Warm starting problem on TDi75

AndyP

Member
United-Kingdom
Stone cold or hot starting is no problem at all, but if I've done a couple of miles, stopped and gone to start 5-10 mins later it takes a lot of cranking to start. This is quite predicitable, This has been the case for a couple of years but as I seldom do short journeys it doesn't matter much. Now, just recently a new development, after a difficult restart the car runs fine for anything from a few hundred yards to a mile or so and then stops responding to the throttle and just glides to a halt. After a lot more cranking it will restart, sometimes after a shortly stalling again and after that it will be absolutely fine. This has also happenned a few times now and is becoming very annoying and to add to the irritation there are no ECU codes. Aside from that the car runs perfectly. This appears to be temperature related so I changed the fuel temp sensor, but it made no difference. Any thoughts or inspiration?
 
Which scan tool are you using? Similar to our cam position sensor failure though VCDS reported it. Injector loom?
 
Either injector loom or fuel issue. Tandem pump leaking / worn, air in fuel line, blocked or dirty fuel filter. How good has the servicing been?
Injector loom really does have a hard life with the connections to the injectors being covered in oil. As it costs far less than the Tandem pump I would go for that first. Check all the rubber hoses around the Tandem pump for signs of diesel contamination, if found replace them and the pump A.S.A.P.
Check the 2 fuel pipes on the Tandem are in good condition and tight. Replace the fuel filter. As far as codes go the injector loom does not (normally) trip a code, nor does the Tandem. If using VCDS fuel trim may point out and injector fueling mismatch which in turn could point to the loom but do not rely on that. If you end up changing the Tandem - a better option than the reseal kit, then fit a Bosch replacement not a Luk and get a good quality OEM gasket as not supplied with the pump. If the rubber hoses have been attacked by diesel then they will ALL have to be replaced at the earliest as you could suddenly find all coolant lost. If the fuel temp sensor had failed, that does show in VCDS, although if it has failed or become disconnected the only difference I found was the ECU uses a slightly different fuel map using a default temp for the fuel. There is a pressure tapping on the Tandem pump that CAN be used to diagnose the pump , but you will need to find a good diesel specialist to use it. Plenty of threads on here about injector loom and Tandem pump.
 
I think you broadly have a choice of two possibilities here; its either a fuel issue or a sensor/wiring issue, I'm betting on a fuel issue. Here is what I would do to confirm or rule out a fuel issue. Temporarily replace a section of the fuel return line, which runs from the tandem pump (lower fuel connection) to the tank, with a clear plastic tube. If the problem is air in the fuel you will see this air bled out as bubbles in the clear tube when trying a difficult restart. Return fuel can get quite hot after the engine has heated up so be careful not to leave the clear plastic tube in place other than for lowish temperature diagnosis.

If you see air bubbles you then know it is a fuel line issue. Given the symptoms (being able to drive for a mile before it stops again) I would suggest the the leak would not be in the engine bay but either on the connections to the fuel filter or tank.

Note: Since the feed fuel line is always under negative pressure you might not see a small leak as it may only ever be sucking air in and not leaking diesel out.

If no air in clear fuel return line then you are back to looking at sensors and wiring.

Hope that helps.

Trevor
 
I think you broadly have a choice of two possibilities here; its either a fuel issue or a sensor/wiring issue, I'm betting on a fuel issue. Here is what I would do to confirm or rule out a fuel issue. Temporarily replace a section of the fuel return line, which runs from the tandem pump (lower fuel connection) to the tank, with a clear plastic tube. If the problem is air in the fuel you will see this air bled out as bubbles in the clear tube when trying a difficult restart. Return fuel can get quite hot after the engine has heated up so be careful not to leave the clear plastic tube in place other than for lowish temperature diagnosis.

If you see air bubbles you then know it is a fuel line issue. Given the symptoms (being able to drive for a mile before it stops again) I would suggest the the leak would not be in the engine bay but either on the connections to the fuel filter or tank.

Note: Since the feed fuel line is always under negative pressure you might not see a small leak as it may only ever be sucking air in and not leaking diesel out.

If no air in clear fuel return line then you are back to looking at sensors and wiring.

Hope that helps.

Trevor
Just a thought , I don’t know if it’s possible to see the fuel return pressure via VCDS & compare actual against expected readings
Hth
Keith
 
I have two scanners, basic Torque and the free RossTech (I think). I'm with you guys on the fuel theories, how else can you stop a diesel, right? So what I didn't mention is that it's done 183K and servicing has been spot on, initially by Audi and the last 130K by me.
The fuel filter is good, there are no (obvious) leaks on the tandem pump or anywhere else. What doesn't make sense is why only when warm, if it's an air leak problem on the return pipe then surely it should manifest on any restart, even overnight, why only when warm? That is why I tried the fuel temp sensor first. Just for kicks I have also taken apart the fuel sender unit under the back seat and looked at the return pipes - also no leaks. Tandem pump, as before, also why when warm, the rest of the time it is perfect, pulls fine etc. The only little thing nagging at me now is that it has been remapped, so I might put the standard map back, but I still don't see how that would affect the stalling, but I can see that if there was an issue with that maps that it could cause poor starting when warm.
As for injector wiring loom, this is not showing the normal failure profile for that - that seems to be misfire on a single cylinder, usually under load and getting worse with time. Also, electrical wiring/connections typically fail when hot, or when cold, not when warm, so I think I'm considering that unlikley for now.
 
It is not an air leak in the return pipe that you are looking for. An air leak in the return pipe wouldn't matter a jot as it goes back to the tank. The reason for the clear tube on the return pipe is to see if the engine is bleeding air out of the fuel. Said bled air ends up in the return pipe.

My thinking was that if you had just the right small leak at the fuel filter (near rear wheel) or the tank connection then the car starts and runs for a mile or so with the fuel in the pipes or filter (after the leak). The slug of air from the leak is drawn through the system until it reaches the engine. If its only a small amount of air the system will self bleed and you wouldn't notice anything if you keep engine running. But if you stopped the engine just as the system is trying to bleed the air then it will be difficult to restart. Hence hot start is ok as the air has been bled out a while ago; cold start is ok since the leak is a long way from the engine; luke warm start difficult due to slug of air. Your description of it stopping again a few hundred yards down the road might suggest the leak is getting worse.

Just a theory that could be confirmed or denied with a length of plastic tube.

Trevor
 
Wiring loom as I said has a very tough life and at 183K will be well past its best. It can work perfectly sometimes, then when it decides not to play can be various symptoms including misfires, reluctant starting, leaks, fuel economy change, stalling etc. Can have single or multiple issues. remember the cam that drives the Tandem will be worn as will the internals be so at that mileage I would replace both. Any other sensor would at very least give a code or warning lamp. Another thing to bear in mind is the A2 does not have a seperate fuel pump, the Tandem has to drag it from the tank through the filter and then to the injectors so works hard, any wear could let the fuel back down under gravity. Fit a short length of clear fuel pipe just before the inlet side of the Tandem and see if you are having to "prime" the Tandem.
 
Based on the information available, my suspicion is that the camshaft or crankshaft position sensor could be dying a slow death. You wouldn't be the first person to struggle with intermittent slow starting issues due to either of these sensors providing the ECU with dodgy readings. Dodgy sensors don't always trigger fault codes, and can be temperature sensitive during their dying days.

Cheers,

Tom
 
It is not an air leak in the return pipe that you are looking for. An air leak in the return pipe wouldn't matter a jot as it goes back to the tank. The reason for the clear tube on the return pipe is to see if the engine is bleeding air out of the fuel. Said bled air ends up in the return pipe.

My thinking was that if you had just the right small leak at the fuel filter (near rear wheel) or the tank connection then the car starts and runs for a mile or so with the fuel in the pipes or filter (after the leak). The slug of air from the leak is drawn through the system until it reaches the engine. If its only a small amount of air the system will self bleed and you wouldn't notice anything if you keep engine running. But if you stopped the engine just as the system is trying to bleed the air then it will be difficult to restart. Hence hot start is ok as the air has been bled out a while ago; cold start is ok since the leak is a long way from the engine; luke warm start difficult due to slug of air. Your description of it stopping again a few hundred yards down the road might suggest the leak is getting worse.

Just a theory that could be confirmed or denied with a length of plastic tube.

Trevor

Yes, sorry, I wasn't thinking clearly about that. I agree that is a good test.
 
Based on the information available, my suspicion is that the camshaft or crankshaft position sensor could be dying a slow death. You wouldn't be the first person to struggle with intermittent slow starting issues due to either of these sensors providing the ECU with dodgy readings. Dodgy sensors don't always trigger fault codes, and can be temperature sensitive during their dying days.

Cheers,

Tom

Yes, good suggestion. In the past I have had the occasional crank/camshaft sensor error but not recently, I did buy a replacement for one of them, can't remember which but haven't got around to fitting it, so I will definitely give that a try. However, I still would not expect a "warm" failure from an electrical component.
 
Wiring loom as I said has a very tough life and at 183K will be well past its best. It can work perfectly sometimes, then when it decides not to play can be various symptoms including misfires, reluctant starting, leaks, fuel economy change, stalling etc. Can have single or multiple issues. remember the cam that drives the Tandem will be worn as will the internals be so at that mileage I would replace both. Any other sensor would at very least give a code or warning lamp. Another thing to bear in mind is the A2 does not have a seperate fuel pump, the Tandem has to drag it from the tank through the filter and then to the injectors so works hard, any wear could let the fuel back down under gravity. Fit a short length of clear fuel pipe just before the inlet side of the Tandem and see if you are having to "prime" the Tandem.

I agee that it feels like the tandem is being primed, but why only when warm, surely if it needs priming when warm, it would still need priming if it had cooled downn as well?
 
I agee that it feels like the tandem is being primed, but why only when warm, surely if it needs priming when warm, it would still need priming if it had cooled downn as well?
Not if the leak is some way away from the engine as theorised above.
 
Not an expert on fueling mapping, but if the engine is warm it may be on a different part of the maap and is more sensitive to loom or tandem pump issues. Warm could mean the rubber has softened and is porous letting air in, that may seal again as it gets hotter, the wear in the pump is not holding fuel in the line, the fuel filter is letting in air. You now must find out if the pump is having to suck fuel all the way up, if bubbles are coming from the supply pipe or if the fuel is dropping back from the pump. Could also be a resistance change in the oil soaked loom. Now at the point where if you can't find out if it is fuel issue, I think we will just go around in circles. The mileage of the vehicle I would not complain about changing the cam and crank sensors as well as the loom and Tandem pump.
 
First thing to do would be to scan with VCDS. Sensors/loom should cause fault codes whereas fuel issues may not.

RAB
 
I have two scanners, basic Torque and the free RossTech (I think). I'm with you guys on the fuel theories, how else can you stop a diesel, right? So what I didn't mention is that it's done 183K and servicing has been spot on, initially by Audi and the last 130K by me.
The fuel filter is good, there are no (obvious) leaks on the tandem pump or anywhere else. What doesn't make sense is why only when warm, if it's an air leak problem on the return pipe then surely it should manifest on any restart, even overnight, why only when warm? That is why I tried the fuel temp sensor first. Just for kicks I have also taken apart the fuel sender unit under the back seat and looked at the return pipes - also no leaks. Tandem pump, as before, also why when warm, the rest of the time it is perfect, pulls fine etc. The only little thing nagging at me now is that it has been remapped, so I might put the standard map back, but I still don't see how that would affect the stalling, but I can see that if there was an issue with that maps that it could cause poor starting when warm.
As for injector wiring loom, this is not showing the normal failure profile for that - that seems to be misfire on a single cylinder, usually under load and getting worse with time. Also, electrical wiring/connections typically fail when hot, or when cold, not when warm, so I think I'm considering that unlikley for now.
I've had 2 scanners as well, similar to yours and neither showed issues unless the EML came on. Now I got a vag401 and it shows other bits too even if EML not on, so might worth to ask a mate with vcds to do another scan
 
General rules (but there can be exceptions) on cam and crank sensors which may help in your troubleshooting -

Cam position sensor tells engine ECU when each cylinder is approaching TDC on the firing stroke.
If it fails the crank sensor takes over but the engine can spin over several times before it works out the right firing stroke = so longer cranking time.
The engine will start and run with the cam sensor disconnected - but will just crank over a few more times before firing up

Crank speed sensor does everything else but if it fails it's normally the only sensor which will stop the engine running. Normal failure mode is to cause engine hesitation and eventually stop the engine when the sensor gets hot. In the early stages it may not trigger any fault codes.

Usually only happens if cam timing has not been done correctly or the belt has skipped a tooth but if the cam and crank sensor signals are out of sync the engine won't start. Quick check is to unplug the cam sensor and it should fire up.

Cheers Spike
 
General rules (but there can be exceptions) on cam and crank sensors which may help in your troubleshooting -

Cam position sensor tells engine ECU when each cylinder is approaching TDC on the firing stroke.
If it fails the crank sensor takes over but the engine can spin over several times before it works out the right firing stroke = so longer cranking time.
The engine will start and run with the cam sensor disconnected - but will just crank over a few more times before firing up

Crank speed sensor does everything else but if it fails it's normally the only sensor which will stop the engine running. Normal failure mode is to cause engine hesitation and eventually stop the engine when the sensor gets hot. In the early stages it may not trigger any fault codes.

Usually only happens if cam timing has not been done correctly or the belt has skipped a tooth but if the cam and crank sensor signals are out of sync the engine won't start. Quick check is to unplug the cam sensor and it should fire up.

Cheers Spike

Thanks, that is pretty much what I expected. I have ordered a cam sensor and will replace that at the weekend. After that is eliminated we move into fuel diagnostics ...
 
Based on the information available, my suspicion is that the camshaft or crankshaft position sensor could be dying a slow death. You wouldn't be the first person to struggle with intermittent slow starting issues due to either of these sensors providing the ECU with dodgy readings. Dodgy sensors don't always trigger fault codes, and can be temperature sensitive during their dying days.

General rules (but there can be exceptions) on cam and crank sensors which may help in your troubleshooting -

Cam position sensor tells engine ECU when each cylinder is approaching TDC on the firing stroke.
If it fails the crank sensor takes over but the engine can spin over several times before it works out the right firing stroke = so longer cranking time.
The engine will start and run with the cam sensor disconnected - but will just crank over a few more times before firing up

Crank speed sensor does everything else but if it fails it's normally the only sensor which will stop the engine running. Normal failure mode is to cause engine hesitation and eventually stop the engine when the sensor gets hot. In the early stages it may not trigger any fault codes.

Usually only happens if cam timing has not been done correctly or the belt has skipped a tooth but if the cam and crank sensor signals are out of sync the engine won't start. Quick check is to unplug the cam sensor and it should fire up.

Cheers Spike
Spike's explanation is infinitely better than mine. :)
 
Back
Top