E39 M62TU 540i Misfire

[pgnuta 4]

Thought i would document my progress tracking down this issue, partially as a guide and partially as a warning for future historians. If anyone has any suggestions of things to try then feel free to pitch in.

The problem initially presented itself as the check engine light blinking intermittently and then finally going on solid. While this was happening the car would run a bit rough and when the check engine light came on solid it lost a lot of power.

I checked the codes using a cheap OBD2 scanner in the form of an ELM327 bluetooth adapter (http://www.trademe.co.nz/motors/car-parts-accessories/tools-repair-kits/meters-testers/auction-922793925.htm) and an android app (https://play.google.com/store/apps/details?id=com.pnn.obdcardoctor&hl=en).
This android app comes with my recommendation because it is free and will read and clear codes without stupid in-app purchases. The code that it was producing was P0302.

The research that i did seems to indicate that the blinking check engine light is telling you that the cylinder is misfiring right at that moment, if it happens too often then the computer just gives up and shuts down that cylinder until the next time that you switch the car off. That is represented by the check engine light going on solid and the car feeling sluggish. I also noticed that the fuel consumption gauge would go right off the scale at the 30L/100K end when it happened and it wasn't caused by my right foot.

I did some research into how the computer detects a misfire and its quite interesting so some of that stuff can go in here too. The computer uses the crank angle sensor to detect misfires, if the angle on the crank isn't exactly what it should be after a cylinder ignition then it registers a misfire.

On most applications, OBD II uses a “block learn” strategy to track misfires. Any misfires that occur are recorded every 200 revolutions of the crankshaft. The tally is stored in 16 memory blocks, so every 3,200 rpm the misfire count starts over again.

As long as the number of misfires in any given memory block remains below a certain value, there is no problem and OBD II gives that block a “pass” rating. But if the number of misfires in that block is too high, the block receives a “fail” rating. By averaging the pass/fail ratings of all 16 blocks, the OBD II system gets the big picture of what’s actually going on. By all accounts if the total misfires on a cylinder exceeds some threshold (unknown) then it will shut off the fuel and spark to the cylinder completely so-as to avoid damaging the catalytic converter.

One downside to using the crank angle sensor to detect misfires is that it can be fooled by rough roads. Most manufacturers will therefore disable the misfire detection while rough roads are being traversed. Most manufacturers will also disable the misfire detection while the engine is being cranked over with the starter motor. My experience so far finds this to be the case but additionally it seems to not register misfires when the engine is cold or when the car is not moving (not sure which). This makes it difficult to diagnose the problem because i need to take the car for a drive to trigger the fault but i need to avoid getting the car too hot to work on afterwards.

So with a cylinder misfire its generally spark or fuel and because the problem was sticking to the one cylinder and i thought i could smell fuel then i thought it was probably spark. I thought i'd crack on and do some basic checking, try swapping the coils and plugs around to see if i could make the problem move to another cylinder.

The BMW V8 engine is a beauty to behold, i'm still in love with this car despite the fact that it hasn't run right for me yet. The plastic covers on the top and on each bank are easily removed with a flat-head screwdriver and a 10mm socket. First take off the top cover by pressing down firmly on each of the four round buttons, you may need to hold the cover up slight while doing this but each button should click to indicate that its disengaged (unless of course it wasn't engaged in the first place like mine). With that removed and put aside carefully then attack the cover on each bank, if you look carefully you'll see two square inserts in the middle of each cover, each of the inserts has a notch on the bottom that you need to use a flat-head screwdriver to lever out. Be very careful not to damage, one of mine was broken and someone has glued it in place because they obviously didn't know what they were doing - If anyone has a spare anywhere then i'd appreciate a contact. Finally get your 10mm socket in there and remove those bolts holding on the covers and you're done.

Below is the firing order and the cylinder numbers:

Firing order
1-5-4-8-6-3-7-2

Cylinder numbers
4-------------8
3-------------7
2-------------6
1-------------5
Front of engine

Coil swap

Now you can see all coils. The M62 has individual coils, one for each cylinder. Each coil is retained by two 10mm nuts. Get your 10mm socket in there and give them a tweak to loosen then and then take them off by hand to avoid dropping them into the engine.

I decided i would swap around the cylinder 1 and cylinder 2 coils. Removing the plugs from the coils is as simple as pulling up the silver retainer clip and pulling the plug out sideways, if you find it easier you can slightly lift the coil out of the head to give more room to disconnect the plug before completely removing the coil. Swapping the coils and putting the 10mm nuts back on by hand and tweaking tight with the socket completes the job. Make sure you don't tighten the nuts too tight or you'll damage the heads. Also make sure that you remember to replace the two earth lugs on cylinder 2 before putting the nut on. I think that there are also two earth lugs on cylinder 7.

Afterwards taking the car for a drive again after clearing all the codes shows no difference. Next step swap a plug around...

Plug swap

So off with the two coils again, cylinder 1 and 2 because they are the easiest to get at. Pulling the plugs i found that the socket attachment plus the 100mm extension from my Fuller Pro socket set gave me the correct length and enough clearance from other engine bits. Starting with the strong-arm bar to loosen and then winding out by hand to remove. Inspecting the plugs i found them not particularly interesting in any respect, not wet, oily sooty or anything of note. I did however note that they were the factory original Platinum NGK BMW stamped plugs with the four lugs on them, given that my car has done 216,000kms they are long past due for replacement so i had some hope that i was on the right track there.

So swapping them over and putting them back in, i always insert plugs by hand because if you don't then you risk cross threading them and ruining your head. There should be no resistance until you get right to the bottom of the thread, if there is then back it off a bit and try again, there might be some gunk in the thread. At the bottom of the thread give them a tweak to tighten but not too tight, as these are used plugs they have no 'quarter turn' tightening recommendation to squash down that washer.

Coils back on and everything back together and off for another drive. No difference... bugger.

Plug swap V2

In the meantime i've been on Trademe and managed to pick up some NGK BKR6EIX plugs. Got 4x plugs for $30 and the other 4 plugs from a second trader for another $60. Pretty good price for plugs that retail for $30 each. Also of note is that when the first batch turned up they had stock number 2272 written on the box, i wasn't sure whether they were the right plugs because by all accounts its very inadvisable to gap them yourselves so you've got to buy the right ones first up. Most of the online documentation i read mentions stock number 6418 so i started to worry. Turns out the stock number isn't much to worry about and the 2272's work just fine. There are three varients of the BKR6EIX plugs available: BKR6EIX, BKR6EIX-11, BKR6EIX-P. You want the base BKR6EIX ones.
In the end the second lot of plugs were stock number 2272 too so that was a happy coincidence.

Replaced all the plugs using the technique described above, making sure to be very careful putting in the new plugs so-as not to cross thread them or damage the electrodes. Everything back together and off for another drive.

No difference... bugger. Although the performance when it is running on all 8 seems very good, it shows promise!

So back to the garage to try some other things.

MAF

I thought i'd unplug the MAF and see if that helped, started the car in the garage and then pulled the plug, it had a bit of a cough and settled back to its usual idle. Now its usual idle still sounds rough but i'm not sure as i don't know if i've ever heard it idle properly. I've got an INPA cable coming from Aliexpress so i guess that will tell me a bit.

Anyway with the MAF unplugged i went for another drive, it seemed to hold out for longer before shutting down cylinder 2 but that might have been my imagination. Ultimately it still produced the same code P0302 and now with an additional code about the MAF which i expected, back to the garage and put it back together.

Plans for the future

So there are a couple of things that i've got left to investigate:

There is quite a loud unexplained whistle coming from the throttle body when the ignition is on. This happens whether or not the car is running, it's just harder to hear when the car is running but its definitely there. The throttle position readout on my android app appears to read correctly though so i don't have any reason to suspect that there is a fault there, its just odd.

Secondly the snapshot from the OBD computer when the cylinder shuts down shows short term fuel trim % on bank 1 as -100%. Reading up on this i can't find any documented cases where the STFT has ever read -100% on a stock vehicle that isn't being messed with like with an aftermarket computer. Obviously the computer thinks that the mixture is super rich and its trying to wind back the fuel on that bank and there are a number of potential causes of this:

1. False reading/ artefact of cylinder misfire
   It could be that because cylinder 2 is misfiring so badly the unburnt fuel is causing a false reading on the O2 sensor.
   
2. Stuck / Leaky injector
   Could be a faulty injector in cylinder 2 is over-riching the mixture in that cylinder to the point where it won't ignite. I'm considering trying to swap two of the injectors around to see if the problem moves between cylinders but i'm concerned by what i read about the complexity of the operation and the potential to damage seals, o-rings etc.
   
   I gather that you can lift the whole fuel rail, injectors and electrics from the motor to access the injectors by removing 4x 10mm bolts from the fuel rail itself but not having attempted this i cannot confirm. Just standing looking at it i can't see how it could possibly be that simple.
   
3. Overpressure in the fuel system
   It could be a partially blocked fuel return could cause an over-pressure but i would have thought that it would present itself in more that one bank and certainly more than one cylinder. In any case i wouldn't know how to test this.
   
4. Stuffed O2 sensor
   Its probably worth going for a drive and monitoring the output from the STFT and any relevant O2 sensor outputs to see if there is likely to be a general fault with the sensor
   
   I've only done very light reading on the subject of O2 sensors but i gather that there is one before the CAT on each bank and possibly one after the CAT but then i haven't closely looked at the exhaust system to see how many CATS there are and such. Perhaps thats something i will do next.

I'm also waiting for my INPA cable and i'm not ruling out the possibility of a vacuum leak although i would have thought that that would present itself as a lean condition, not a rich one and it would most likely not just effect cylinder 2.

I'm also not ignorant to the fact that cylinder 2 is the last in the firing order but i'm not sure if there is anything to that.

More as it develops and in the meantime if anyone has any bright ideas from anything that i've written then please post a reply. Bear in mind i'm limited to what i can achieve with a multimeter and a variety of tools.

Edited July 26, 2015 by pgnuta

[B.M.W Ltd 950]

Was it just a faulty coil ?

[pgnuta 4]

Nope see above my edited and now complete post, sorry took me a long time to write. More as it develops but i'm keen for some ideas.

[B.M.W Ltd 950]

I'd be using a proper scanner first before going through all the rigmarole youv'e gone through. ^^^^ up there

[westy 614]

Codes are crap, live data is what you need using a proper scanner as Glenn suggests.

Anyway, do a quick check of the middle (large) plug to the DME. The thermostat can pump moisture through the harness if it's bad and corrode the plug and will give you all sorts of random codes and, um, stuff. Best disconnect the battery first.

[pgnuta 4]

Is a 'proper scanner' the INPA software and the 20 pin cable that i've ordered or something else?

[dkonsta1 109]

20 pin obd is the best way despite the fact you also have a obd2 under the steering wheel.

2 things from my side. Correct spark plugs are bkr6equp (3199)

Not sure if something else will work properly.

Also I have a spare MAF brand new if you want to make a test. You need to allow time though because your ecu takes a while to adapt the new readings. Good luck with it. Cheers

[hybrid 1043]

O2 sensors show on the short and long term fuel trims ... if they're off by a lot the DME is compensating and it will run like a bag of dicks.

There are lots these engines problems manifest themselves as, coils (cylinder shutdown), spark plugs, cam timing chain is stretched/guides worn (runs rough) as westy said, water ingress into the data-map thermostat control wire, just going near them when your in a hurry, I dont think the MAF (or lack of signal) will cause it to run rich unless youre like tony and lives down some sh*t dirt road which rattles the internal MAF wiring loose which cause the DME to jump in and out of alpha-n mode and in doing so shuts the engine down.

Sooo .. when you say loss of power .. does the whole thing shake when it has no power and feel generally un-balanced?

Your friend is using INPA and live data ... now its a pile of sh*t as far as software goes but will give you the key live data you need to check if the thing is in range or not.
toss away the generic OBDII software, its for noobs and the VTNZ, use BMW specific software as it will give you the most useful descriptions.

[Alex Carter 0]

Is a 'proper scanner' the INPA software and the 20 pin cable that i've ordered or something else?

you will also need a 20-pin to 16-pin extension cable to work with 16-pin INPA

[pgnuta 4]

Yeah I ordered the usb 16 pin plus the 16-20 pin adapter plus whatever software is on the dvd it comes with. I gather that the software is generally available on the Internet, I've found a bunch of dead links from other forum articles but eventually got a torrent from somewhere. I'll mess with the software a little more when I get the cable though.

Sent from my iPhone using Tapatalk

[pgnuta 4]

Well a brief update to this... I got the cable and software load from China the other day and found an XP laptop to use it with.

I attempted to load it following the instruction video that came on the DVD which was a cellphone recording of a youtube video of a cellphone recording of a Chinese Windows XP laptop being installed with the software. The basis of the installation and the tweaks were all there but something went wrong in the installation process and the software wouldn't run properly. After digging around elsewhere i found out what was wrong and started again...

From what i understand the issue is that english versions of the INPA data files stopped being available after a certain version so you have to make sure that the data files are copied in from a version that had them. I ended up with so many different versions of archived and extracted copies of the software that i eventually get a combination that worked and had all the data files in the right place.

I didn't have any problems getting the USB cable working in the end, that was fairly straightforward.

So last night i had some time to mess with the car and plug in the laptop, i took a couple of screenshots and exported the error codes but they are on my XP laptop at home so i'll have to post more later with those specifics.

Initially i had a look at the rough idle monitor and it looked very strange. My version of INPA measures the roughness of each cylinder from 0 - 1000. Most of the cylinders were all over the place but within the boundaries except for cylinder number 3 which was way out at 2500, it had a ! in front of the number which i guess is the software telling you that this is REALLY not normal. Cylinder number 2, the suspect in all this, was sitting on 0. I unplugged a couple of the coils one by one to see if anything changed and nothing really obvious presented itself. About this point i decided to go exploring the rest of the software so took a screenshot of that page and went to look at the error codes.

In the codes section i found 5 error codes. Off the top of my head i don't remember them all now but one was a misfire on 'Zyl 6'. So this is getting strange, OBD thinks that Cylinder 2 is doing something strange, the roughness output looks bad for Cylinder 3 and the code is present for Cylinder 6. However i took this with a grain of salt as i didn't know how old these codes were so i thought i would clear the codes and check them again tomorrow after taking the car for a drive.

Now this is where i may have made a fatal mistake... When i was clearing the codes i also saw a function to 'clear adaptations'. I didn't know what it was and i really should know better then to press random buttons but i pressed it anyway. Immediately the car's idle dropped to almost a stall and a second later recovered to normal. I thought that it was a bit strange but didn't think much of it, packed the car up and went inside to do some more reading.

This morning i went to start the car to drive to work and it idled very badly, much worse than before, barely running. I tapped the throttle and it almost stalled and then the engine failsafe warning flashed up on the dash. I turned the car off and started it again and the same thing. I thought i would try the 'unplug the MAF' trick to see if that helped but it only made it worse, when i started the car it would immediately stall. Anyway i was out of time at that point so left the car as it was.

I'll try to read the codes again tonight but i'm not hopeful of a good outcome here, after clearing the codes with INPA last night i didn't immediately get any codes repopulate so i'm not sure what causes codes to populate, whether it is the act of starting the car or whether it needs to be driven a short distance, if it is the later then i'm out of luck because with the engine in this state i doubt i could even engage a gear let alone drive it anywhere.

More to follow.

[B.M.W Ltd 950]

I would suggest getting someone with the right diagnostic gear & experience to check you car and get on with your life. All of this up there ^^^^ for a probable very simple diagnosis and repair now complicated by your attempt to diagnose and correct it. Clearing adaptions on fly by wire systems need reinitialising throttle values

Edited August 18, 2015 by B.M.W Ltd

[pgnuta 4]

Yet another update.

Thanks to Glenn's suggestion i found some info about how the computer reinitialises its peripherals after you clear the adaptations, for future reference the procedure is to turn the ignition to position 2, don't start the car, leave it like that for 30 seconds and you'll hear a bunch of clicking as the computer relearns what its got connected to it. Then turn the ignition off for 10 seconds. Then start the car and it should be back to normal. So that resolved my failsafe issue.

Afterwards i cleared all the codes and took the car for a drive to see what popped up; General cylinder misfire code and 'Zyl 6' Misfire code. We're getting somewhere now.

So off with cylinder 6's coil and swap it with cylinder 5 and clear the codes and off for another drive. Now we've got a cylinder 5 misfire, ah ha!

I've bought a replacement coil off trademe and we'll see how that goes when it arrives.

In the meantime its interesting to consider how the standard OBD tools pointed towards cylinder 2 instead of cylinder 6. Obviously the BMW software is designed specifically for BMWs but the software is not interpreting the codes and coming up with its own ideas, it's simply reading what the DME has stored there. Ultimately what i'm saying is that surely if the DME new enough to store the BMW specific 'cylinder 6 misfire' code then it could have also stored the generic OBD cylinder 6 misfire code instead of the cylinder 2 misfire code. Long story short i'm sold on the 'proper tools' and wouldn't recommend trusting the OBD plug, at least not on a car equipped with a 20 pin plug under the bonnet and certainly not with generic software.

[hotwire 352]

So, is that a new coil you have bought? I hope so, Otherwise you probably continue chasing your tail.

With total respect, & i appreciate you may be wanting to familiarise yourself with diagnosing, this has been a long drawn out exercise to confirm a faulty coil.

[Guest cortez]

Holy moly that's one big ol post there cat man ! have ya figured out all the hu harr yet ?

[pgnuta 4]

Yeah was just the coil in the end, i wasn't going to bother updating this thread because everyone seemed to have missed the point of my post.

The point was that i wanted to document the process for the next person that came looking for the same information. Ultimately if no-one contributes knowledge in general to the great internet ecosystem then what does it become?

If there is one important lesson that i've learnt from this exercise its that i need to keep it to myself for fear of running afoul of the opinionated and the agenda pushers.

[hotwire 352]

Don't take it the wrong way Tim. Most replies to this post have been positive & virtually all advice has been correct - types of diagnostics, live data etc, knowledge etc.

There are a multitude of pitfalls to fall into with one having no or limited knowledge & equipment to diagnose with. Incorrect codes, out of parameter codes etc

I have seen this so many times at work where someone has has replaced parts on a whim or a fault code, only to end up with the car at work to fix the real problem.

Glad you have it sorted though & you will now have a better understanding of things & no doubt be pleased that you sorted it yourself

[darek4488 0]

I just registered only thank you, pgnuta.

I am so glad that you wrote all this down, because it seems like my E53 X5 with M62B44 seem to have the identical symptoms and from what I have checked so far I seem to have the exact same issue you had. What I have read here is very helpful to my case. I know that you sir are no longer active, but I just want to encourage more people to document their troubleshooting in a similar fashion. You guys might think nobody will ever use it, but years later some bloke from the opposite side of the world would speak English and he'll be browsing for hours to find out what's wrong with his bimmer.

Cheers from Poland

 

[pgnuta 4]

You're very welcome, i'm glad my post could help someone.

I ended up selling my beloved 540 after it developed a nasty case of 'randomly not starting and/or stalling while driving'.  I decided that as much as i loved its power and comfort i just needed a reliable car.

Best of luck with your issue!