Persistent O2 Sensor code

Hi. I drive a BMW 335i twin turbo and I've recently had some misfire issues and o2 sensor codes. I was previously getting cold start misfires on Bank 2 (cylinders 4-6) and o2 sensor codes for bank 2 (P2298, P112D, P166F). In an attempt to fix this I did the following
- replaced all 6 spark plugs
- replaced all 6 coil packs
- replaced the valve cover gasket ( as i noticed some oil dripping into the spark plug holes for cylinders 4-6)
- replaced injectors for cylinders 4-6
- replaced air/fuel sensor for Bank 2 (cylinder 4-6)

The cold start misfires seem to have gone but i still get the o2 sensor codes. When i log my car, upon deceleration I think my car goes into open loop. When i look at my logs ( datazap.me/u/peanuts/afr-bank-2-28-0?log...ta=12-16-17-21-26-28 ), the AFR for Bank 1 is about 234.xx and AFR for Bank 2 is stuck at 28 and remains stuck here. When I restart the car everything is back to normal. This issue is very intermittent. I've pretty much changed all the key components on bank 2 yet the issue seems to be pertinent to bank 2.
A google search of P2298 suggests "O2 Sensor Out of Range During Deceleration Bank 2 Sensor 1" and suggests that the o2 sensor be replaced (which i have already done). I am completely stumped. Any assistance very much appreciated.

thanks
sam

It looks like you completely drop that sensor. When it is acting up, I would pull a wiring diagram and do your basic circuit testing as Paul demonstrates pretty well in his first few chapters. Obviously it sounds like you will have to leave the car running during testing, since everything resets back to normal after a key cycle.

Look for an open when the car is acting up:

B2S1 Virtual Ground Failure –P112D - Pumping Current Failure –P112D
(Virtual Ground) or (Pumping Current) An open circuit in line Virtual Ground or in the line Pumping current can be detected if the sensor signal stocks near lambda 1. The sensor non-activity can be detected by the Oxygen Sensor Signal Monitoring During Fuel Cut-off (signal voltage below e.g. 2,1 V *) in fuel cut off).

Hope this helps!

In addition to the great advice above (AFR sensors are among the black arts to me) ---- From the definition I found BMW P116F is a code pointing to cylinder 3 injector ?

TheTechWhisperer wrote: It looks like you completely drop that sensor. When it is acting up, I would pull a wiring diagram and do your basic circuit testing as Paul demonstrates pretty well in his first few chapters. Obviously it sounds like you will have to leave the car running during testing, since everything resets back to normal after a key cycle.

Look for an open when the car is acting up:

B2S1 Virtual Ground Failure –P112D - Pumping Current Failure –P112D
(Virtual Ground) or (Pumping Current) An open circuit in line Virtual Ground or in the line Pumping current can be detected if the sensor signal stocks near lambda 1. The sensor non-activity can be detected by the Oxygen Sensor Signal Monitoring During Fuel Cut-off (signal voltage below e.g. 2,1 V *) in fuel cut off).

Hope this helps!

Thanks alot for this advice.
Just thought I'd mention I had this o2 sensor issue about 8 months ago on the same bank and the mechanic changed the o2 sensor and the issue was gone. This new sensor had about 5000 miles on it before it came back, hence i just replaced the sensor again to no avail. Based on your advice are you suggesting it is a wiring issue? If so, why is it only on deceleration. As mentioned it is very intermittent and for while there the issue was gone for 2 weeks.

I will try do some circuit testing over the weekend and hopefully this will reveal more.

The reason the fault occurs on Deceleration is because that is when it is monitored.

However, if you look at your pids again you will see that your Fuel Trims on Bank 2 eventually drop out to no activity at all, regardless of whether you are on Accel or Decel. Fuel trims are really the best way to monitor AFR wideband sensor activity, because with AFR sensors the Engine Computer uses slight variances in amperage and processes it in some sort of algorithm to make sense of Oxygen content. Unlike Zirc O2's that give us an actual raw varying voltage to interpret, AFR's don't give us much for a raw reading- (at least I haven't met anyone who can interpret AFR slight mA changes) so we have to depend heavily on the Fuel Trims when using a scan tool on AFR's.

Yes, it could be a wiring issue, like an intermittent open. Heck, it may even be a rubbing wire shorting somewhere, dropping the signal- it's hard to speculate without digging in more. Perhaps when the sensor was first replaced the tech disturbed the harness just enough to keep the wire off of the spot where it was rubbing for awhile.

Don't rule out pin connections at the DME or even the DME itself having an internal intermittent problem. That's why it's so important to do your basic electrical checks while it is acting up to see what's really going on.

The first thing I would do is pull a wiring diagram and read up on Description & Operation of the circuit from whatever database your shop uses.

One thing I always tell my students for successful electrical diagnostics is: "Always create an expectation before you take a measurement" by utilizing the last 2 things I mentioned. Otherwise, you will start to justify your measurements and get confused & frustrated. That will ultimately lead to parts slinging and wasted $.