Vehicle details:
WDB2100622BXXXXXX
2001
E240
M112 2.6L V6 +-215000 km

I have a rough idle with noticeable misfires.
I connected my OBD2 reader and found code P0305(Cylinder 5 Misfire Detected).
I left the vehicle to idle and checked the values returned by the lambda sensors and noticed voltage dips on the second bank sensor, see attached images. I plotted bank1 Lamda voltage, bank2 lambda voltage and engine RPM, engine temp at 92C.

From the data I assume that there is an injector on bank2 that did not inject fuel as the low voltage registered indicates an oxygen level close to atmosphere. Conversely a spike in voltage would indicate en excess fuel and therefor a faulty coil pack. And from the misfire in cylinder 5 I assume injector 5 is the culprit. I removed the electrical connector to injector 5 and an open circuit fault was immediately stored indicating that the ECU will detect a wiring problem to the injector. My feeling is that injector 5 is at the end of its life causing it to not inject at times and also causing the fluctuations in bank2 lambda sensor voltage. It fluctuates wildly when compared to the bank1 sensors.

Can any one with experience with the ignition system provide insight into my assumptions?

Attached Thumbnails

     

It could also be dirty, and/or the seal could be leaking.

I would suggest buying some new o-rings (same top and bottom), cleaning around it thoroughly (to prevent dirt entry) and then pull it and inspect it and swap it with a neighboring cylinder, renewing the o-rings in the process. Reset codes and see what you get.

Good luck.

before removing the injector and other bits, replace the spark plugs and retest, If the problem persists then swap an ignition coil from another cylinder.

I had a misfire problem and the culprit was worn out OE spark plugs at 160,000 miles.

__________________
2012 BMW X5 (Beef + Granite suspension model)

1995 E300D - The original humming machine (consumed by Flood 2017)
2000 E320 - The evolution (consumed by flood 2017)

Can't Know, will give it a try this weekend. Are you referring to air being sucked in around the injector?

Zulfiqar, I ruled out the plug as I assumed this would show a higher voltage on the lambda instead of a low one due to unburnt fuel entering the exhaust. I have piston slap, and suspect cylinder 5 as the culprit again so have been eager to pull those plugs for a while to check for fouling due to oil....its just such a mission to get to them! The car has 215 000 km so nearing the point where you had to replace plugs.

One question I have is that if there is a spark problem and the fuel is sent down the exhaust with a full compliment of oxygen, will this show a high or low voltage? Lambda sensors are ideally designed to measure the amount of CO entering the exhaust?

I also could not get the system to run closed loop. I went for a drive and ended with coolant temp at 92C but the control was still open loop due to low temperature. Where does it get the temperature to determine the change to closed loop?
This accounts for the voltage levels around 0.8v which is too rich in my mind. Stoichiometric is at 0.47v.

the O2 stream can be questionable as its an old sensor and old sensors go lazy as time proceeds.

your best remedy would be to start with changing the plugs and swapping coils - once both runs fail then you are left with two possibilities - fuel or compression.

Thats the fastest way to determine the fault with the least expense.

__________________
2012 BMW X5 (Beef + Granite suspension model)

1995 E300D - The original humming machine (consumed by Flood 2017)
2000 E320 - The evolution (consumed by flood 2017)

Agreed. Going to swap coil 5 with one on bank 1 tomorrow and remove those plugs to see what they look like. Just need to rule them out first if possible.
I don't think age is causing the problem as the data from the two banks are reasonably consistent and the bank 2 sensor is picking up fast dips indicating quick response to the change in O2 level.
Will post the findings.

Can you swap the injector to see if the problem moves?

Also, I just looked at the graph and noticed the different axes. Which bank is up around .83 volts (senses too rich mixture)?

__________________
Prost!

Moving the injector is the most work so Im going to move the coil first and replace the plug, just looking for a source of compressed air first to clean out the plug cavity.

Referring to image 3:
Range 0.785 to 0.840 is the yellow line, Bank1 Lambda(colour not visible on the axis)
Range 0.1 to 0.9 is the blue line, Bank 2 Lambda
Range 680 to 725 is the red line, Engine RPM

The colours stay the same for all the graphs but the ranges on the axes differ. Bank2(blue) is the problematic one. Both of the banks are running rich because the fuel system is running open loop due to insufficient temperature.

Just exchanged Coil 5 with coil 1 and recorded the attached data.
Coulours refer to the same sensors as described above.
Misfires are still present although the data looks better than yesterday. Yesterday I tested over a 15min period but couldn't today due to rain.

Attached Thumbnails

     

how are the plugs? are they factory original to the engine?

__________________
2012 BMW X5 (Beef + Granite suspension model)

1995 E300D - The original humming machine (consumed by Flood 2017)
2000 E320 - The evolution (consumed by flood 2017)

Don't know yet, going to try and remove them today to take a look.
I will only replace ignition components with OEM.
The coil packs are OEM, I noticed coil5 has already been replaced by the previous owner as it has a newer part number.
The previous mechanic that worked on the car also spilled a lot of oil on the second bank. I found a lot of thick oil inside the electrical connections to the coils, have started cleaning that out too.

I took the car for a drive to go buy a compressor. This time with coil 1 and 5 swapped. Recorded data during the trip there and during the return trip. Bank1 O2 levels were still very consistent and bank 2 still showed misfires although not as sever as before. I attached the graph of the return trip, recording started just after starting the car in it was still in high idle. Bank 2 is running leaner than bank 1 and fluctuates too much.

I cleaned out the engine with compressed air and removed all the coils for cleaning and inspection as well as the plugs on cylinder 5. All the coils looked good except for the one originally on cylinder 5. This one's low voltage connector as well as the plug connector were covered in thick oil. Upon removing the wires leading from the coil to one of the plugs I saw that in stead of a nice copper pin there is only blue dust inside. The copper pins has completely disintegrated on the wire, see pictures in the attached links.

The plugs on cylinder 5 are also still very good, see attached pictures.
I also removed a plug from cylinder 1 for comparison with one the of cylinder 5 plugs.

I am going to replace coil5 and the corroded cable on monday and test again. I still suspect the injector as the car did drive better today and from the data I could see the dips in bank 2 O2 levels were less severe than before.

Coil pack connection corroded:
https://www.dropbox.com/s/xts9ev8twbruwi0/20140201_155114.jpg
https://www.dropbox.com/s/mjte9ntrrypm0kt/20140201_154958.jpg
https://www.dropbox.com/s/ll2eysjzn0ysxx6/20140201_154939.jpg
https://www.dropbox.com/s/tau29qe5p8tt20b/20140201_154846.jpg

Plug a and b from cylinder 5:
https://www.dropbox.com/s/jrqplbmsq0sku21/20140201_162201.jpg

Cylinder 1 plug(top) compared to cylinder 5 pulg(bottom)
https://www.dropbox.com/s/ry2a6yxq1r3in8z/20140201_162606.jpg

Attached Thumbnails

 

I removed the fuel rail this morning, it came out quite easily. The most work went into removing the portion of the engine harness that was in the way.

After removing the rail I noticed that injector 5 was missing the bottom O-ring and retaining clip. I found the O-ring stuck in the cavity but the clip is nowhere to be found. I attached a thin pipe to my vacuum cleaner with a stocking as filter and after about 15mins I only extracted little bits of hard crud. I assume this is whats left of the clip as the bottom of the O-ring is burnt hard and the top is still a bit soft. This might be the cause of the funny sound emanating from this cylinder, it sounds like piston slap, yet doesn't sound like piston slap.

Fuel rail:
https://www.dropbox.com/s/1sixds1cn0xbtf2/20140202_150132.jpg

Injector 5:
https://www.dropbox.com/s/c6q9wbvw98cjxvn/20140202_150244.jpg

Im going to have to remove the head on bank 2 to see whats going on in there, something I was avoiding. Im also building an injector cleaner to properly clean the injectors and test them afterward.

This simple job just turned very big....

I have started with injector after getting all the parts today and writing the the test program. I used a mixture of 50% petrol and 50% injector cleaner. The mixture was placed in a bottle with a hand pump used to pressurise ans spray poison. The output of the bottle is fed through an in-line filter to the input of the injector. I used an Arduino to control a little relay that switches 5V to the injector. The duty cycle varies and before every change in duty cycle it opens the injector for about 0.5 s. If you are going to do this just remember to take the operating time and release time of the relay into account. And don't smoke. And remember to depressurise the setup before removing the injector.

Link to a video of the setup and injector 4 test cycle #2:
https://www.dropbox.com/s/sqxqoiyjy4jx6tk/20140203_231227.mp4

I did notice on injector 5 that it was not squirting as well as the others and after about 30 seconds it was operating exactly like the others.