I am about out of ideas, so I'm lofting this to the hive, hoping for a fresh perspective.
Car -- 1998 E36 M3. 150k mi. Stock.
Issue -- Intermittent cylinder 1 misfire. Once the CEL illuminates, it shuts off that injector and I have a dead miss. Restart clears the dead miss. Very predictable pattern.
Fault Condition -- Cold start, about a 3 miles into a drive cycle. Coolant temp >= 190, usually about 194. Occurs after a low load coasting down hill. After clearing it, it will not reoccur until another cold start. Only thing in the freeze frame data that is weird is a spike in short term fuel trim -- Was +28% on bank 1. LTFTs are 4-8% STFTs are otherwise +- 2-4%. On a restart, the STFT goes back to normal.
Troubleshooting done so far:
- New plugs.
- New O2s.
- Moved the coils around to see if it moved (no.)
- Compression test -- 172-170-170-168-165-165.
- Replaced the fuel injectors (thought #1 was leaking down during coasting and then over-rich when the injector fired back up.)
- New CCV, all vacuum hoses, intake plenum boots, intake and TB o-rings
- Cleaned MAF.
- Smoke tested to check for leaks in the intake system (after all the new stuff) -- tiny wisp of smoke from the ASC butterfly, but no major leaks, and that's upstream from the common plenum.
What haven't I thought of? Thanks!
You pretty much covered all the bases, if I was in your shoes I would probably swap the injector next. It's a piece of cake as long as you don't lose the clip ...
In reply to Slippery :
Thanks, all 6 injectors are freshly replaced from FCP Euro. No change.
The only other hint, maybe: when I bought the car a few years ago, I would get a code for the charcoal canister valve stuck shut. I tested the valve and it was working. Replaced it with a charcoal canister out of Pull-a-part and the code never came back. I also replaced the fuel/vapor separator tank and fuel tank pressure sensor from the donor car, which was really freakin clean. My only hunch is maybe some EVAP or purge process is causing a lean condition??? Grasping at straws.
Food for thought . . .
Parts tossed at problem so far:
- New plugs.
- New O2s.
- Replaced the fuel injectors (thought #1 was leaking down during coasting and then over-rich when the injector fired back up.)
- New CCV, all vacuum hoses, intake plenum boots, intake and TB o-rings
Troubleshooting done so far:
- Moved the coils around to see if it moved (no.)
- Compression test -- 172-170-170-168-165-165.
- Cleaned MAF.
- Smoke tested to check for leaks in the intake system (after all the new stuff) -- tiny wisp of smoke from the ASC butterfly, but no major leaks, and that's upstream from the common plenum.
Did you use a genuine BMW CCV?
Based on... the contents of my head, so... probably in left field.
Fuel rail layout on these? If they feed at the rear and return from the front (IIRC like my E30) could there be some obstruction at/near the #1 injector?
Unfortunately a clog there seems like it would probably interfere with the return and if anything force a trim the other direction...
Tried looking at #1 and any other spark plug for clues about conditions during combustion?
In reply to Jesse Ransom :
Correct, the rail is fed from the rear and the return comes after #1.
I had a 1999 M3 that went through what seemed like a lot of coil packs, maybe 2 or 3 in one year once, so I came in here thinking it was one of these. I guess I'm no help.
So...
The misfire happens on a cylinder:
- That has compression
- That has a coil that works elsewhere
- That has a new injector just like the rest
I think "air" generally can be assumed, and "compression good" mostly means "air is correct/reasonable."
So something is wrong with either fuel or spark.
- If spark is a problem, it's probably upstream of the coil (I feel like there's probably some room for vagueness in chasing a timeline on when coils were swapped relative to spark plugs; you'd think it would *eventually* move with a bad coil if the coil caused plug #1 to quasi-foul, but it does leave a little room for the problem to stay with the plug.)
- If the problem is fuel, it's upstream of the injector (or the new injectors included a dud that happened to wind up in the same spot. Once again just noting the highly improbable but conceivable). "Upstream" on fuel means both electrical and fuel supply. Connectors, grounds, actual driver... It'd be so weird to have a fuel rail issue that only impacted one injector...
And that's kinda where I'm at. I think we're into the realm of something reasonably arcane and weird, but somewhere there's a concrete reason. Now it's a matter of locating it.
I would put a bore scope in that cylinder and give it a really good inspection. I am thinking there is something wrong that is "fixing" its self when it gets warm.
Thanks all for the input. Will keep you posted.
Just for giggles, I pulled the plug to the EVAP purge valve and car ran great and no issues across several drive cycles today.
dean1484 said:
I would put a bore scope in that cylinder and give it a really good inspection. I am thinking there is something wrong that is "fixing" its self when it gets warm.
I have a bore scope and will do so and post some pics next time it happens.
Jesse Ransom said:
So...
The misfire happens on a cylinder:
- That has compression
- That has a coil that works elsewhere
- That has a new injector just like the rest
I think "air" generally can be assumed, and "compression good" mostly means "air is correct/reasonable."
So something is wrong with either fuel or spark.
- If spark is a problem, it's probably upstream of the coil (I feel like there's probably some room for vagueness in chasing a timeline on when coils were swapped relative to spark plugs; you'd think it would *eventually* move with a bad coil if the coil caused plug #1 to quasi-foul, but it does leave a little room for the problem to stay with the plug.)
- If the problem is fuel, it's upstream of the injector (or the new injectors included a dud that happened to wind up in the same spot. Once again just noting the highly improbable but conceivable). "Upstream" on fuel means both electrical and fuel supply. Connectors, grounds, actual driver... It'd be so weird to have a fuel rail issue that only impacted one injector...
And that's kinda where I'm at. I think we're into the realm of something reasonably arcane and weird, but somewhere there's a concrete reason. Now it's a matter of locating it.
Arcane sums it up! Perplexing and it freaking runs great other than when it does this BS. :)
I unplugged the purge valve from the wire harness on a hunch and haven't had any other misfires since.
Does the ECU lean out the mixture while the purge valve is open?
I suspect the evap hose might be damaged south of the purge valve, creating a lean condition when the valve is open. Maybe this would yield the high short term fuel trim, trying to compensate for richer fuel mixture when purging?
I'll segment out the evap system and check for leaks. I know OBD2 looks at the tank pressure and runs a test to verify if it's leaking. Not sure if the purge vent hose past the tank is tested? I would assume the DME would look for the mixture to richen when the valve was open to indicate it was working?
Might put an LED on the purge valve connector and plug it back in to see if there is correlation between the valve opening and the misfire.
Sounds like you might be on the right track but if it isn't it, look at your valve train. That can cause issues specific to a cylinder.
