E39 M62TU runs crappy at one specific temp

[keltik 15]

Hey guys.

My 540 gets its lumpy idle on at 65*c coolant temp. Plugged an ODB tool in and discovered at almost exactly that temp every time, the idle will start sounding more like a cammed up LS and the fuel trims for both banks go straight to max (+20%). Still runs great under load and the problem only really shows when idling. Soon as coolant temp hits 75*c it goes back to a smooth idle.

Its been a constant problem since ive had the car.

Never really worried about it until this morning when it ran like a bag of dicks and set the CEL. Will pull the codes this afternoon, but until then - would anyone like to guess what my problem is? I'm willing to bet the codes will just be "Missfire ALL THE CYLINDERS!".

If I had to guess (and i do), I'd say that this thing has a fuel pressure problem. Do these have a multi-speed fuel pump? I reckon its set to drop pump speed down to low at a certain engine temp but fuel pressure is falling off too much which is why the trims are going up to max rich...until the engine warms up enough to run on the leaner setting.

[BreakMyWindow 1870]

Havn't heard of this one before. Be interested to see what the DME reports once you've cleared the fault codes and the issue has re-occured.

[keltik 15]

Yup. Missfire cylinder 1,2,3,6 and maybe 8 came up on the scan tool. Either way it was most of them. Wasnt confined to one bank. I think there was also a fuel trim lean code but couldnt quite read it before the technician I stole had to run away and do some real paid work.

This afternoon, I kept an eye on the trims while it warmed up and noticed both banks crept towards 20% enrichen, then one bank wavered around between 18-20% and the other bank reported 0% trim....which is interesting (maybe the DME discarding that sensors data due to implausible values?). Revved it up and both trims got back to around 0 then back to idle started creeping richer.

Will monitor the problem more and try to record more solid data on when the problem occurs. Might get the wideband o2 on it if I can summon up the effort to rig it up.

Edited April 14, 2015 by keltik

[keltik 15]

Well i still have no answers. Got a 20pin cable on the way from Chong-Kongapore. Will see what live data i can get with that. Might try snag the wideband 02 this week and see what it says.

Heres a video, you will need headphones or good speakers to pick up on the engine note. I tried EQ'ing it to make it more obvious but my software just isnt set up for that. So dont bother if all you have is laptop tincan speakers.

https://youtu.be/zS6v5l5n7pQ

[BM WORLD 1258]

vac leak ??

pcv etc

bad maf

[keltik 15]

Finally resolved. Got INPA running on my laptop, plugged in and checked out the important things:

Fuel long term trims were quite high. When running cold / warming up, the ECU was having to dump in quite a large enrichment correction and was running out of adjustment triggering my check engine light.

When hot, the ECU was able to compensate. There must be a point when the warmup enrichment has tapered off and the ECU has less room to compensate, causing the problem to only show up at a narrow temperature range.

So i checked for vacuum leaks, thought about replacing the CSC valve thing on the back of the manifold....but that looked hard.

I eventually kept a good watch on all of the parameters in INPA as the car warmed up a few times and noticed intake air temp was wrong at low air flows. It was showing 40degrees C on a cold engine that hadnt been run that day! Off idle, the air temp would drop to something like 24 but thats still a big jump from the 15 outside. Idle air requirements didnt seem to vary much with A/C or power steering loads...so that was odd.

Took a chance and replaced the airflow meter. Didnt want to pay the big tax for a BMW or Range Rover one so ended up slamming in the Volkswagen one. Works great! The internet was right, its a brilliant low cost alternative. $170 compared to $800 for a BMW one or $500ish for a Range Rover one.

The car is now running awesome, it drives better even tho i didnt notice anything wrong with the performance. Shifts are smoother and pulling away from intersections is crisper. Idle quality is still only a 9/10 but I'm very picky. Perhaps there is something else going wrong but I'll keep an eye on things and see how we go for now.

Edited June 13, 2015 by keltik

[BreakMyWindow 1870]

The proper bosch one ex o'seas is not all that expensive.

I too tried the vw maf and thought job done. Got the correct maf and got the 540i low end torque back.

[keltik 15]

The proper bosch one ex o'seas is not all that expensive.

I too tried the vw maf and thought job done. Got the correct maf and got the 540i low end torque back.

Problem came back so gather round kids and learn about how i finally fixed it for reals;

Codes for misfire all cylinders and adaptions maximum enrichment reached.

VW MAF sensor reads about 5% richer than the BMW one, this can be seen in that my long term fuel trims are 95% compared to %103 with the BMW sensor in there.

For a short time, this extra fuel due to the maf sensor change caused the car to run a bit richer which allowed it to compensate for the lean condition and make the problem appear to go away.....until the overall fuel trims adjusted and brought everything back to square one.

So it had to be a vacuum leak. I spent a couple of mornings getting rowdy with the brake clean and finally found the culprit, this f**ker right here!

So $50 worth of seals from Coombes later, I've taken it for a good long drive and here are the fuel trim numbers before and after;

As we can see, the long term trim has taken out the error from the wrong maf sensor and the short term trims are close enough to spot on.

After a few cold starts to verify success, i'll do my happy dance.

[dkonsta1 109]

Very well done and structured troubleshooting. Good on you.

[Olaf 3243]

as a design engineer used to say to us techies when we'd sleuthed out and found a solution to a pre-production issue: "good sussing".

what was the gasket/seal in question?

I was wondering from your descriptions if it was a temperature sensor somewhere in the chain sending an erroneous signal to the DME. Air leaks cause some pretty interesting issues with our vehicles, eh.

cheers

Olaf

[keltik 15]

as a design engineer used to say to us techies when we'd sleuthed out and found a solution to a pre-production issue: "good sussing".

what was the gasket/seal in question?

I was wondering from your descriptions if it was a temperature sensor somewhere in the chain sending an erroneous signal to the DME. Air leaks cause some pretty interesting issues with our vehicles, eh.

cheers

Olaf

Throttle body adapter plate on the front of the manifold was the culprit. Not a common area to leak it seems.

Problem is definitely fixed! Took me a week to find out as the alternator died for the 2nd time right after I fixed this.