Strange Engine Starting Issue

[glenmore3685]

Hello

I've got a strange issue with my car (2007 2.0T Saab 9-3 Aero auto estate).

The car starts fine from cold and runs with no issue. If I drive the car and get it up to temperature, then park it and switch off and come back to it about 5/10 minutes later it starts and dies immediately:

Video

If I use the accelerator then it will stay running, run rough for a few seconds and then be completely fine (as I did on the 3rd starting attempt on the video).

I took the car to my local specialist and they were a bit stumped. We mistakenly thought the issue only occurred when the car is full of petrol and they replaced the evac purge valve as there were no codes on the ECU and they were pretty stumped at what else it could be and this is a common problem for the 2.8T engine. This did not resolve the issue however.

I then tried a couple of straight forward fixes myself and cleaned the MAF and MAP sensors and then throttle body, again these have made no difference.

I'd love to get to the bottom of this issue but want to try and avoid the 'replacing parts until I find the right cause' scenario!

As the car will start if you give it some gas makes me think its not an immobilizer/ignition fault, it feels like it could be fuel starvation, although I'm clearly no expert.

Any help would be most appreciated.

[E-bmw]

I know you don't want to do a willy-nilly-part-change approach, but you may need to without any decent pointers.

My first simple & cheap thought would be fuel & air filters and plugs just to rule them out for a baseline.

My best guess would be idle control valve if I were a gambling man, but I am speaking generically as I have no specific experience of your particular engine.

[glenmore3685]

No idle control valve on my engine, there is a tiny gap around the butterfly to control idle.

Throttle body is probably at the top of my list on what to replace first, although your suggestion of the basic items first is probably a good shout. Just feel if it was plugs or filters the issue would be more consistent and appear more often than warm starting?

[bearman68]

I wonder if it is an inlet pipework leak?
When it's hot and starts, it's running on the ECU map, and it's just running a little lean, so cuts out.

Would be interested in the MAP output too.

Just a thought.

Cheers

[mighty kitten]

Measure fuel pressure when you restart it . These have the Pressure Reg on the pump in the tank and I've had an issue with one with poor performance that was the o ring split . If nothing else it will rule out the fuel supply side

[bearman68]

I've just bought a very similar car, so keep us posted.

BTW, how is the fuel pressure measured - is there a fuel pressure sensor?

[glenmore3685]

Will do, not sure how I can test the fuel pressure - hopefully someone in the know comes along soon :-)

I've had the car 10 months and I'm pretty please with it aside from this issue. I could live with it but the idea is for my wife to be the main user so I want to get it sorted as I don't think she'll appreciate a car with an intermittent starting problem!

CPS is another common failure point on these cars but I don't get any error codes and my garage said if that was on the blink the rev counter doesn't move when you try and start it either.

[E-bmw]

Fuel pressure can normally be monitored either manually at one end of the fuel rail with a gauge on what looks like (and is) a car tyre type shrader valve, or by plugging a reader into the ecu, the cheap £30 ebay code readers/reset devices will suffice.

[bearman68]

I've just had a look at the engine management wiring diagram and not found an electronic method of measuring fuel pressure. Therefore I would assume that fuel pressure is controlled via the traditional vac valve in the fuel line. These are pretty reliable, so I would imagine the fault lies else where.

[glenmore3685]

Thanks for checking.

Do you think fuel is the most likely suspect? As you can see from the video it does appear to start, albeit for a split second and then it completely dies. Would this be synonymous with too much fuel (flooding) or too little? If there was a fuel starvation issue would it not cough and splutter and then stall?

I have ruled out ignition initially as a dab of the accelerator sorts its out, but I'm no expert so happy to be corrected.

[E-bmw]

I do not know (obviously) but my thought with fuel would be as follows.

When you start up there is still some fuel pressure in the fuel rail, this very quickly is burnt and the engine stops.

Quite what changes between that & the second start, I am not sure, but you need to start eliminating things, hence why I said to start with a known baseline with new air/fuel filter & eliminate the plugs while you are there.

Then as mentioned start checking things & fuel pressure is a good starting point after the basics.

[bearman68]

E-bmw said:

I do not know (obviously) but my thought with fuel would be as follows.

When you start up there is still some fuel pressure in the fuel rail, this very quickly is burnt and the engine stops.

Quite what changes between that & the second start, I am not sure, but you need to start eliminating things, hence why I said to start with a known baseline with new air/fuel filter & eliminate the plugs while you are there.

Then as mentioned start checking things & fuel pressure is a good starting point after the basics.

There is merit in that argument for me - in which case the fuel pump is the suspect component.
If you have no particularly sophisticated tools,it's worth just lifting the back seats and listening for the fuel pump when it doesn't start.
The more 'professional' method of doing this is to monitor the output of the O2 sensor to see if the output goes lean or rich. (I don't think the scan tool will be fast enough for this).
It is possible to fit an in-line fuel pressure gauge into the line to see what the pressure is doing. But you need to understand what the desired is for this.
You can also look to see what injection duration you have on the injectors - will give you an indication of rich or lean.

Cheers

[glenmore3685]

Cool, thanks both. Will try and replicate it and go and listen out for the fuel pump. Then I'll start to narrow it down by replacing the basics if it doesn't lead to anything.

[Auntieroll]

There is often a check/non-return valve on the fuel pump outlet,this maintains the rail pressure on switch off,it almost seems like vapourisation from the rail getting heat soaked and boiling the fuel within, as sometimes experienced on a very hot day in the "good old" days of carbs.

[Auntieroll]

There is often a check/non-return valve on the fuel pump outlet,this maintains the rail pressure on switch off,it almost seems like vapourisation from the rail getting heat soaked and boiling the fuel within, as sometimes experienced on a very hot day in the "good old" days of carbs.

[Peanut Gallery]

Auntieroll said:

There is often a check/non-return valve on the fuel pump outlet,this maintains the rail pressure on switch off,it almost seems like vapourisation from the rail getting heat soaked and boiling the fuel within, as sometimes experienced on a very hot day in the "good old" days of carbs.

This. I had loads of this sort of problem with my manual fuel pump carb cars. (as opposed to electric fuel pump) - drove me mental on hot days.

Somewhere fuel is vaporising, and there is just enough fuel left to get the car to fire. What happens if you turn the ignition on, wait for 5 seconds or so for the fuel pressure system to bring itself up to full pressure, then try start?

[glenmore3685]

Peanut Gallery said:

This. I had loads of this sort of problem with my manual fuel pump carb cars. (as opposed to electric fuel pump) - drove me mental on hot days.

Somewhere fuel is vaporising, and there is just enough fuel left to get the car to fire. What happens if you turn the ignition on, wait for 5 seconds or so for the fuel pressure system to bring itself up to full pressure, then try start?

Just to clarify it is an electric setup in my car. I will give it a few seconds next time and see, but I'm sure I've given it an adequate gap in previous attempts but it didn't seem to make a difference but I'm not 100% sure.

[glenmore3685]

Also just to clarify if I try and start the car immediately after I've switched it off, its fine. The problem only manifests if the car is left for 5 minutes or more, or once the engine is completely cold its fine.

Edited by glenmore3685 on Monday 9th November 11:31

[Peanut Gallery]

glenmore3685 said:

Also just to clarify if I try and start the car immediately after I've switched it off, its fine. The problem only manifests if the car is left for 5 minutes or more, or once the engine is completely cold its fine.

To me this starts pointing all the more at a point in the fuel system that is gathering heat and vaporising the fuel.

Just after turn off, the fuel in the system is still cool, and has not vaporized yet.
5 min later, the fuel has vaporized.
Once the engine is cooler, the fuel has turned back into a liquid.

It probably wont work, but have you thought about insulating any fuel lines in the engine bay where heat might collect once the car stops moving? - Checking fuel lines are not resting on engine block, head etc?

I am not saying this is definitely the problem!

[glenmore3685]

bearman68 said:

If you have no particularly sophisticated tools,it's worth just lifting the back seats and listening for the fuel pump when it doesn't start.


Cheers

OK so I tried to replicate it today and switched the radio off so I could listen out for the fuel pump. Replicated the issue and I could hear the fuel pump working quite hard when I tried to restart it. Not sure what that proves now though!