Another starting problem, but a little bit different this time.

[Demelich]

I have developed a starting problem in my 91 318is.

Cold starts, for example, its about 30F this morning, it started right up, although it is a bit rough for a few seconds.  The idle surged between 600 and 1000rpms for about 5 seconds before settling down to a nice idle.

However, yesterday, it was about 60F and it would just crank.  It tries to start a little bit, but does not catch, but then I push the gas pedal down a bit, it fires right up. Then, if I kill the engine and start it again after waiting a few seconds, it fires right up.  If I wait 10 or 15 minutes, it does not start, and just cranks until I push the accelerator.

Note that both of the starts listed above were after the car sat all night, so the only change was ambient temperature.

I do smell fuel when I have the starting issue.

-I tested the resistance on both the cam and crank sensors and they were good.
-My ICV appears to be fine, it hums and vibrates. Also, I have two of them and recently swapped them as a test with no change. Cleaned it up with carb cleaner when I swapped them.
-Airflow sensor swings freely, and the sensor looks fine. I also have two of these, I ran resistance tests on both of them and they seem OK
-Before winter, I redid the mess of hose underneath the intake manifold. I also replaced the temp sensor.  I have also sprayed carb cleaner on the various hosing looking for a leak, with no luck
-I am trying to replace my O2 sensor, but that sucker is stuck.  However, I do not believe that has anything to do with starting the car.  Note: I did unplug the O2 sensor and drive around the block.  The car was warm and I noticed no difference in performance nor starting.
-Fuel pump is only 3 years old, and turns on when cranking.
-New spark plugs and ignition coils. I did not replace the wires, but they look to be in decent condition.

Possible issues
-I tried to perform diagnostics on my TPS, but was unable to get any good readings for any of the tests. I believe the TPS does help control the ICV. I wonder if a failing TPS could be why my car starts as soon as I give the car some throttle, if perhaps it is not reporting a closed throttle condition properly.
-Fuel regulator. Could this be the cause? On a cold start, extra fuel is injected, on a start in warm weather, it is not. Then, when I start the car immediately afterward, it works fine, I still have fuel pressure. So my thought is that perhaps, after sitting a while, I am losing fuel pressure, but the cold start system takes kind of bypasses the issue.


Battery is good, and once started, the car runs and idles great.  I would just like some opinions before dropping cash on a new TPS, fuel regulator and possibly ICV.

[DesktopDave]

Sounds like you've been around that engine pretty well.  I have a similar issue with hard starts below freezing and rough running when it's cold.  You're correct about the O2 sensor being ignored at start.

I'd suspect the TPS.  Did you do the stomp test for stored codes?  It'll confirm that one.  I'd also test the ambient air sensor in the AFM, that also helps to determine cold start enrichment.

I also had trouble with a resistor pack installed on the AFM.  It looks like a wiring extension, but it actually changes the AFM output to the DME.

I've also heard of a small check valve that some BMW owners put in the pump supply line at the pump.  It's an OEM part for some M20 cars, might be a good idea for you if none of the above sensors are wrong.

[Demelich]

As for stomp codes, I get a code for coil pack 1.  Ive replaced the coil pack yet still get the code.  I suspect it has to do with my EAT chip, as I have read about that issue before.  No code for TPS, but then, I believe I have also read that it can still check out OK as far as the DME is concerned, even if its going bad.

I did resistance tests on the AFM temp sensor and it came back good.  Where is this AFM resistor pack?  Are you talking about the black box connected to the AFM where the wiring connects?

I have read about the check valve as well, but there should be a cause for it.  Either the check valve on the fuel pump has bad, in which case I would go with the check valve.  If the fuel pressure regulator is the issue, I would rather just replace it.

Ive got to order some stuff from Pelican anyway, so I may as well pick up a new TPS.

[DesktopDave]

The AFM pack was inserted in the wiring harness.  It looks like a wiring extension - techs just pulled the AFM connector and put this little harness in the middle.  It's not very common.  Here's a pic:

[Demelich]

Actually, I do have one of those on my car.  Part number on the rubber boot at the AFM end is 17345163, but I am not finding it on RealOEM or Pelican Parts.

Whats the purpose of this resistor pack?  Some sort of quick and easy fix for some issue?

[Demelich]

I pulled the resistor pack/cable, and it ran like it did before, with the same starting issue.  So that eliminates that I think.

I do have a new TPS on the way, but the more I think about it, the more it seems like a fuel pressure issue, just because after turning the car off, the engine will fire up right away immediately afterward.

I also picked up a new fuel filter, because mine is way over the recommended 6000 mile service interval.

[DesktopDave]

Mine too, I'm planning on dropping the tank to replace all the hoses & filter this spring...  I'm hoping that it's just a clogged filter instead of the FPR.

I was told that the harness resistor pack was meant to cure some driveability issues.  A German page attributes it to cold-start jerkiness and driveability issues in the M40 & M42.  BMW really took some time to make this engine as smooth as a 4cyl could be...seems like we're undoing all the patches to make it into a more honest machine!  Hydraulic motor mounts, dual-mass FW, extensive intake pre-heating...I'm surprised they didn't add balance shafts.

Found this page too.  Granted, it's the e30 M3 service manual, but it's still pretty useful.  The adapter is part of a service campaign mid-'91 on this page.  Note the last six items, all are well-known problems with our car.

[Demelich]

As a possible resolution, I think that I have determined that my problem is in fact a fuel pressure issue.

There are two hard fuel lines going up underneath the intake.  I pushed the fuel line closest to the firewall up and heard an air leak as the system depressurized.  So I either have a cracked fuel hose, or a bad seal up in that mess.

I will not know until I get in there and replace the hoses, but it seems like a reasonable solution.

[DesktopDave]

There are two really short fuel hoses under the intake manifold connecting the hard lines to the fuel rail.  Maybe that's the leak?

[Demelich]

Yep, thats what I was thinking.  I replaced the whole mess under the intake manifold, but I did not even touch the fuel lines.

Guess that was a mistake, since Ill have to pull at least the upper intake manifold off again.