No heat inside the car (great mistery, checked already ECTS and thermostat)

RGL

Free breather
Joined
May 14, 2013
Messages
114
Reaction score
16
Points
18
Location
Way up north
Hi! My A3 is constantly providing me with new puzzles to solve, but this one is going over my head.

So as soon as outside temperature dropped below 0*C, then I lost basically all heating inside the car. If the climate control is set on AUTO, then there is no air blowing at all. If I manually set it on HI and maximum volume, then I only get lukewarm air. My first idea was to check the coolant temp and this is where it gets weird.

11qr88o.jpg

The red light on the dash is unrelated (airbag, bad switch on passenger airbag).

As you can see from the picture above. The dash temp gauge is at 90*C, but when I checked the numeric value through the climate unit on channel 51 (which should show ECT, right?), then it only showed 30*C maximum. Mind you, the picture was taken after 60 minute drive, so it cannot be 27*C surely.

My first idea was, that it must surely be a faulty ECT sensor, which is feeding wrong numbers to the climate control, making the climate control think, that the coolant temp is only 30*C and thus restricting warm air into the car. Today I changed the ECT sensor (blue to green) and still the number on the climate control unit didn't go past 35*C and still no warm air inside the car (even though dash showed 90*C). Service has already checked the thermostat and it worked and the whole head warmed up evenly (at least so they've said). I've checked the pipes from the radiator and they are warm, so I more-or-less believe the dash gauge showing 90*C. The temperature must be definitely more than what the climate control unit is showing. So now I am running out of ideas. One possibility is that the climate control unit itself is broken, but then again ECT would then be the only thing where the climate control is messing up then, because otherwise it is working prefectly (all buttons do as should and I cannot basically detect any other mishaps from the climate control).

Haven't had a chance to hook up VCDS-Lite yet, but can do it soon-ish. Even though I doubt, that I would get much wiser.

Any ideas oh wise men? :worried:

Car: 1997 A3 1.8 (non-turbo; AGN)

More weird symptoms I've noticed:
- Sometimes randomly after a very long run (60 minutes of highway), heating will actually start to work normally, but not always.
- If the car has been parked in a warm place (heated garage with +10*C) before, then the car often heats up normally, but again not always. And this is even when I exit the garage as soon as I start the car, without giving it time to sit idling in the warm garage.
 
Id be checking the coolant level, and it thats ok then changing the thermostat.

If you don't know the cars history, then a coolant flush through would be a good idea (before changing the thermostat)

The symptoms are very indicative of a failed thermostat (which always fail in the open position)

:racer:
 
Have you check the block for ECT via engine module?
If it's still 30 when vehicle warmed up, your going to need to confirm wiring between sensor and ECM.
Is your new ECT sensor OEM?
 
Id be checking the coolant level, and it thats ok then changing the thermostat.

If you don't know the cars history, then a coolant flush through would be a good idea (before changing the thermostat)

The symptoms are very indicative of a failed thermostat (which always fail in the open position)

:racer:

Thanks for the reply! Coolant level is normal and the coolant itself was changed recently.

I initially also suspected thermostat and got it checked. Also the ECT reading from climate control shows only 20-30*C and even for a thermostat stuck fully open that is way too low IMO (should be at least 50-60).

Have you check the block for ECT via engine module?
If it's still 30 when vehicle warmed up, your going to need to confirm wiring between sensor and ECM.
Is your new ECT sensor OEM?

Thanks! I will check the engine module as soon as I can get my hands on the cable (soon).

The new ECTS isn't OEM, but also not exactly a total cheapo, it's Febi Bilstein. Of course I cannot fully rule out, that the new sensor isn't faulty also, but the odds of it showing exactly the same issues and temperature readings (and no other problems) as the old sensor, are quite low I think.
 
The new ECTS isn't OEM, but also not exactly a total cheapo, it's Febi Bilstein. Of course I cannot fully rule out, that the new sensor isn't faulty also, but the odds of it showing exactly the same issues and temperature readings (and no other problems) as the old sensor, are quite low I think.

I agree completely, but 'twould be folly to assume ;)
 
Connected to VCDS Lite and it confirmed as I suspected that the dash gauge shows the correct coolant temperature and the climate control unit shows wrong. Engine module measuring block values matched the dash gauge showing 90*C, while climate control showed 25*C.

Couldn't connect to HVAC module (got "too many communications" error). Later I read online that it can be due to the cheapo diagnostic cable and changing settings (latency) might help. Will try again tomorrow.

At this point I'm super frustrated since the temperature outside is -15*C blizzard and driving a cold car in this weather is awful.

But the hunt continues...
 
Are you sure the climate unit is not displaying the actual cabin internal temperature it's detecting? If the HVAC unit thinks it's 25 or 27 inside, then it won't keep on pumping in hot air (as 25/27 is pretty hot!) You could try another HVAC unit or also try cleaning in the small grille which is used for the unit to detect the temperature? Let us know how you get on if you don't catch too cold!
 
Are you sure the climate unit is not displaying the actual cabin internal temperature it's detecting? If the HVAC unit thinks it's 25 or 27 inside, then it won't keep on pumping in hot air (as 25/27 is pretty hot!) You could try another HVAC unit or also try cleaning in the small grille which is used for the unit to detect the temperature? Let us know how you get on if you don't catch too cold!

Thanks!

No 51c is for ECT. Checked today that the cabin temperature is channel 2 and 3 and those both showed around 15*C at the end of a drive.

I might go for another unit if I can find one cheap from the breakers. Don't want to spend too much on wild guesses and my current unit looks mint, no pixel loss and no button wear (doubt that I'll get a nice looking unit for cheap).

Before doing that I'll try to see later today whether I can connect to the HVAC module in VCDS and I get any more clues there. Then, once the weather improves a little, I will probably check whether the climate control wires are tightly plugged in behind the dash. I remember them being quite short, when I tore the dash apart a few years back. If that doesn't help then I'll shop around for a used unit.

Unless of course I get new ideas.
 
I would have thought that this indicates an issue with the heater matrix
" If I manually set it on HI and maximum volume, then I only get lukewarm air"

what is the temperature of the feed and return pipes to the matrix? might be a blockage of some sort in the matrix?

if the return pipe is cool , indicates not enough flow through the matrix

unless the heat control flap inside the air handling unit is stuck in the cold position

can you exercise the control flaps from VCDS?
 
Thanks!

I have to have a look at the heater matrix pipes, off the top of my head I'm not sure where they are exactly located under the bonnet, but logically thinking they shouldn't be too difficult to find. I will check the pipes, but changing heater matrix being a total PITA job, then I would like to rule out other things first and leave that as last.

I'm not sure whether I can test the flaps on VCDS-Lite. I will try to check this tonight.
 
vcds lite will not exercise the flaps, but I think gives feedback on position

the matrix can be reverse flushed in situ, there are some vids on you tube

are you sure about the channel on the climate control unit, I found a few different listing for the channel definitions,

is it always on 30, or does it change as the engine warms up from cold?
 
vcds lite will not exercise the flaps, but I think gives feedback on position

the matrix can be reverse flushed in situ, there are some vids on you tube

are you sure about the channel on the climate control unit, I found a few different listing for the channel definitions,

is it always on 30, or does it change as the engine warms up from cold?

Pipes to/from heater matrix are hot.

Managed to get into the HVAC module this time. Yes, cannot do any output tests with VCDS Lite. What is however interesting is, that HVAC sees coolant temp as 0*C (see screenshot 7 attached). I guess this means, that it isn't receiving any information at all and those 30*C (which I see under channel 51) on the climate panel are some made up numbers by the HVAC unit itself?

Also something weird happened after driving around 30 minutes and idling (motor running) for another 15-30 minutes while looking around in VCDS. The heat suddenly came on. The climate control screen, which previously reported coolant temp as 30*C started rising and went to bang on 90*C (even though VCDS showed the real coolant temp to be 99*C in the engine module). Meanwhile I also checked the HVAC module and it was still seeing coolant temp as 0*C (as in screenshot 7). Weird eh?
So while the heat was on, I took a drive for 5 minutes, then parked the car, shut engine off and turned it straight on again and no heat again. Climate control unit showed again coolant temp as 30*C and in the next 5 minutes started gradually climbing to 90*C and turning heat back on. Meanwhile HVAC module in VCDS reported coolant temp. still as 0*C.

This is getting weirder by the minute. The climate unit has to be getting a bad signal or is faulty itself. I checked the ECTS wiring under the bonnet and couldn't see much, because most of it is in the protective plastic sleeve. However as far as I could see there was no damage. Next up would be to check the wires behind the dashboard I guess? It could be also the case of pure unluck, in which case I replaced a faulty ECT sensor with a faulty ECT sensor, but this is rather unlikely as the problems are identical with no new symptoms.

PS! Screenshot 6 should be showing the flap motor position.
 

Attachments

  • Screenshot_7.jpg
    Screenshot_7.jpg
    35.9 KB · Views: 152
  • Screenshot_6.jpg
    Screenshot_6.jpg
    38.7 KB · Views: 111
If still wanting to test ECT:
Can you send a ref voltage up the relevant temp sensor circuit and read out block with VCDS. This will confirm is sensor/wiring is defective.
 
If still wanting to test ECT:
Can you send a ref voltage up the relevant temp sensor circuit and read out block with VCDS. This will confirm is sensor/wiring is defective.

Hmm...I've never done anything like that before.

I presume I need a multimeter and then I unplug the ECTS and hook up the multimeter instead and monitor any changes through VCDS. I think I cannot read the voltage directly from the ECT sensor in VCDS HVAC module? Then again the current should change the value of coolant temperature in HVAC module if the wires are fine, right? Am I on the right track?
 
interesting stuff

my reading of the wiring diagrams is that the HVAC unit gets its temperature reading not directly from the ECTS but from somewhere else, possibly the instrument panel. This because it does not show a temperature sensor input to the HVAC, but does show some links to the instrument panel.
If this is the case then the fault could be
1 output from instrument panel
2 wiring between them
3 fault in HVAC


I would start by checking the wiring between them, i.e. make sure plugs are seated correctly, no damaged wires


the display on my HVAC also was limited, it did not show the full range of temperatures, it max out at 75c, leading me to change a stat, when there was nothing wrong with it

I might be wrong on this, anybody got access to the plug wiring for the HVAC unit?
 
Hmm...I've never done anything like that before.

I presume I need a multimeter and then I unplug the ECTS and hook up the multimeter instead and monitor any changes through VCDS.

No you would need a something like a powerprobe capable of inserting a ref voltage, as a dummy sensor, within the plausibility range of the original sensor. You don't want to fire 12v up the wrong line. When you see the smoke, you won't be able to get it back in :p

If you want to send me your vehicle details including VIN no, I'll pull the circuit diagram and can advise on system setup and on further checks.
Before that though, having re-read you OP, the first paragraph describes what is commonly a temp blend flap issue. While the other observations made to this point are causing some confusion, they may not be in the right direction. A simple check can be a visual inspection of the blend flap motor (varying levels of stripping required) or, even better, running a test program on the HVAC motors.