Bimmerfest - BMW Forums - View Single Post - How do diagnose a sudden and disturbingly unsettling cold-engine misfire (vanos?)
View Single Post
Old 02-24-2010, 04:09 PM
bluebee's Avatar
bluebee bluebee is offline
Seek to understand,^Value
Location: San Jose, California
Join Date: Mar 2008
Posts: 25,236
Mein Auto: 02 BMW 525i M54 auto 130K
Originally Posted by bmw_n00b13 View Post
Spark plugs, MAF, CPS, ... coils ...The VANOS issue can't be helping, but it almost certainly isn't the root cause.
I appreciate EVERYONE's help! That's why I'm here! I love you guys!

However, I'm probably one of the slowest and least-informed people here (about how the BMW works) ... so please bear with me on this quest for truth.

My approach is that every part is presumed innocent until proven faulty. I generally test to PROVE that a component is faulty so that I may sentence that component to the trash bin with a clear conscience (bear in mind this methodical approach has drawbacks ... it took me a year, for example, to resolve the ABS trifecta issue, mainly because I finally had to simply accept the sad fact we'd never come up with a definitive test that "proved" the culprit). Point is, I feel it's more helpful if we can believably UNDERSTAND what we're up against, and then reliably outline a test sequence which fingers the culprit and prevents fuel-cut-off recidivism.

True to the original request, let's concentrate on defining the debugging sequence in this thread for anyone in a BMW who comes up with the OBD code P1349: Misfire Cylinder 4 With Fuel Cut-off (specific to BMW).

First, it behooves us to understand these things:
1. What are the symptoms?
2. What is the error telling us (and not telling us).
3. What is the debugging and repair sequence.

What are the symptoms:
- Rough idle upon cold-engine start <--- may be temperature & humidity related; may be related to car sitting for two or more days prior
- Rough running lasts only for the duration of the "ignition cycle" <--- This is the most amazing clue if we can take advantage of it for debugging!
- The prior and next starts are generally perfectly smooth <--- This is another most amazing clue if we can Colombo it properly!
- Service Engine Soon is lit with the resulting code of P1349
- Some say it happens after changing fuels (I certainly changed mine as I was testing smog results under different octane-rated fuels)

What is the OBD II code telling us (and not telling us):
- A single cylinder, #4, misfired
- Fuel was cut off

Note: I found only one reference that explains the difference between a "misfire" and a "misfire with concomitant fuel cutoff".
"...with cylinder cut-off" means the DME has ceased to fire the fuel injector, because the fuel is not being ignited.

My assumption is that the #4 misfire was so bad that the engine electronics noticed the stumbling and they (somehow, magically?) cut off the fuel to the cylinder. If this is the case, it might explain why the rough idle only lasts the duration of the trip. The assumption is that whatever caused the cutoff "reset" itself upon an ignition cycle.

If that's the case, then a possible miscreant might be whatever that circuit itself is that determines it's time to shut off the fuel to cylinder #4.

Even so, thinking about this avenue brings up more questions than it answers. For example, how would you cut off the fuel to just a single cylinder anyway? Don't all the injectors fire into the intake manifold simultaneously? Do they just cut off the electricity to the "closest" injector to cylinder #4?

Similarly, there is nothing wrong with replacing the air filter (Lord knows I have no idea when mine was last replaced, for example), however, I doubt a blocked air filter will affect ONLY cylinder #4, and even then, that it would affect that cylinder so badly that it would cut off the fuel supply to cylinder #4; and even after all that, that it would totally go away on the next start after waiting a minute.

Along that vein, I'm sure VANOS affects valve timing, but, if VANOS affects ALL valves equally, then it seems not likely (to me anyway) that VANOS would badly affect only the valves (do we have 2 or 4?) associated with just the one cylinder #4. I admit I don't know how VANOS works. Do you know if VANOS seals have the potential of affecting only one set of valves to a single cylinder?

It seems to me the guilty party is something unique to cylinder #4. So, I might first ask of every suggested culprit ... Which ones are unique to cylinder #4?
a) coil (definitely!)
b) plug and/or plug boots (definitely!)
c) air filter (not likely)
d) cam position sensor (??? I have no idea if cylinder #4 has its own sensor)
e) vanos (??? I have no idea so that's an open question)
f) maf (??? I don't even know what MAF stands for yet)
g) cps (??? I guess I need to learn a new vocabulary)
h) vacuum (hmmm... is vacuum specific to cylinder #4?)
i) oxygen sensors (2 precat, 2 postcat, hmmm... is an o2 sensor specific to cylinder #4?)
j) coolant leaking (maybe ... coolant could leak into just cylinder #4)
k) contaminated fuel (not likely to affect only cylinder #4)
l) engine electronics (this could potentially be related to only cylinder #4 but I don't know how yet)

What is the OBD II code NOT telling us:
- I see no indication that more than one cylinder is involved.
- I see no indication the fuel cut off first; it appears the misfire happened first (if I read the code correctly and discern between a simple "cylinder misfire" and a "cylinder misfire with concomitant fuel cutoff").
- It seems the misfire was so severe, that the fuel had to be cut off (but how?)

What is the debugging sequence:
Note: I don't really think we can come up with a believable debugging sequence until we can state we understand what the code is telling us and what controls the firing and fuel on cylinder #4. Bear in mind that the E46 has this random cold-start misfire problem in spades.

By reading scores of related misfire posts, and combining all the proposed "solutions" into a reasonable sequence, so far, the preliminary debugging sequence (which, I admit, is more of a guessing sequence than a debugging flowchart) is:

0. When you notice a BMW cold-start stumble ... especially after sitting for a couple of days in damp cold air ...
1. Shut the ignition off & wait 30 seconds to restart the engine
2. The stumble may go away but look for the SES light on the cluster
3. If you see an SES, check for a P1349 code and reset the MIL
4. If you have a P1349 code, swap ignition coils between #3 & #4
5. If the intermittent start stumble moves to cylinder #3, replace the coilpoolman coil recall information here)
6. If the start stumble returns on cylinder #4, swap or replace the plug
7. If the stumble returns on cylinder #4, swap or replace the plug boots
-----< after this point, I think we're moving away from the specific cylinder >-----
8. If it returns on cylinder #4, clean the MAF (hot film air-mass meter) with CRC & replace the air filter
9. If it returns, replace the "cam position sensor" (is there one specific to cylinder #4?)
10. If it returns, replace the Cam Position Sensor CPS (is this specific to cylinder #4?)
11. If it returns, replace the post-cat oxygen sensor (the pre-cat 02 sensor is not implicated)
12. If it returns, replace the VANOS seals (can variable valve timing affect only a single cylinder?)
-----< after this point, we're really shooting in the dark >-----
13. If it returns, replace the CCV (aka CVV oil separator valve)
14. If it returns, replace the fuel filter (one guy suggested replacing the fuel pump)
15. If it returns, add a bottle of Seafoam or Techron concentrate to the fuel
16. If it returns, clean the ICV (idle control valve) with carb cleaner
-----< people really suggested all these items in the respective threads
17. If it returns, replace the hose from the MAF to the engine
18. If it returns, replace any cracked "T-connection after the MAF sensor"
19. If it returns, replace the "valve cover gaskets (VCG) into the spark wells"
20. If it returns, clean or replace all the fuel injectors
21. If it returns, replace the "coolant temperature sensor"
22. If it returns, replace the alternator to raise the voltage (yes, this has been suggested)
-----< ok, these were listed as solutions but they aren't feasible >-----
23. If it returns, move to Georgia (from Michigan
24. If it returns, drive the car every day (don't let it sit for two days)
25. If it returns, check the "intake boot" for cracks and replace if necessary
26. If it returns, replace the "throttle position sensor" (TPS)
Note: Don't laugh; I've read dozens of threads and each of these has been posited as the "solution" to this intermittent problem!

In summary, the goal (of this thread) is together, for us to come up with a believable debugging sequence for all to benefit.

The first part is to understand what could affect only cylinder #4 and the fuel cutoff to cylinder #4. Intuitively, I suspect it's that circuit itself that is faulty (i.e., the circuit that shuts off the fuel to cylinder #4); but having said that, I have no idea what the circuit is that shuts off the fuel to cylinder #4. Do you?

Please help all by contributing answers to some of the questions in this thread (and bringing up new relevant questions to answer).

Note: Multiple BMW models seem affected by this intermittent cold-start stumbling that goes away with the next ignition start; so does Toyota, e.g., NHTS May 2003 Bulletins 00903 (Toyota Echo) & 00703 (2002 Toyota Camry) variable valve timing (VVTI) malfunction.

Unfortunate acronyms used in this post (most of which I had to look up):
- OBD (On Board Diagnostics)
- DME (Digitial Motor Electronics, i.e., the engine computer)
- VCG (Valve Cover Gasket)
- ICV (Idle Control Valve)
- CCV or CVV (Oil Separator Valve)
- VANOS (variable valve timing)
- MAF (mass air meter)
- CPS (Cam Position Sensor)
- CAT (catalytic converter)
- TPS (throttle position sensor)
- MIL (malfunction indicator light, i.e., the SES)
- SES (service engine soon, aka CEL check engine light)
- VVTI (Toyota variable valve timing)
- NHTS (National Highway & Traffic Safety Administration)

Last edited by bluebee; 02-25-2010 at 07:47 PM. Reason: I will update with new information for as long as it will let me
Reply With Quote