1991 325i SE - Engine cutting out/not starting - advice?

[Nightshift]

Car developed a new fault last week.

Took the car out last weekend, came to a stop at some lights, engine died, oil light came on, wouldn't restart - not even turn over. Just sat there dead as a dodo. Pushed it to the side of the road, realised I didn't have my phone (nice) - so pegged home to call out recovery. 15 minutes later went back to the car, tried it for luck, and it started (after some hesitation). Drove home and then went round the block a few times to see if it happened again - no problem. Idling happily, all smiles.

A few days later, as it had been sitting there covered in snow, thought I should give it another run - being cautious, just drove it round the neighbourhood a bit; engine cut out/died three times: each time when I was dropping into 2nd and decelerating to turn into a junction or similar. Same symptoms - oil light (and other ignition lights) come on, engine dead. Difference being that this time it would start again every time with no problem.

Went out again this morning. Fired up no problem, idling OK. Fine, I thought, I'll take it out. Turned the ignition off, popped in to get my coat, came back, wouldn't start. All ignitions lights, check panel lights OK (though again the oil light stays on), but no life - won't turn over, just makes one click, then clicks repeatedly if you keep the key turned to the start position.

Another thing I noticed mid-week was how low the idle speed was considering the freezing temperature - around 600 rpm. Was under the impression it should be higher than that (around 7-800) normally and around 1000 when cold? Might be unrelated but maybe a glitch with the ECU?

Other info: a few weeks back the car died on the way back from sussex to London - backfiring, loss of power, dying then restarting. Eventually it would just turn over but not start. After getting towed home and getting a diagnostic guy out, the ECU returned no fault codes apart from suggesting a fault with the ECU itself (which he thought was a false reading) - he reckoned the fuel pump was the problem, which I replaced and which seemed to due the job.

I'd suspect the starter motor if it wasn't for the cutting out; I'm suspecting the Crank Position Sensor, or perhaps the ECU itself. Any advice from anyone who's experienced similar issues very welcome!!

Work done in the last 12 months (not everything - just what might be relevant)
- Inspection 2
- new coolant temperature sensor
- injectors cleaned and refurbed
- air intake boot, breather hose, vaccuum hose etc, replaced
- ICV cleaned
- new fuel pressure regulator
- new fuel pump

Things I've checked so far
- grounds; front strut drivers side, glove box, under steering column all seem fine
- C191 connector and the connector clipped to the firewall - all good, no corrosion
- connections for main relay and fuel pump relay - look OK, though haven't tested them
- battery connections

Thanks in advance for any pointers anyone can offer.

[Nightshift]

Anyone? Looking like a tow to the garage...

[Grrrmachine]

Almost word for word, this is what happened to my old engine (M20B20). Fault was the blue temp sensor; specifically the wiring had come loose at the plug.

Considering how the temperature's dropped in the last month, it wouldn't surprise me if your temp sensor is causing the engine to overfuel, causing it to flood after a few minutes which is why it conks out when you dip the clutch, and then won't restart for 15 minutes.

The PROPER technique is to get access to the ECU plug and test the resistances across the whole loom. But I'm 99% it's the temp sensor and they're only a few quid if you want to get it swapped over.

More info:

http://www.e30zone.net/e30zonewiki/inde ... emp_Sensor

[Nightshift]

Thanks Grrrmachine.

Suspected the blue temp sensor because of the idling characteristics. It was replaced spring last year, so if there's a fault it's in the wiring, one of the connectors, or in the way the ECU is reading/receiving the info. C191 connector (under the inlet manifold) looks fine so it looks like I'll need to check the resistances at the ECU.

Surely this wouldn't affect starting though? Currently the engine doesn't even turn over, the solenoid just clicks repeatedly. I'd suspect flat battery but lights etc. all working fine (no multimeter available for testing) and the battery being flat wouldn't have caused engine cut out.

I think the issues are related (cut out/not starting) as every time it cut out the oil panel light came on, and that is on all the time now, even though oil level is fine. Perhaps the oil pan to body ground: http://www.rtsauto.com/ground-cable-rep ... ventative/

As usual, cheers for any pointers.

[bigwinn]

I had this but it was the CPS- couldnt see a fault with it until it was removed, then you could see it was worn due to it being refitted incorrectly, presumably on a belt change. The rubbing on a pulley had worn down to the wires which caused a short. This only happened when the engine was hot, it died as you approached junctions then could not restart...

Its worth the short job of removing just to check

[Nightshift]

Bigwinn - cheers.

Definitely on my list to check - I suspected the CPS when I had the misfiring/cut out issue that the diagnostic guy said was fuel pump. If there was a problem with it wouldn't the ECU log a fault though?

[Grrrmachine]

CPS would only have an effect if the engine was already spinning. If you can't even crank the engine then the CPS won't be the issue.

I'd be looking for the basics, such as corroded earth straps and loose connections around the engine bay. Tighten up all nuts on the alternator and starter motor too.

[Nightshift]

Damn I wish I had a nice dry garage to work in... rather than kerbside.