth exponentially more than my ancient and dusty degree - and definitely in this context.

RE: Getting the clear plastic thing off - should I just commit to breaking it? It appears to be held in place by the AC and Defrost/Recirc buttons and those look to be soldered in place. From what I can tell, breaking it is about the only option... any tips or tricks from your experience?

RE: Voltage on the 11V rail - Nope. Well, not entirely nope - I'm getting ~0.28V, but I think that qualifies as 'nope' where 11V is expected.

I'm still leaning the direction of sending this to you, if you have the time to have a look at it - I've just spread myself too thin as of late and so this is 'just one more thing', if you know what I mean