Little bug in the CSL program

[MpowerE36]

Hi everybody,

I think I’ve found a little bug in the CSL program. It began when I saw this video on YouTube :

https://www.youtube.com/watch?v=pLuBJZRSNRI

I noticed that RF from MAP sensor was very strange : at idle (~300mBar) we have about 40% which is huge and at WOT (~1000mBar) we have about 30% which is very little.

So I checked the code of the program and I didn’t see any problem in the calculation of the RF from MAP but I noticed a strange thing in the way the program reported the word in the DS2 respons. You can see it below :



I modified the code like this :


(There are many ways to do the correction)

Now I have consistent values of the RF from MAP sensor :

https://www.youtube.com/watch?v=Q0_Ek1BHqbo

[Tomba]

Always impressed with this kind of research.

On MSS50 ECU I see something other. On DS2 the MAF air flow seems to overflow. Like the value is 12/16 bit while only 8 bits are shown. The reported value looks like a jigsaw profile when air consumption is increasing. Another method would be to log it by regular values but these are quite slow.

[SliM3]

Ha, that’s one of my old videos!

For transparency sake, that data log was populated using a GM 1 bar MAP sensor so the linear scale is quite different than the stock sensor. Minimum pressure is around 150 mbar with an offset of nearly 102 mbar.

Going to eventually swap to an AEM pressure transducer which has a linear scale that actually starts at 0 & .5 volts. Those older GM sensors aren’t really the best choice for this application.


Sent from my iPhone using Tapatalk

[MpowerE36]

I use a Bosch MAP sensor and before the correction I had exactly the same result you had in your video. Voltage and MAP pressure values are consistent.

[SliM3]

I use a Bosch MAP sensor and before the correction I had exactly the same result you had in your video. Voltage and MAP pressure values are consistent.

Good info! I never logged any other sensors so didn't have a baseline to compare against, so this is good to know.

[karter16]

Thanks very much for this - Was looking at my logs of "RF from MAP Sensor" and thinking it looked very strange.

[ppm008]

CSL program doesn't use RF from MAP Sensor
RF is only calculated by TPS.
You can pulloff MAP Sensor and there will be no difference in driving
MAP sensor is only used if one TPS has malfunction

[MpowerE36]

It is almost correct. In normal operation the final rf is the rf from tps + long term adaptation based on the rf from map sensor. This correction is limited to +/-2.5% of rf. There is configuration byte for the rf calculation in the csl sw. One of these configurations defines the rf from map sensor as the final rf but bmw didn't choose this configuration by default. Maybe they were not 100% confident in this estimate of the rf in the 2000s.

[karter16]

There is configuration byte for the rf calculation in the csl sw.

You're referring to [Master]0x008e5e4 (0xe5e4 in the xdf) right?

[MpowerE36]

You're referring to [Master]0x008e5e4 (0xe5e4 in the xdf) right?

Yes

[Slideways]

What does RF stand for or what is it?

[bmwfnatic]

What does RF stand for or what is it?

relative filling, amount of air in cylinder.