Intermittent Bank 2 Misfire and Rough Idle

[Contra]

SOLVED. Summary at the bottom of this post so you don't have to read the whole thread if you don't want to:

This is kind of a long post, but I've been having a weird misfire as of late and wanted to see if anyone has any idea whats going on with my car.

Car:
2008 BMW M3 6MT
94k miles
RBs done at 77k
TAs done at 80k
Spark plugs done at 91k
RKP Street Xpipe that definitely leaks a bit

The first time it happened I was letting my car warm up after work, sittiing on my phone while it idled for a little after starting it. I noticed the car starts to shake and the idle becomes rough. I get limp mode and a SES light. I immediately turn the car off and scan for codes and get the following:

2729 DME: Oxygen Sensor, coast-mode diagnosis before cat, bank 2.
27B1 DME: Secondary Air system flow rate, bank 2
2B39 DME: Misfire with cutoff, cyl 5.
2B3A DME: Misfire with cutoff, cyl 6.
2B3B DME: Misfire with cutoff, cyl 7.
2B3C DME: Misfire with cutoff, cyl 8.
2B41 DME: Misfire with cutout, several cylinders

I clear the codes and restart the car, because the O2 sensor codes have also been randomly coming and going and I just replaced the bank 2 precat o2 sensor 3 weeks prior so I assumed those were stored codes. After I restart the car it runs totally normal. I drive it home with no issue.

I go 12 days with no issues until i get the same rough idle, shaking, and limp mode at a red light, pretty soon after leaving work (car was not fully warm). As soon as I get limp mode I cut the ignition off, wait 15 seconds, and restart the car... and its totally normal again. I scan the car for stored codes after I reach my destination and get the following:

275E DME: Oxygen sensor limit position error bank 2
2B46 DME: Misfire impairing exhaust emission cyl 5.
2B47 DME: Misfire impairing exhaust emission cyl 6.
2B48 DME: Misfire impairing exhaust emission cyl 7.
2B49 DME: Misfire impairing exhaust emission cyl 8.
2B4E DME: Misfire impairing exhaust emission, several cylinders.

I end up having to go to a wedding the following day, so the car sits for 2 days before I work on it. I get it up on the lift and check all the o2 sensors are plugged in and none of the wires are damaged. They look fine, so I try starting the car. It doesnt start well and immediately starts misfiring. The codes above remain the same.

I swap the bank 1 and 2 O2 sensors to see if the misfire changes sides. I restart the car and its definitely still misfiring and idling poorly. So I clear codes to see if the misfire changed sides and the car wont give me another SES. It idles poorly but the light won't come on so the codes weren't posting.

I call it a night because at this point it was late and the car is loud and I didn't want to bother neighbors.

The next day I start the car and its totally normal. Can't get it to misfire. I drive it for 20mins, do a few pulls to redline, let it idle for 5 mins and nothing. No codes, starts fine, idles normal.

Anyone have any idea whats going on?

RESOLUTION:

I ended up replacing:

-All 4 O2 sensors (3 times)
-Secondary air pump, valves, hoses
-Coolant temp sensor
-Valve covers, valve cover gasket, spark plug tubes
-Spark plugs
-Got a unmodified OEM X Pipe
-All exhaust gasksets and hardware

And STILL had the same problems.

Ended up stripping the O2 sensor plug harness to check for wire damage and re-sheathing the wires. Problem Solved. Weird.

[amrazM]

these cars are fairly sensitive to a leak from the header to middle exhaust, if you know theres a problem there you should address that first. the o2 sensor has reached the limit of its trim adjustment.

[Redd]

One bank misfires usually point to exhaust. If you're tried swapping the O2 sensors then it's likely an exhaust leak. Check your exhaust ring gasket between the front cat and the headers.

[DrFerry]

Originally Posted by Contra

This is kind of a long post, but I've been having a weird misfire as of late and wanted to see if anyone has any idea whats going on with my car.

Car:
2008 BMW M3 6MT
94k miles
RBs done at 77k
TAs done at 80k
Spark plugs done at 91k
RKP Street Xpipe that definitely leaks a bit

The first time it happened I was letting my car warm up after work, sittiing on my phone while it idled for a little after starting it. I noticed the car starts to shake and the idle becomes rough. I get limp mode and a SES light. I immediately turn the car off and scan for codes and get the following:

2729 DME: Oxygen Sensor, coast-mode diagnosis before cat, bank 2.
27B1 DME: Secondary Air system flow rate, bank 2
2B39 DME: Misfire with cutoff, cyl 5.
2B3A DME: Misfire with cutoff, cyl 6.
2B3B DME: Misfire with cutoff, cyl 7.
2B3C DME: Misfire with cutoff, cyl 8.
2B41 DME: Misfire with cutout, several cylinders

I clear the codes and restart the car, because the O2 sensor codes have also been randomly coming and going and I just replaced the bank 2 precat o2 sensor 3 weeks prior so I assumed those were stored codes. After I restart the car it runs totally normal. I drive it home with no issue.

I go 12 days with no issues until i get the same rough idle, shaking, and limp mode at a red light, pretty soon after leaving work (car was not fully warm). As soon as I get limp mode I cut the ignition off, wait 15 seconds, and restart the car... and its totally normal again. I scan the car for stored codes after I reach my destination and get the following:

275E DME: Oxygen sensor limit position error bank 2
2B46 DME: Misfire impairing exhaust emission cyl 5.
2B47 DME: Misfire impairing exhaust emission cyl 6.
2B48 DME: Misfire impairing exhaust emission cyl 7.
2B49 DME: Misfire impairing exhaust emission cyl 8.
2B4E DME: Misfire impairing exhaust emission, several cylinders.

I end up having to go to a wedding the following day, so the car sits for 2 days before I work on it. I get it up on the lift and check all the o2 sensors are plugged in and none of the wires are damaged. They look fine, so I try starting the car. It doesnt start well and immediately starts misfiring. The codes above remain the same.

I swap the bank 1 and 2 O2 sensors to see if the misfire changes sides. I restart the car and its definitely still misfiring and idling poorly. So I clear codes to see if the misfire changed sides and the car wont give me another SES. It idles poorly but the light won't come on so the codes weren't posting.

I call it a night because at this point it was late and the car is loud and I didn't want to bother neighbors.

The next day I start the car and its totally normal. Can't get it to misfire. I drive it for 20mins, do a few pulls to redline, let it idle for 5 mins and nothing. No codes, starts fine, idles normal.

Anyone have any idea whats going on?

Your fuel tank breather valve has very likely failed and is the culprit. I advise replacing it. When it goes bad most all (and perhaps more) of the symptoms you listed above are present. Search for threads on the topic and you'll find similar stories. Initially, the "close gas cap" dash warning appears, then "Service Engine Soon", then "check engine" light and "Increased Emissions" warning. The car will usually require a long crank after filling up with gas. Moreover, when codes are pulled you will see misfire listed for each cylinder and sometimes O2 sensor codes since they will log the lean mixture from long crank misfires. It's a cheap part and a relatively easy DIY. When (if ever) was your fuel tank breather valved replaced? They have about a 60K mile average 'half-life' in my experience. Hope this helps and fixes the mystery for you..

[Contra]

Originally Posted by amrazM

these cars are fairly sensitive to a leak from the header to middle exhaust, if you know theres a problem there you should address that first. the o2 sensor has reached the limit of its trim adjustment.

I think my exhaust leaks are from the cat-to-RKP X-pipe connection. It’s a slip fit, and I think the shop I had do the install cut the pipe after the cat slightly too short so no matter how tight I make the clamp it leaks.

Pics attached.

I’ve been trying to find an OEM X-pipe near me to remedy this, but I’m honestly leaning towards just having the X pipe welded to the cats and calling it a day. Was kind of hopeful I could go back to OEM then sell my cats and the RKP pipes to recoup some money.

[Contra]

Originally Posted by Redd

One bank misfires usually point to exhaust. If you're tried swapping the O2 sensors then it's likely an exhaust leak. Check your exhaust ring gasket between the front cat and the headers.

I bought and installed 2 new OEM gaskets 12k miles ago. I'd be a little surprised if they failed already, but I'll check them out the next time I get under the car.

[Contra]

Originally Posted by DrFerry

Your fuel tank breather valve has very likely failed and is the culprit. I advise replacing it. When it goes bad most all (and perhaps more) of the symptoms you listed above are present. Search for threads on the topic and you'll find similar stories. Initially, the "close gas cap" dash warning appears, then "Service Engine Soon", then "check engine" light and "Increased Emissions" warning. The car will usually require a long crank after filling up with gas. Moreover, when codes are pulled you will see misfire listed for each cylinder and sometimes O2 sensor codes since they will log the lean mixture from long crank misfires. It's a cheap part and a relatively easy DIY. When (if ever) was your fuel tank breather valved replaced? They have about a 60K mile average 'half-life' in my experience. Hope this helps and fixes the mystery for you..

I actually have a new OEM one ready to go, but I bought the OEM hose clamp and I don't have the correct pliers to install that clamp. Might run to home depot and get a normal hex/phillips head hose clamp and knock this out.

I have had the fluctuating idle after filling up for a little while, but its never been severe. Just a slightly surging idle. No close gas cap warning.

Also I'm only getting bank 2 misfires. I figured if this issue stemmed from the breather valve then it would be both banks misfiring, right?

[Contra]

3 week update: Car still hasn't misfired since. Still haven't changed anything other than switching the bank 1 and bank 2 O2 sensors around. Really strange.

I've been putting off changing the fuel breather valve just to see if the missfire comes back, but it hasn't. Car needs an oil change this weekend, so I'll probably install it then.

[aggieshan]

I had misfire on bank 2 last year. I had similar error codes. My misfires were actually due to engine oil overfill which messed up crank case pressure. So just in case you have added some oil or changed oil, you could check/estimate the oil level in your car.

[Contra]

Originally Posted by aggieshan

I had misfire on bank 2 last year. I had similar error codes. My misfires were actually due to engine oil overfill which messed up crank case pressure. So just in case you have added some oil or changed oil, you could check/estimate the oil level in your car.

I did have to add a little oil a week or so before the first misfire, but I didn't get an overfill warning. I added ~1/2qt when the display said "+1qt". Reset the oil level and it was at the first tick above the middle line.

Doesn't the car tell you its overfilled if you add too much?

Also, the oil level is already back down to the middle mark so I couldn't have put that much more than the middle. I normally have to add ~1 qt per 7,000 miles or so when I change the oil.

[aggieshan]

Originally Posted by Contra

I did have to add a little oil a week or so before the first misfire, but I didn't get an overfill warning. I added ~1/2qt when the display said "+1qt". Reset the oil level and it was at the first tick above the middle line.

Doesn't the car tell you its overfilled if you add too much?

Also, the oil level is already back down to the middle mark so I couldn't have put that much more than the middle. I normally have to add ~1 qt per 7,000 miles or so when I change the oil.

In my case, the oil level indicator was telling me 3/4 low after my track day, but the car actually had OK amount of oil. So i added 3/4 qts oil in. My car showed that the oil level is full but doesn’t say it is overfilled. So the next day or after another short drive, my car started misfire. I guess my case is not very common, but oil level uncertainty on these cars are quite common. There is a delay between oil added and engine symptom responses.

[Contra]

This weekend I changed the oil and tackled the fuel tank breather valve. Got a new OEM one installed and noticed this hose clamp for the idle control unit piping had broken (circled in red).

This could be the culprit, but also I checked my pics from 2020 when I did my throttle actuators and it was broken even in those photos.

Might cut this one and replace with a normal hose clamp so I can slip it on without having to removed the whole tubing “T” section.

[RocketyMan]

Thanks for updating! Let us know if putting a clamp on it solves your problem.

[Contra]

Originally Posted by RocketyMan

Thanks for updating! Let us know if putting a clamp on it solves your problem.

Honestly the problem hasn't returned since swapping the O2 sensors from bank 1 to bank 2... which makes no sense to me.

Not sure when I'll get around to replacing the clamp. It'll become a priority if the misfires return. I'll definitely keep this thread updated if something changes.

[RocketyMan]

Originally Posted by Contra

Honestly the problem hasn't returned since swapping the O2 sensors from bank 1 to bank 2... which makes no sense to me.

Not sure when I'll get around to replacing the clamp. It'll become a priority if the misfires return. I'll definitely keep this thread updated if something changes.

The WB O2 sensors have 5 wires going to the sensor in the exhaust, and in my experience they are very sensitive. I've had CEL before complaining about the O2 and I saw that the wiring was all twisted (like someone took it off before). I untwisted the wiring and it *SEEMS* to be working now.

You could've disturbed the wiring enough by switching sides that now it's functioning, in addition to perhaps getting a better clamp on the exhaust header there.

[Contra]

Update: The misfire came back......

Since my previous post, I found an OEM X-pipe off a 26k mile car for sale locally, bought it, and installed all new OEM gaskets, new header connection bolts, and exhaust supports. I did this to solve my exhaust leaks and hopefully stop killing O2 sensors so frequently.

A month or so after the OEM x pipe install, the misfire returned with the exact same symptoms and codes one morning while warming the car up. Turned the car off, cleared codes, and restarted it. Still misfiring.

So I had it towed to a shop to investigate since I was going to be out of town for work for 2 weeks the following day. The shop was less than helpful and didn't really seem to know the cause either. I had them replace the busted clamp on the idle control valve piping (photo in my previous post) and they said the misfire went away.

Picked the car up when I returned from my work trip and it ran great for about a week. Then the check engine light came on again during a cold start one morning, and I felt the car start to shudder and hunt for idle. I gave it a little bit of gas and it smoothed out. Scanned for codes and got 2729 and 27B1 again.

Ordered a new O2 sensor (even though the one in the car is less than 5k miles old) and will be installing that when it arrives.

[Redd]

Remove your plenum and air intake. Drive around and see if it still misfires. Do this after replacing your O2 as they could have been damaged by frequent misfiring.

[Contra]

Originally Posted by Redd

Remove your plenum and air intake. Drive around and see if it still misfires. Do this after replacing your O2 as they could have been damaged by frequent misfiring.

Like drive around with the throttle bodies exposed? Would that be just to rule out issues with the plenum? I also got 8 new throttle body couplers (the rubber things that hold the plenum onto the throttle body) while the shop had the car since mine were pretty tired and sad looking.

[Redd]

Originally Posted by Contra

Like drive around with the throttle bodies exposed? Would that be just to rule out issues with the plenum? I also got 8 new throttle body couplers (the rubber things that hold the plenum onto the throttle body) while the shop had the car since mine were pretty tired and sad looking.

Yes with the throttle bodies exposed. This helps to eliminate the entire intake system as potential problem. If the misfire goes away you likely have an air leak somewhere in the intake system.

[RocketyMan]

Originally Posted by Redd

Yes with the throttle bodies exposed. This helps to eliminate the entire intake system as potential problem. If the misfire goes away you likely have an air leak somewhere in the intake system.

I would not do this. This engine has ITBs... individual throttle bodies. Which technically means each throttle body has its own vacuum, hence needing the vacuum pump for the brake booster. There is an idle control valve that comunizes the vacuum on both banks. Vacuum doesn't start till AFTER the throttle plate on each ITB.

Taking the plenum off has little to no effect on vacuum. This pressure would technically be atmospheric in the plenum.

If we're concerned with misfire, a good way to verify this is to look at the spark duration and combustion quality panel in ISTA. Need to rule out the ignition. But since it's on bank 2 cylinders only, means it's something on that side. What are the o2 readings in closed loop for bank 2? The other fault code makes me think there is a suspect exhaust leak on bank 2 like was mentioned before.

Curious...

[Redd]

Originally Posted by RocketyMan

I would not do this. This engine has ITBs... individual throttle bodies. Which technically means each throttle body has its own vacuum, hence needing the vacuum pump for the brake booster. There is an idle control valve that comunizes the vacuum on both banks. Vacuum doesn't start till AFTER the throttle plate on each ITB.

Taking the plenum off has little to no effect on vacuum. This pressure would technically be atmospheric in the plenum....

This was how I identified a leak in the intake plenum which multiple smoke/water tests failed to detect. Just sharing my own experience.

And for the record the engine runs fine in alpha-n mode without the plenum. I would not recommend doing it for extended periods as there is no filter but a couple minutes just to localize a possible air leak is fine.

[RocketyMan]

Was it a supercharged engine? Hmmm...I have heard of part of the interior tube part cracking. Interesting...