Weird mixture adaptation values...

[Albert850i]

Hi there,
After a few years (!) of debugging, my 850i finally runs nicely and on all 12 cylinders. Oxygen sensors are in closed loop all the time (I mounted two small gauges on the dash to monitor this), so I am happy. No fault codes.
However, reading out the DME's yesterday using DIS, I noticed that both mixture adaptations LL (idle I guess) are way out, see images. On bank 1-6 it is 84 (should be around 0) and on bank 7-12 even 127. Again the engine runs and idles smoothly.
Please let me have your comments (is the engine running fine or not?).
Thanks, Albert NL
IMG_6726.jpgIMG_6727.jpgIMG_6725.jpgIMG_6723.jpg

[Timm]

Negative adaptations at low load is usually poor fuel-rail pressure regulation, so make sure you are getting a good vacuum at the lines that run from the manifolds to the FPR's. Without the vacuum the FPR's will increase the fuel-rail pressure which should be 2.5 Bar at idle and 3 Bar at WOT.

[Albert850i]

Thanks Timm (I watched your video on mixture adaptation).
The corrections are +84 and +127: are you saying these positive corrections are due to a mixture being too rich? This is critical I as want to understand what causes these high figures.
I checked fuel pressures before and measured 3.0 bar during running and 2.5 when sucking on the (new) FPR hose. So probably OK.
BTW I installed 2 brand new Bosch oxygen sensors 15 miles ago, perhaps the DME's are still learning? See also attached image.
The BMW diagnostic procedures specify "93 mixture adaptation LL" to be between -48 and +32, so my values are currently way out of spec. However the engine seems to run fine so I guess the adaptations are doing their work properly..
Afbeelding1.jpg

[Timm]

Those are negative adaptation values - you can see the edge of the diamond pointer at the extreme left of the slider scale. You should have 2.5 Bar at idle as the manifold vacuum will be high reducing the fuel pressure. So, if you have 3.0 Bar at idle there's your problem!

[Belgiumbarry]

Timm, i think Albert meant 3 bar with FPR disconnected , 2.5 bar with vacuum on . So this is good i suppose ?

[Timm]

Yes, that would be fine.

[Albert850i]

Thanks, now I see the blue diamond at the very left end of the scale. So I guess mixture at idle way too rich. Interesting. I will re-measure fuel pressures just to be sure.
Something else jumped to mind: a while ago I installed 12 brand new upgraded model Bosch injectors (the blue plastic ones), perhaps these required the DME's to shorten injection times.

[Timm]

'Upgraded' always make my heart sink when it comes to injectors - unless you are going to use forced induction and a re-map of the DME then the standard injectors are perfectly matched. I wonder what is upgraded about them - do they let more fuel in for the same injector dwell?

[BMSman]

Are these adaption values in DIS 'Mixture Adaptation LL and TL' the same as 'Mixture additive (idle)'/'Mixture Mult(normal)' in Inpa?

What does LL and TL stand for?

[gon850]

Hi

and for negative values... I have -38 and -10 AND -50 -11 ?

Where do I need to look?
new lambda, no error codes, runs fine... maybe some very minor vibration on idle, good fuel consumption but needs oil frequently 1 qtr/1500 kms

Thanks

[Albert850i]

I believe LL stands for Leerlauf = idle and TL stands for Teillast = part-load.

[shogun]

Yes, LeerLauf = idle, TeilLast = partial load

[Albert850i]

'Upgraded' always make my heart sink when it comes to injectors - unless you are going to use forced induction and a re-map of the DME then the standard injectors are perfectly matched. I wonder what is upgraded about them - do they let more fuel in for the same injector dwell?

Timm, Bosch seem to have replaced the original 0280150715 with an improved 4-nozzle version 0280155885. As these "upgraded" injectors can be found for around 10 Euro each, I decided to swap them just to be sure. I believe the flow rate is about the same.
Afbeelding1.jpg

[Belgiumbarry]

10 € each ???

[Timm]

As long as there is no error code associated with the mixture adaptations then the engine is running fine - the adaptation values have pulled the air/fuel mixture back to where it should be. There will be variations in the MAF, fuel pressure, injectors and the efficiency of the combustion all leading to the adaptation values that the DME's have stored. It is worth saying that the Lambda sensors are usually spot-on as they are comparators (narrow band) rather than true Oxygen meters.

I am not sure what the actual adaptation value (on the M70) is before it will throw up an 'adaptations at limits' error - but it does seem to be rather large.

[Albert850i]

10 € each ???

Yep, search for the Bosch number on Aliexpress and you will find them (I paid 109 Euro for 12 pieces incl shipping).
Note that also Bosch has them made in China..

- - - Updated - - -

I am not sure what the actual adaptation value (on the M70) is before it will throw up an 'adaptations at limits' error - but it does seem to be rather large.

Indeed no fault codes, so the adaptations appear quite dramatic but are still effective. I will check again after a few hundred miles.

[Albert850i]

Update: engine runs fine on all 12 cylinders, load signals L and R both 1.6 ms, so equal. No DME errors.
However, still the mixture adaptation at idling is way out. Bank 7-12 even shows -127 which is off the scale (running way too rich?).
I don't like this. Fuel pressure measured OK, even the MAF's replaced with after market ones, all no change.
Suggestions?
screendumps 1-6 7-12.pdf

[Timm]

You may not like it much - but your engine is running fine with no error codes! It really leaves two things - fuel pressure too high at idle (it should be 2.5 Bar) or the uprated injectors let too much fuel through for a given dwell - that's it really.....

[Albert850i]

Thanks Timm, I agree that too rich a mixture can only be caused by too high fuel pressure, or too much injector delivery. Will try with the old injectors again and see.

[Belgiumbarry]

Yep, search for the Bosch number on Aliexpress and you will find them (I paid 109 Euro for 12 pieces incl shipping).
Note that also Bosch has them made in China..

- - - Updated - - -



Indeed no fault codes, so the adaptations appear quite dramatic but are still effective. I will check again after a few hundred miles.

that's a crazy price Albert....due my supercharger i needed bigger injectors ( about 300 cc instead of the oem 150 cc ) . I found them in Germany , same Bosch type in dimensions but 330 cc , used in the '90s in Audi Quattro. They were around 100 €/piece!
With the kit at the time i had smaller ones ( assume a Ford type ) with extension caps on , were also near 70 €/piece.... but when i had a serious fuel leak at a extension cap , i didn't thrust it anymore and bought the Bosch .
Later i found out that the O ring grooves in the extension caps were not made as it should be , the O ring was also sideways pressed in it , so no clearance to "work" under pressure.

amazing you find them at 10 €/piece....

[Albert850i]

Attached screenshots from INPA confirming the way-out-of-range adaptation values.
I will see whether the original injectors make a difference.
IMG_8820.jpegIMG_8821.jpeg

[Timm]

Confusingly INPA is showing the values as positive - ie the mixture is too lean rather than too rich. I think I would rather trust INPA and not immeditaely blame the injectors and instead try and find a vacuum leak!

[cartoonz]

gonna state the obvious, since nobody else will... $10 "Bosch" injectors from Ali express? Counterfeit. Bet you a dollar to a dime that if you put those on a test machine, you'd quickly see why...

[Albert850i]

Really can't tell whether these are OEM Bosch or fake, see photo.
Anyway I changed bank 1-6 back to the original injectors after ultrasonic cleaning, little difference unfortunately.
I am confused as to whether the adaptation problem is caused by generally too rich or too lean mixtures.
DA838949-766E-408B-98AA-D81985DFB27A.jpg

[Timm]

As I often use INPA I can confirm that the initial mixture is too lean (as measured by the Lambda sensors). But it is worth saying that the disparity can be due to MAF problems as well as vacuum leaks.