m20 dying issue...AFM?

[Lurker27]

Cliff notes version:

1. 325is, 1989, autotragic.

2. recent work: timing belt, spark plugs, rewire fuel pump (pretty new), replace fuel filter, seafoam in tank and intake, spliced C191 connector, new thermostat, cleaned ICV.

3. Issue: . First, the car is difficult to start. It turns over, but in the initial surge of revs, it may or may not "catch" and idle at a normal rate. If it gets to 1000 or so I can nurse it along with gas. I drove home toggling back and forth in neutral and with brake and gas to modulate my speed while feeding the car the whole time.

Any time I am driving and I have my foot off the gas, it may stall out (loses all throttle responsiveness and rpms slowly drop to zero, rather than all at once). It's almost like it misses a beat and then is unable to recover to a normal driving mode. Once it stalls out, it usually starts back up. If I attempt to start the car and it does not reach idle, quick subsequent attempts are much less likely to succeed (I have to wait 2-3 minutes before a restart is likely to succeed).

The car drives GREAT when it is not dying.

[Knockenwelle]

Check TPS? And also connectors for all critical sensors.

[StereoInstaller1]

FIRST motherfucking FIRST thing to check is the main relay.


Pop the cover off the 3 relay pack just in front of the left strut tower. Pull the front one...if you see ANY corrosion AT ALL, get a new one.

Did this stumble extrordinare just start, or has it gradually gotten worse?

[Lurker27]

Gradually gotten worse in that it seems to occur more frequently.

TPS and wire have been checked, are fine. I multi'd all the connectors, nothing seems to be amiss.

Motronic throws no codes (1444)

Hadn't thought it could be main relay, amazingly...I assume new ones are cheap enough to order on a whim?

[jlevie]

With the recent work, my first suspicion would be an intake leak.

[Lurker27]

Problem existed before the work, and idle is fine - wouldn't a vacuum leak cause increased idle speed.

[StereoInstaller1]

That damn main relay has caught me so many times now. I say if it is 10 years old, change it.

[Lurker27]

Will do, thanks a lot. I'm sourcing a known good AFM and will check out my vacuum connections also.

[Fidhle007]

I would suspect a vacuum leak or AFM as well, let us know what you find!

[Lurker27]

Here is the update:

I ordered a new main relay and AFM (well, known good used AFM!). I also picked up new injector seals.

I'm testing the CPS/ICV tomorrow. I tested the ICV and cleaned it this summer, the ohm reading was slightly high but it seemed to run just fine and the car does have a good idle, RPM wise.


Tonight, though, it fucked me over. I was dropping my girlfriend off at her house, and idled in the driveway for maybe 10 minutes, then decided I was wasting gas and turned it off, and when I went to restart it, it put the revs up to ~1000, then died as per usual. It also did this same symptom a week or so ago (lots of idling -> turn off -> hard starting) The usual symptom of "the more you wait since last start attempt, the easier it gets to start" persisted, but I was constrained for time and could only wait ~15 minutes (it ALLMMMMOOOOST made it to started!). Then she wanted to go to bed, so I walked the mile-and-change home (thank God for the girl-next-door).

In the middle of this I checked motronic and got a 1221 code - I believe this is lambda (air-fuel) error. I feel like (and this may be crazy), the car floods the cylinder or something while it idles, and so that messes up the fuel air, and it takes time to vaporize that stuff off to the point where it has a clean slate to start again.

That could jive with a bad spot/out of tune/dirty AFM, I think. I am hoping like crazy that this fixes my issues, after the stuff I've already checked and will be checking tomorrow, I'm out of ideas.

[Fidhle007]

Lambda code could indicate a faulty O2 sensor but this really sounds like a bad coolant temp sensor. On a cold start, the motor is being run as if it's cold but once it's warmed up, the ECU can't see that it has so re-starting becomes an issue until the motor has a chance to cool down to a point where it will run within the parameter provided the partially-blinded ECU.

[Lurker27]

CTS has been replaced, issue persists. Main relay replaced, issue persists. I'm convinced it's not CPS or Idle Control valve, both check out OK and I have spark during the failures, just no throttle response (not really a CPS symptom). It dies very...very....slowly, still running for maybe 10-15 seconds with no throttle response whatsoever.

And then starts right back up. It did it to me last night on the Commodore Barry Bridge in Philly, coasting down one side. (again, no pedal -> death). I'm beginning to suspect a maladjusted or otherwise faulty TPS, in addition to the AFM. I'll probably clean my injectors and Seafoam the car again just to be sure, but this is definitely an electronic gremlin. I know nothing about the evaporative emissions system - could this be in there?

Thanks....

[jeer]

Have you checked the connector under the intake manifold, they have been know to hold moisture and corrossion on the pins and cause major headaches on m20 owners.
then a bad dme/ecu would be my next step givin the measures you have already taken.

[wagonizzle]

I don't think anyone has suggested fuel pressure regulator...

[Lurker27]

The C191 conenctor under there has been spliced and replaced - The local BMW guy suspected that too, so you're in good company.

FPR is interesting, my delivery did look fine though and I'm not really sure that fits the symptoms. I'll check out the unit though.