Two options to help you hopefully find the culprit:

1) Switch to the rotation conformity page, it's the next one from the selective mass adjustment. Let it sit there for maybe 5 seconds, then switch to the mass adjustment again. You'll feel the car shake violently and at this moment, the worst injector will probably be above the +-5 level. You can repeat this test until the car warms up, after which it will probably not show itself so easily. Try to see which one keeps crossing the +-5 line and go from there.

2) Physically switch injectors 1 and 2, then code them to your car correctly. If then inj. 2 will be in the negatives like currently inj. 1, you've found it. If not, it is gonna be inj. 3.

Happy hunting!

 

I played around a bit in INPA just now to compare the codes on the injectors to the codes in the ECU, and they dont match at all!

Here is what is in INPA: (The engine was cold when I took this screenshot)

And what I read off the injectors myself:
Cyl1: 7ZRF3W
Cyl2: 8HSD6E
Cyl3: 8H8061
Cyl4: BSLR7A

What is really annoying is that I found receipts from an earlier owner that injectors 1 & 3 have been changed by an authorized BMW shop 7yrs/70.000km ago...

Should the engine be running when I change the codes? And are the circles in the codes zeros (0) or Os?

 

It will be a zero.
I'm betting that when you get them properly coded, the car will run much better!

 

Thanks, shold the car be running when I change the codes?

I got an error when I tried with the engine off...

 

Oh yeah, forgot that part. The engine has to be running, yes. If I recall correctly, DIS can do it also while the engine is off but I'm not 100% sure. Just turn it on and do it that way

P.S. I noticed you have an automatic - these are best tested with the gearbox in "D" with your brake pressed down. That gives the engine a slight load that stresses the injectors a bit - helps you find the offending piece easier (if there is any after the coding!)

 

Hello again. When coding the injectors I got an error when using "0" but it worked with an "O". Anyways, the car doesn't run any smoother, maybe even a little bit worse.. I will try what you suggested in the first post tomorrow!

 

Odd, since I've dealt with a lot of injectors (at least 3 sets) and they've all been zeros not O-s.. Oh well, you live and learn
With the codes corrected, you'll have a better chance at finding the misbehaving injector(s) since now the quantity compensation works properly. I personally use the SMA <-> Rot. conform. screen toggling with good success, but be warned, sometimes they do fail in pairs so it's not impossible that both of them are shot

 

Thought I'd update: I replaced injector 1 first with and overhauled one as I was able to do so without taking off the intake manifold(!) You just remove the oil cap seal, the camshaft sensor and the oil cap, then I was able to get engouh leverage on the diesel feed pipe.

That did not change anything, the readings in INPA stayed the same, so I put the injector from cylinder 1 into cylinder 3. And that changed the readings for the better, but it also became apparant that cylinder 2 is now the one in the worst conition of the four. This was also apparent before I changed them using your method of SMA<-> Rot. Conf. The only problem is that the ECU seems to compensate by increasing/decreasing the amount for another injector when one is bad, so it seems like it's 50/50 which one is broken. Maybe it is different depending on if the injectors are clogged or simply worn out..?

So now I will check my motor mounts again as the reason for the vibration at idle.

I also managed to brake two of the screws holding the manifold in place, but I was lucky and was able to just screw the broken parts out again and replace the whole screws with new M6x40 (original is M6x42 I think).

Sorry for a strange and messy post, but maybe someone in the future googleing around will find it useful, hehe

 

Thanks for the update. As you say it really helps when searching similar issues.

I think you are right when you say the ECU tries smooth out the running of the engine by 'trimming' the fuel to the cylinder that is 180 degrees away, however I'm not sure how this will work on six cylinder engines.

 

Usually the injector correctional value has to go above ~+3 to show as poor idle. So those values on your pictures don't indicate any clear problems with injectors. The changes those individual injector codes make is so small that it's usually not shown on any characteristics.

So there are no fault codes? Have you tried jerking/bending the injector wires near the plug if you could replicate misfire? How does the rail pressure reading behave with live data? Have you tried unplugging the rail pressure sensor at the front end of fuel rail?

 

There are no fault codes and I have not tried unplugging the rail pressure sensor. I will try bending the wires abit, however the almost stalling/misfire only happends when the engine is cold and I release the gas quickly.

When driving around with the live data, all the values in the Nom./Act. page are pretty much spot on all the time. If you see the link in the first post you will see more SMA values (I have replaced inj. 1 and 3 since that). The Rot. conf. / SMA - method was quite effective, after checking/replacing #1 and #3, it's evident that #2 is the worst one now (goes up to ~3.5 when changing back and fourth).

Lately I have started thinking maybe there is something with the torque converter of gearbox also as downshifts 2-1 are quite rough when the car is cold (I have an auto). I read somewhere that hard downshifts from 2nd to 1st is usually engine related rather than gearbox, and that is why I have been looking at the injectors. No vibrations at speed and all other shifts are fine though. I checked all the glow plugs and tried starting the car with the MAF disconnected and also the MAP, no change.