Missed start, EML, no code

[Deanxm]

Evening gents

Thinking out loud here,

Done a lot of miles in the XM recently but today i had something similar to an old issue i thought i had cured and probably have but im not sure.
From time to time i used to get a warm restart issue, if anyone who follows my blog on clubxm will remember this was tracked to low fuel delivery pressure, this issue never gave an EML if i remember correctly though and would be woolly with no fuel smell when it did start.

Today i had a similar issue, a warm restart (20 ish miles driven, stopped for 30mins) went back to the car to start and it just cranked for longer than usual and no start, cranked again for about 7 seconds which is an eternity for this car, still nothing so while still cranking i gave it some right foot and it just started, no splutter or farting just fired right up, it smelled of fuel something awefull though as i reversed through the exhaust so nothing like the fuel pressure issue, as soon as it fired i also got the EML light on and it stayed on for the half hour drive 5 or 6 miles crawling through the roadworks to the office but it ran like it was submersed in honey. Restarted fine straight away as a test and again half hour later when i went home with no EML light in sight after turning the engine off the first time.

Now it sat a while as my diag laptop died and i had to take that apart and tighten down the CPU to get it to work (not relevant), plugged the car in and no fault code, everything is fault code clean which is damn annoying.

Im thinking it was getting fuel due to the smell so its got to be ignition related but then why would it run fine? the silky idle after the issue would suggest a duff sensor the ecu was ignoring on the drive back to the office but that would surely log as a fault on the ecu, and it hasnt? so its got to be a sensor that gives variable reading (air temp, coolant temp) that the ecu wouldnt see as faulty but given wrong info the ecu may have trouble starting and not know why hence the EML with no actual fault logged.

Coolant and air temp sensors are giving believable reading though as seen on live data screen of Lexia.............. im going to sleep on it, or ignor it and wait another year to see if it happens again

Since you probably fell asleep 3 paragraphs back anyway, im also wondering why a dinosaur of an engine management system such as the XM's would be being fed gear selection info from the gearbox .

D

[Mandrake]

Is it a petrol engine ?

If so are you sure it isn't vapour locking in the fuel rail ?

[Deanxm]

It is Petrol, the 8v XU10 engine.
Vapour lock was the problem last time when delivery pressure was low due to an aging pump, this meant the fuel rail didnt pressurise and open the regulator to create a circuit back to the tank, the fuel rail just acted like a dead leg if you see what i mean?.
When i had that issue it would run really rough for a few seconds when it did start and you could reproduce the symptoms by letting the engine cook for just the right amount of time.

Having said that its always puzzled me that the fuel pump does not prime when you turn the ignition on, only when the engine is cranking or running, it starts well though even when sat for ages, every single time the engine goes through two compression cycles and fires, except this time.

you have made me think now though, if the fuel pump doesnt prime, it could vapour lock but the rail holds fuel pressure for a very long time so it shouldnt make any difference other than to flush the vapour out of the rail and replace with cool fuel prior to cranking. mmmm

I will find a Motronic injection manual and see if the fuel pump should be priming at ignition on, i know it never has in the 12 years ive had it.

D

[Mandrake]

It is Petrol, the 8v XU10 engine.
Having said that its always puzzled me that the fuel pump does not prime when you turn the ignition on, only when the engine is cranking or running, it starts well though even when sat for ages, every single time the engine goes through two compression cycles and fires, except this time.

I think it's more of a safety issue - if the pump could come on for any significant length of time with the engine not running then that could pump fuel through a split hose and onto a hot engine in the case of a crash - most ECU systems will cut the fuel when no engine RPM signal is detected so that the fuel pump doesn't keep pumping fuel out through a split fuel line into the wreckage of an accident.

On my Xanita V6 the pump will run for half a second when the key is turned on after it has been off for a long time, however subsequent key off/on cycles will not trigger the pump again. If it's vapour locked half a second is not nearly long enough to clear the vapour lock though.

you have made me think now though, if the fuel pump doesnt prime, it could vapour lock but the rail holds fuel pressure for a very long time so it shouldnt make any difference other than to flush the vapour out of the rail and replace with cool fuel prior to cranking. mmmm

Vapour locking is more likely to be caused by pressure drop when the car is left standing rather than the pump not running before startup. On an ideal system when you turn the engine off the fuel rail holds its pressure for a long time (30+ minutes) so that the heat soak from the engine doesn't boil the fuel in the rail. As long as the rail remains pressurised for long enough so that the danger of heat soak passes it won't vapour lock.

Some ECU's like that in the Xantia V6 will actually run the fuel pump briefly (about half a second) two or three times over a half hour period to "top up" the fuel pressure to help prevent it dropping too quickly. (Even though the engine is off and the car is locked!) However if either the non return valve in the fuel pump has a slight leak, or the pressure regulator (in a return system) doesn't quite seat, fuel pressure can be lost in under a minute and that will almost certainly vapour lock even though it won't affect normal running.

From what I've seen the non return valve leaking back slightly in the in tank pump is probably the most common cause, which would mean either a new fuel pump or an extra non return valve fitted somewhere in line with the fuel filter. If you have a fuel pressure gauge and a schraeder valve on the fuel rail, try monitoring the fuel pressure after the engine is turned off - if it maintains most of the static pressure for at least 20-30 minutes its ok, if it disappears in a couple of minutes I would start there.