A3 2004, cranks but won't start

   #1  

mpiroglu

Verified VCDS User
Verified
Joined
Jul 25, 2017
Messages
7
Reaction score
1
Location
Australia
VCDS Serial number
C?ID=312228
A3 2004 8P1 2.0FSI manual, 192770km-119781mi

What happened
:
It was working perfectly until I changed the oil. I forgot to warm the engine before changing the oil, ending up overfilling. Also during servicing I unplugged the throttle body and cleaned it with spray, then re-plugged.

Then when I started it, it would shake badly while idling. Realising I overfilled, I took some oil out and re-balanced the oil level. Waited a day and it would still shake badly. I waited for a couple of more days and when I try starting it would crank but wouldn't start for a while. I connected my VCDS and saw these logs:

Code:
3 Faults Found:
16716 - Knock Sensor 2 (G66) 
            P0332 - 002 - Signal too Low - Intermittent
18711 - Leak in Air Intake System 
            P2279 - 008 -  - MIL ON
17978 - Engine Start Blocked by Immobilizer 
            P1570 - 008 -  - Intermittent
I cleaned the error codes and did a throttle retraining. This fixed the issue and I idled for a couple of minutes, reving occasionally up to 3000rpm. Then I parked it next to driveway downhill. Wait a couple of more days and it won't start. VCDS throttle retraining won't change anything this time.

Pulled the car to level driveway again, retrain the throttle but no change. It's not starting. Then the engine gave these errors:

Code:
17951 - Angle Sensor 1 for Throttle Actuator (G187) 
            P1543 - 002 - Signal too Small - Intermittent
17580 - Angle Sensor 2 for Throttle Actuator (G188) Signal too Low 
            P1172 - 002 -  - Intermittent
17987 - Throttle Actuator (J338) 
            P1579 - 001 - Adaptation Not Started - Intermittent
I think I cleared above errors again to do another throttle training, and after that the car won't start at all (but cranks), and VCDS auto-scan gives the error below and nothing else (tried before and after start attempts, now it's always the same):

Code:
01314 - Engine Control Module 
            013 - Check DTC Memory - Intermittent
ross-tech site tells to fix the ECM problem before anything else for this error.

Since I seem to have induced this problem myself, I don't know if I should do a general won't start troubleshooting or not (and I don't know where to start).
 
   #2  

Jack@European_Parts

NostraJackAss
Joined
Jan 29, 2014
Messages
20,686
Reaction score
8,802
Location
Montgomery, NY, USA
VCDS Serial number
C?ID=57337
Did you recharge battery and then disconnect with reboot?

How about lubing TV with light lube and readapt plus pull plugs to crank and clean cylinders with FP unplugged, than run dry, install plugs/FP and restart?
 
   #3  

mpiroglu

Verified VCDS User
Verified
Joined
Jul 25, 2017
Messages
7
Reaction score
1
Location
Australia
VCDS Serial number
C?ID=312228
Did you recharge battery and then disconnect with reboot?

How about lubing TV with light lube and readapt plus pull plugs to crank and clean cylinders with FP unplugged, than run dry, install plugs/FP and restart?
I did recharge the battery recently to keep cranking, but I didn't disconnect the battery, so the ECU always had power. I only had disconnected the TV to spray-clean it. Do you mean I should disconnect the battery so that the ECU reboots?

And I'll shortly be reading on how to pull plugs. Also FP is fuel pump or fuel pipe :confused: I'll try finding it on my workshop manual but what would be the easiest to unplug the FP? And I'm assuming after plug pull and FP unplugged, just crank it for a while so that the cylinders would clean themselves?

Thanks!
 
   #5  

mpiroglu

Verified VCDS User
Verified
Joined
Jul 25, 2017
Messages
7
Reaction score
1
Location
Australia
VCDS Serial number
C?ID=312228
I finally found time to work on it. Thanks for the suggestions Jack, I was quite in dark before.

The spark plugs were replaced about 30k ago, but when I took them out and checked them, their metal parts were all covered with a black coating. I used a dremel to clean them all and after that the car started perfectly the first attempt and started running quite smoothly.

There was oil around two of the plugs which makes me wonder if it's time to change the engine cover gasket (there's certainly oil spills around the gasket area). But then the other two plugs didn't have any oil.

Also I wonder how all four plugs got fully covered with the black coating. Could the many attempts of cranking do it or it was just the time to clean them and they can still kind of perform like that?

Thanks!
Matt
 
   #6  

mpiroglu

Verified VCDS User
Verified
Joined
Jul 25, 2017
Messages
7
Reaction score
1
Location
Australia
VCDS Serial number
C?ID=312228
Oh well, I spoke too early. Next morning it won't start again. I disconnected the battery for 10min and retried everything but no joy.

Then I wondered if it's the TV, because on two occasions before it ran I had disconnected the TV to do cleaning. I removed it again and checked its terminals. Two pairs have around 0.6ohms initially and they go up to 1.4ohms when I manually open the valve. I don't know if these values are within limits but at least they change. One thing I wonder is the valve is not fully closed by default, and I can't remember if it would sit fully closed when there's no power.

Is there a way to understand if this TV is defective? The previous scans did point to this possibility.as I posted earlier:

Code:
17951 - Angle Sensor 1 for Throttle Actuator (G187) 
            P1543 - 002 - Signal too Small - Intermittent
17580 - Angle Sensor 2 for Throttle Actuator (G188) Signal too Low 
            P1172 - 002 -  - Intermittent
 
   #7  

mpiroglu

Verified VCDS User
Verified
Joined
Jul 25, 2017
Messages
7
Reaction score
1
Location
Australia
VCDS Serial number
C?ID=312228
It's been a while and I have been spending some weekends on the car, which hasn't started yet. I had previous coolant leak under the intake manifold which I hadn't inspected before, so I decided to do some more deep dive.

I removed the timing belt and reinstalled (new bolts, etc), double checking the marks and timing belt position seems correct. Unrelated but replaced coolant thermostat housing as it was leaking. Also the short valve cover breather hose had a crack, replaced that with new one. Intake valves had lots of carbon build-up, I cleaned them as well. Valve cover gasket was leaking oil (and the third spark plug was oily as I mentioned before), I replaced that gasket as well. One of the injectors (fourth one) came out as I was pulling the lower intake manifold (integrated with the fuel line), thus I cleaned it with a spray jig. I also pulled the third injector out to compare, they both spray well so no clogging, then I replaced all injector gaskets and two of the injector's ceramic gasket. Put everything back on and the problem is still the same.

Then I did a compression test. Car spec is 10.5bar with 3bar diff max and 7bars min for each cylinder. I have 10.25, 10.25, 12, 7.6bar. The 12 one is the one with oil in it (most probably from the old valve cover gasket leak), that's why I'm thinking these values are passable. I haven't done a wet test on fourth cylinder yet, but I'm guessing from the intake valve carbon buildup, maybe the exhaust valves are similar/worse and that might be causing low compression on it (I'm guessing it isn't head gasket as I never had issues/symptoms that would point to that).

When I crank, the car occasionally stutters / tries to start, but never does.

As I mentioned before, ignition on gives the usual fuel pump/rail pressure sound. I removed the variable cam timing solenoid and checked resistance (I think it was 7-8ohms, although it seems awfully low), saw it functioning (moving up and down) with 12V, seems plausible but I don't know the expected resistance.

I removed the crankshaft position sensor (i.e. engine speed sensor) and the resistance values are fishy. I mean I can't make anything out of them. It's a three pin sensor, 1-2 and 1-3 are open circuit, and 2-3 is 770ohms. Another unusual value is the voltage coming to pin 2 and pin 3. They are both 2.48V when the ignition is on. Pin 1 is apparently ground. As expected, car doesn't try starting when this sensor isn't connected. And goes back to usual 'tries to occasionally spark mode' after connecting it.

Unfortunately, when I had the intake valve off I haven't checked the knock sensor 2 (G66) which was an intermittent problem during last year's VCDS report. I wonder if I should spend another 3-4h removing intake manifold again just to check that sensor. Mind that VCDS currently doesn't say anything about it (or any other sensor for that matter). Maybe it'll be faster as I don't have to flush the coolant this time, and the fuel rail can stay where it is (intake manifold is two pieces on this car).

Thanks for anyone reading, and I'd greatly appreciate any pointers.
 
   #8  

mpiroglu

Verified VCDS User
Verified
Joined
Jul 25, 2017
Messages
7
Reaction score
1
Location
Australia
VCDS Serial number
C?ID=312228
The last two days were my learning to read the wiring diagrams and hunting the voltage problem I saw on pin2 & 3 of engine speed sensor. Not finding any mention of G28 (engine speed sensor) (or G66 - knock sensor 2) on the elsawin I have, I was a bit at a loss. But then after deciding to clean some dirty sensors I haven't touched, I checked the fuel pressure sensor, then the G62 coolant temp sensor on the left rear of the engine (and they are on the same harness as G28). Moving the fuel pressure sensor fixed the short between pins 2-3 of G28 and I then realized that it was the cables of the G62 coolant sensor that were chafed and touching each other. After fixing that with some insulation tape the car started, but quickly thought that coolant was too hot and started the fans 10 seconds after starting. I stopped the engine and thought maybe the G62 also went bad because of its previous short and replaced it with another one lying around.

Now I am back to no start state again. I'll wait tonight and will try again after the battery fills and the car rests.

But I think I still have some other short, because 1) I don't like the 2.49V on the pin-3 of G28 (engine speed sensor) - although the car started with that today, and 2) the relay R4 on the engine bay keeps getting warm without anything being on. The problem is, I can't find wiring diagrams for this part of the circuits. Neither elsawin nor my useless haynes manual has it. And also VCDS started giving 00532 Auto HVAC B+ 002 Lower limit exceeded error.
 
  • Like
Reactions: Uwe
   #10  

mpiroglu

Verified VCDS User
Verified
Joined
Jul 25, 2017
Messages
7
Reaction score
1
Location
Australia
VCDS Serial number
C?ID=312228
I had read JPPSG before, and did it again. It's a valuable guideline for sure, and I tried following as much as I can.

Here is the update from my latest checks. I managed to use a hantek 1008 to probe the crankshaft and camshaft sensor signals at the same time while cranking.

I then made an image with both signals, attached below.

https://drive.google.com/open?id=1LgPcoLyn16wBInRDcZOeo7KhmYpSfm0m

Red square pulses are camshaft sensor, and blue signal is the crankshaft one. Crankshaft sensor seems very plausible (seems like an inductive CKP), and although the red-camshaft sensor data seems reasonable as well, it doesn't seem to be aligning with the crankshaft sensor. Also note that I checked the resistances of the camshaft pins between each other, and they all seem very high (8+ Mohms each), but this doesn't mean much I think.

Also, readers should note that my previous theories about 2.5V on crankshaft sensor pins being wrong seems incorrect. I'll assume that these pin1 and pin2 2.5V while ignition ON are the expected values.

Also, the coolant temp sensor that I replaced may not be working either. It is becoming annoying to hear the fans come on each time I crank the engine (actually a couple of seconds after).

Currently, I am considering buying a camshaft sensor & a coolant temp sensor and see if that'll change the signal or fix the issues.
 
Top