Audi A3 petrol intermittent stalling and not starting
Discussion
Hi,
I could do with some help in diagnosing an intermittent issue on my Audi A3 1.6 petrol 2010 car with 75k on the clock.
A few months ago it cranked fine but wouldn't start - didn't even sound like it would fire. This was totally out of the blue, the car has been fine in the previous 2 years i've owned it. I used OBDeleven to check for codes and only found a code for the rear number plate bulb - i had replaced that a few weeks before. I left the car 10 mins or so and it then started fine. I thought no more of it.
A few weeks later it did it again (engine was last started an hour before) but this time it started as soon as i turn the ignition key off and on.
Now since then it hasn't done that but on a couple of occasions it's felt like it's wanted to cut out as i changed gear when turning off a road - but then picked up again.
Forward to today and it's again lost power when coming off a roundabout (again as i changed gear) but this time it didn't recover and died. I pulled over to side of the road, left it a few seconds and it fired up fine.
Now i'm confident when it comes to fixing cars (built 4 kit cars and rebuilt a couple of engines) but not so great on diagnostics. At least if the fault was permanent i could do the obvious checks for fuel and spark but the intermittent nature makes it difficult. The lack of fault codes doesn't help but i guess i could take it somewhere with the proper diagnostic gear to get it checked. At the moment i'm thinking crankshaft sensor or maybe the switch on the clutch pedal (that has to be depressed to allow it to start) but they are guesses if i'm honest.
Any pointers please guys?
Thanks
Simon
I could do with some help in diagnosing an intermittent issue on my Audi A3 1.6 petrol 2010 car with 75k on the clock.
A few months ago it cranked fine but wouldn't start - didn't even sound like it would fire. This was totally out of the blue, the car has been fine in the previous 2 years i've owned it. I used OBDeleven to check for codes and only found a code for the rear number plate bulb - i had replaced that a few weeks before. I left the car 10 mins or so and it then started fine. I thought no more of it.
A few weeks later it did it again (engine was last started an hour before) but this time it started as soon as i turn the ignition key off and on.
Now since then it hasn't done that but on a couple of occasions it's felt like it's wanted to cut out as i changed gear when turning off a road - but then picked up again.
Forward to today and it's again lost power when coming off a roundabout (again as i changed gear) but this time it didn't recover and died. I pulled over to side of the road, left it a few seconds and it fired up fine.
Now i'm confident when it comes to fixing cars (built 4 kit cars and rebuilt a couple of engines) but not so great on diagnostics. At least if the fault was permanent i could do the obvious checks for fuel and spark but the intermittent nature makes it difficult. The lack of fault codes doesn't help but i guess i could take it somewhere with the proper diagnostic gear to get it checked. At the moment i'm thinking crankshaft sensor or maybe the switch on the clutch pedal (that has to be depressed to allow it to start) but they are guesses if i'm honest.
Any pointers please guys?
Thanks
Simon
When the fuel pump is running, can you normally hear it? If so, keep an ear out for that. There are umpteen possible reasons for this type of problem, but the intermittent nature combined with clearing after a key off and without fault codes is a good fit for a failing fuel pump relay. If you're inclined to diagnose it yourself, a test lamp wired across the pump motor and positioned where you can see it would show whether this was a simple electrical problem. Note that some cars have multiple fuel pumps so you might need to check in more than one place. Alternatively, if you have the test equipment and don't mind getting your hands dirty, install a remote fuel pressure gauge. If your OBD scan tool has a working pid for fuel pressure, that would be an even better way to check.
Thanks for the reply. The fuel pump is triggered by opening the drivers door and you can hear it run. I'll look into the idea of checking the fuel pressure as i do have a remote fuel pressure gauge - i'll have to see what OBDeleven can do, i know it has a lot of options for live data.
Alternatively i could just replace the relay anyway. A quick google shows i could get a Topran relay for less than £4. Not sure if that's the one to go for but even if it was a £10 or so it's still worth a gamble. I'll go and do some digging on relay prices and see what an OEM unit will cost.
Off to the garage now to mate my Audi transaxle to BMW V12 engine after doing a top end rebuild. Working on my daily driver is boring in comparison
Cheers
Simon
Alternatively i could just replace the relay anyway. A quick google shows i could get a Topran relay for less than £4. Not sure if that's the one to go for but even if it was a £10 or so it's still worth a gamble. I'll go and do some digging on relay prices and see what an OEM unit will cost.
Off to the garage now to mate my Audi transaxle to BMW V12 engine after doing a top end rebuild. Working on my daily driver is boring in comparison
Cheers
Simon
Bit of an update. I had a fuel filter on the shelf that i hadn't got around to replacing so did that today. Also swapped the fuel pump relay with the horn relay - i couldn't get a replacement on a Sunday so a swap will have to do for now. I'll order a new relay though as no guarantee the horn relay is any better.
Will see what this week brings.
Thanks
Simon
Will see what this week brings.
Thanks
Simon
Another update.
Over the last 2 weeks the car has not started and stalled on numerous occasions (both on with a fully warmed engine) but always when i'm out and about so difficult to diagnose when i'm on my own. Anyway i rigged up a fuel pressure gauge and a spark plug tester and finally managed to get the car to not start on my driveway after a few hours of running the car, leaving it to idle, turn off and on etc.
I can confirm when the car cranks but won't start there is no spark. The fuel pressure is good, i can hear the pump running and the voltage across the pump terminals is 13.5 or so. So at last i've managed to narrow it down to ignition. I believe that means the problem could be;
Ignition coil/module (there is only 1 coil on this car with standard HT leads)
Crank sensor (i replaced this 2 weeks ago on a hunch) with a Febi-Bilstein unit)
Cam sensor
ECU
Wiring
Now i only had a few minutes to find the fault as the car normally starts up after 5-10 mins and sure enough it was about that time yesterday when it started. I could't find any obvious wiring issues in that short time.
I am still not getting any fault codes (i can force a fault code by unplugging a relay so i know the code reader is working). I need a plan of action for when i reproduce the issue (hopefully today).
My plan is to;
1) Measure the coil inputs/outputs when it's running ok and compare to when it's not firing. I am led to believe the control signals being the same would lead to a faulty coil.
2) Check for an RPM signal on the live data (to check the crank sensor is ok)
3) Test if the injectors are trying to fire - Will this help diagnose a camshaft sensor issue? Does a camshaft sensor failure mean no spark?
Does the above plan sound ok? The issue only occurs with a warm engine and didn't occur during the recent snow/ice period so maybe that's a clue.
I'll be grateful for any other suggestions.
Thanks
Simon
Over the last 2 weeks the car has not started and stalled on numerous occasions (both on with a fully warmed engine) but always when i'm out and about so difficult to diagnose when i'm on my own. Anyway i rigged up a fuel pressure gauge and a spark plug tester and finally managed to get the car to not start on my driveway after a few hours of running the car, leaving it to idle, turn off and on etc.
I can confirm when the car cranks but won't start there is no spark. The fuel pressure is good, i can hear the pump running and the voltage across the pump terminals is 13.5 or so. So at last i've managed to narrow it down to ignition. I believe that means the problem could be;
Ignition coil/module (there is only 1 coil on this car with standard HT leads)
Crank sensor (i replaced this 2 weeks ago on a hunch) with a Febi-Bilstein unit)
Cam sensor
ECU
Wiring
Now i only had a few minutes to find the fault as the car normally starts up after 5-10 mins and sure enough it was about that time yesterday when it started. I could't find any obvious wiring issues in that short time.
I am still not getting any fault codes (i can force a fault code by unplugging a relay so i know the code reader is working). I need a plan of action for when i reproduce the issue (hopefully today).
My plan is to;
1) Measure the coil inputs/outputs when it's running ok and compare to when it's not firing. I am led to believe the control signals being the same would lead to a faulty coil.
2) Check for an RPM signal on the live data (to check the crank sensor is ok)
3) Test if the injectors are trying to fire - Will this help diagnose a camshaft sensor issue? Does a camshaft sensor failure mean no spark?
Does the above plan sound ok? The issue only occurs with a warm engine and didn't occur during the recent snow/ice period so maybe that's a clue.
I'll be grateful for any other suggestions.
Thanks
Simon
It's good to be able to narrow it down to the ignition, although that still leaves lots of possible causes.
Each ignition coil will have two connections - 12V and ground. One will be a constant supply, and the other will be switched by the ECU. In my experience the constant supply is usually 12V and the ECU switches the ground, but don't assume that. If you have a coil pack with multiple coils in it, the constant supply side will usually be common.
The constant supply will usually be provided by a relay and if so that can cause the same heat related problems for the ignition that can affect the fuel pump. Since this is the easiest type of problem to diagnose, I'd check for that first.
The constant supply may be connected with the ignition on, or may only be live when the engine is turning with the ignition on. You should be able to find that fairly easily with a volt meter or test lamp. Then see whether the behaviour is different when the problem happens.
If the constant supply side of the coil is OK, the next thing to check would be the ECU switched output. This is much harder to test electrically. One way to test it is to connect a 12V LED across it. This should light when the ECU switches the coil on. At cranking speed the flashes will be very brief but would be visible in low light. Given the symptoms it's unlikely that the coils themselves are failing, but it's worth a sanity check just to confirm that they aren't being switched on when the problems happens.
Assuming that's the case, either it's a problem within the ECU itself (pretty unlikely) or a sensor input is missing. Depending on the ECU design it might need a cam phase signal as well as the crank position signal. I don't know how the tacho is driven on your car, but on the ones I've seen the tacho was driven from the crank position sensor so if the tacho shows the engine speed while cranking, that suggests it is seeing at least that sensor input. To get further, you need to know more about the behaviour of that particular ECU, or test by substitution. One way to work out the behaviour is to disconnect a given sensor while the engine is healthy, and see what the symptoms are (in terms of tacho output, OBDII pids etc). If you see similar symptoms when the problem happens that's a clue to the cause.
Each ignition coil will have two connections - 12V and ground. One will be a constant supply, and the other will be switched by the ECU. In my experience the constant supply is usually 12V and the ECU switches the ground, but don't assume that. If you have a coil pack with multiple coils in it, the constant supply side will usually be common.
The constant supply will usually be provided by a relay and if so that can cause the same heat related problems for the ignition that can affect the fuel pump. Since this is the easiest type of problem to diagnose, I'd check for that first.
The constant supply may be connected with the ignition on, or may only be live when the engine is turning with the ignition on. You should be able to find that fairly easily with a volt meter or test lamp. Then see whether the behaviour is different when the problem happens.
If the constant supply side of the coil is OK, the next thing to check would be the ECU switched output. This is much harder to test electrically. One way to test it is to connect a 12V LED across it. This should light when the ECU switches the coil on. At cranking speed the flashes will be very brief but would be visible in low light. Given the symptoms it's unlikely that the coils themselves are failing, but it's worth a sanity check just to confirm that they aren't being switched on when the problems happens.
Assuming that's the case, either it's a problem within the ECU itself (pretty unlikely) or a sensor input is missing. Depending on the ECU design it might need a cam phase signal as well as the crank position signal. I don't know how the tacho is driven on your car, but on the ones I've seen the tacho was driven from the crank position sensor so if the tacho shows the engine speed while cranking, that suggests it is seeing at least that sensor input. To get further, you need to know more about the behaviour of that particular ECU, or test by substitution. One way to work out the behaviour is to disconnect a given sensor while the engine is healthy, and see what the symptoms are (in terms of tacho output, OBDII pids etc). If you see similar symptoms when the problem happens that's a clue to the cause.
Hi,
Thanks for the comprehensive reply. I went to do the testing but decided to take the coil of the car and test it first on the bench. What i found was badly corroded terminals for cylinders 3 + 4. See a pic below to show what this looks like and a pic of one of the good terminals for cylinder 1 + 2. According to the Haynes manual the coils are split 1 & 4 and 2 & 3 so this seems to show an issue on both sides.
Do you think this could have been a cause of my issue and do you have any idea what has caused this corrosion? I've now replaced the coil and HT leads anyway. A test drive of 35 miles has not shown any issues but i've done 150 miles previously before the issue arises so i'm not getting carried away just yet,
[url]
|https://thumbsnap.com/bPyFkPUw[/url]
Thanks for the comprehensive reply. I went to do the testing but decided to take the coil of the car and test it first on the bench. What i found was badly corroded terminals for cylinders 3 + 4. See a pic below to show what this looks like and a pic of one of the good terminals for cylinder 1 + 2. According to the Haynes manual the coils are split 1 & 4 and 2 & 3 so this seems to show an issue on both sides.
Do you think this could have been a cause of my issue and do you have any idea what has caused this corrosion? I've now replaced the coil and HT leads anyway. A test drive of 35 miles has not shown any issues but i've done 150 miles previously before the issue arises so i'm not getting carried away just yet,
[url]
|https://thumbsnap.com/bPyFkPUw[/url]
It's obviously not right so worth fixing anyway, and see whether that helps the starting problem. Corrosion usually indicates it's getting damp. The plug lead should seal to it but perhaps not well enough. When you fit the replacements, a light smear of dielectric grease on the inside of the boot will help it seal and reduce corrosion.
Gassing Station | Engines & Drivetrain | Top of Page | What's New | My Stuff