Bank 2 misfires (all cyls 5-8), reduced power mode, no other codes (any more)

[IamFODI]

Cliff’s:

- Had misfires on all bank 2 cylinders, rough running, reduced power mode.
- Went away when codes were cleared and adaptations reset, then gradually came back over several startups.
- At first the misfire codes came with precat O2 sensor and catalytic converter codes. Swapping in a different used precat O2 sensor solved the O2 & cat codes, but bank 2 misfires (all cylinders) and reduced power mode eventually came back.
- Now running with a brand new O2 sensor. Idle seems a hair rougher than I'm used to, but no other identifiable issues after ~2500 miles (will update this number periodically).


Questions:

1. While the engine is running with no codes, what can I check to see if there's a subtle problem other than that O2 sensor, and how?
2. If the engine continues to run with no misfires, how long can I wait until I assume the problem is resolved?



Full story:

Recently, with no warning I could discern, my car went into reduced power mode while in cruise on the highway. Fault codes per ProTool:

2789 – Catalytic-converter conversion, bank 1
278A – Catalytic-converter conversion, bank 2
27E2 – Oxygen sensor behind catalytic converter, bank 2, signal
2B39 – Misfire with cutoff, cyl. 5
2B3A – Misfire with cutoff, cyl. 6
2B3B – Misfire with cutoff, cyl. 7
2B3C – Misfire with cutoff, cyl. 8
2B41 – Misfire with cutout, several cylinders
2B57 – [engine protection mode]
2796 – [engine protection mode]
2B5D – Cruise control, deactivation
2729 – Oxygen sensor, coast-mode diagnosis before cat., bank 2

Clearing the codes permitted a short time of decent running, but the codes and reduced power mode came back soon afterward. Ended up doing this a few times on the way home. Sometimes the codes included “misfire[s] affecting emissions”, again on cyls 5-8. Occasionally there was an unrelated code, like a cylinder 4 misfire or a code for the low pressure fuel sensor (recently replaced). But I got 2B39 - 2B3C (misfires on cylinders 5-8) and 2729 (precat O2).

For testing purposes, I swapped in an old precat O2 sensor, cleared the codes, and reset adaptations (hadn’t tried resetting adaptations until this point). It ran slightly rough on the first drive, but didn’t drop into reduced power mode. Misfire codes came back (2B39 - 2B3C) without throwing a CEL. There were no other codes. I cleared the codes for the drive home and everything seemed fine.

Was going to address it further but the engine continued running well as far as I could tell. There was some slightly rough running, but not beyond what I’d expect from having just cleared adaptations and driving through wild ambient temp swings like we’ve had around here.

After several days and 500+ miles, it again dropped into reduced power mode. Again, the only codes were for bank 2 misfires (2B39 - 2B3C). Cleared the codes and reset adaptations, and everything was fine for the rest of the day. The engine wasn’t quite as smooth as I’m used to at certain RPMs (mainly a weird vibration at ~half or more throttle around 2k RPM), but it didn’t throw any codes or do anything really stupid.

I tried to feel for exhaust leaks between the header flange and the primary cat on bank 2, and didn’t find any. Then I replaced the precat O2 sensor with a brand new one. Cleared adaptations and hit the road. I’ve done a little more than 100 miles since then. So far, no codes or reduced power mode. The engine seems to idle a little rougher every time I start it up, but it's smooth and responsive everywhere else so I don’t know what to make of that. Maybe I’m just overly sensitive at this point.

Any thoughts on what might have happened? Hope it was nothing beyond the O2 sensor, but I’ve never seen an O2 sensor fail like this…

Are there any diagnostics I can do? Maybe DME parameters I could monitor that might point to a developing problem, like fuel trims or smooth running?

[spammysammich]

how is your battery voltage? Alternator/voltage regulator ok? Are there any loose chassis grounds?

[Redd]

Yeah check ground strap first and foremost.

Single bank misfires likely exhaust side - either exhaust leak or O2 sensor.

[IamFODI]

Originally Posted by spammysammich

how is your battery voltage? Alternator/voltage regulator ok? Are there any loose chassis grounds?

Battery and alternator both replaced within the past year. Voltage seems fine. Chassis grounds seem secure, though they do have some surface corrosion. They're only 3.5 yrs old but I've done 58k miles in that time, including PA winters. Maybe worth preemptive replacement...

[Sammy82]

Check the gasket between headers an the pre-catalytic converters
Mine was done and it smelled a little bit

If the gaskets are ok, check the wires from the DME.

[IamFODI]

Thanks spammysammich, Redd, and Sammy82.

The issue seems to have slowly disappeared over a few hundred miles after swapping in that brand new precat O2 sensor. The car starts, runs, and drives pretty much as it did before now. Coming up on 2 months and 2500 miles with zero codes and no more limp mode.

The only thing I changed was the precat O2 sensor so I'm assuming that's where the problem was. That resolved the codes and limp mode right away, though it took a few hundred miles for the NVH and throttle response to return to ~normal; maybe I got a ton of deposits in bank 2 that took a while to clear out.

I do think I hear the sound of a minor exhaust leak now, though I checked for a leak while I was having the issues and didn't find one. The noise started just after all the issues cleared up so I don't know what to make of that.

Either way, I'm pretty satisfied that the problem has been resolved, though I still might try some datalogging to be sure and am still interested in other ideas.

[ajks]

When this happened on my own E92 M3, the problem was (as Sammy82 suggests above) that the gasket between the downpipe and the exhaust pipe (i.e. before the pre-cat AFR sensor) had disappeared, causing a big air leak and a throatier roar than usual.

Putting in new gaskets (2x 11627830668) solved the all-one-bank misfires. You would also need new bolts (4x 11621318568) and nuts (4x 18301317898).

[IamFODI]

Thanks, ajks.

Since my previous post, I had the exhaust smoke tested. Found a leak at one of the muffler flanges but nothing upstream of that. The engine has continued to run without any perceptible issue. Will continue to keep an eye on it.