[Resolved]What Causes CSL Conversion tune to Jerky?

[mushitaro]

Thanks to everyone who contributed information to this community.

After running Terra's CSL Tune on my car it jerks below 3500 rpm when running at full throttle. And if this continues, all pistons will misfire.
When I measured the air-fuel ratio, it shows 10.0 up to 3500rpm.
I know this issue has already been discussed, but for those of you who have DIY'd this, how did you tune it?
https://nam3forum.com/forums/forum/s...2025#post42025

＞[Resolved] Jerky can be improved by lowering the coefficient of the WOT map. This allowed us to adjust the air-fuel ratio to 12.5 at all rpm (generally the most powerful value for NA engines).

I've seen the thread below mention how to deal with this issue.
https://nam3forum.com/forums/forum/s...5420#post75420
However, the heated debate does not seem to have converged.
Can anyone understand and explain the conclusion of this discussion?

What does "simple" mean in the thread below?
https://nam3forum.com/forums/forum/s...5#post42095

This is what I did.
① The fuel injection coefficient of the full throttle map was lowered step by step.
> does not resolve.

② I turned off the catalyst protection control using Martin's tool.
> does not resolve.

My vehicle is January 2004 SMG2(japan).
【Equipment】
・Turner CSL Style Intake Kit
・MSS54HP
・MAP sensor
・Stock M3 cam
・Euro header(Stock)
・Euro Cat(Stock)
・Stock muffler

The cam offset is corrected to stockM3 using Tuner Pro.
Converting the IAT scale to CSL using Martin's tools.
I used testo to confirm that the map sensor is working.

I've attached the BIN with my changes.

[Slideways]

Does it pass a VANOS test?

[mushitaro]

Does it pass a VANOS test?

This is the result of the VANOS test. I think it's within acceptable limits.

Test file : C:\EDIABAS\ECU\MSS54_VANOS_TESTS.TST
Start : 02.07.2023 15:38:51.8
Nr 1( 1) : 02.07.2023 15:38:52.1
apiJob("mss54ds0","STEUERN_EVANOS1_FRUEHANSCHLAG", "","")
JOB_STATUS = OKAY
EVAN_ISTWERT = 0 00 00 ..
EVAN_SOLLWERT = -10 FF FF FF F6 yyyo
EVAN_ISTWERT_EINH = Grad KW
EVAN_SOLLWERT_EINH = Grad KW
Nr 2( 1) : 02.07.2023 15:38:57.6
apiJob("mss54ds0","STEUERN_EVANOS1_SPAETANSCHLAG", "","")
JOB_STATUS = OKAY
EVAN_ISTWERT = 59 00 3B .;
EVAN_SOLLWERT = 75 00 4B .K
EVAN_ISTWERT_EINH = Grad KW
EVAN_SOLLWERT_EINH = Grad KW
Nr 3( 1) : 02.07.2023 15:39:03.0
apiJob("mss54ds0","STEUERN_EVANOS1_VERSTELLZEIT"," ","")
JOB_STATUS = OKAY
EVAN_VERSTELLZEIT_FRUEH = 248 00 F8 .o
EVAN_VERSTELLZEIT_SPAET = 223 00 DF .s
EVAN_VERSTELLZEIT_FRUEH_EINH = ms
EVAN_VERSTELLZEIT_SPAET_EINH = ms
Nr 4( 1) : 02.07.2023 15:39:09.5
apiJob("mss54ds0","STEUERN_EVANOS1_DICHTHEIT",""," ")
JOB_STATUS = OKAY
EVAN_STATUS = EVANOS-1-Einheit vorzeitig weggedriftet
Nr 5( 1) : 02.07.2023 15:39:24.0
apiJob("mss54ds0","STEUERN_AVANOS1_FRUEHANSCHLAG", "","")
JOB_STATUS = OKAY
AVAN_ISTWERT = 0 00 00 ..
AVAN_SOLLWERT = -10 FF FF FF F6 yyyo
AVAN_ISTWERT_EINH = Grad KW
AVAN_SOLLWERT_EINH = Grad KW
Nr 6( 1) : 02.07.2023 15:39:29.5
apiJob("mss54ds0","STEUERN_AVANOS1_SPAETANSCHLAG", "","")
JOB_STATUS = OKAY
AVAN_ISTWERT = 49 00 31 .1
AVAN_SOLLWERT = 55 00 37 .7
AVAN_ISTWERT_EINH = Grad KW
AVAN_SOLLWERT_EINH = Grad KW
Nr 7( 1) : 02.07.2023 15:39:35.0
apiJob("mss54ds0","STEUERN_AVANOS1_VERSTELLZEIT"," ","")
JOB_STATUS = OKAY
AVAN_VERSTELLZEIT_FRUEH = 178 00 B2 .2
AVAN_VERSTELLZEIT_SPAET = 229 00 E5 .a
AVAN_VERSTELLZEIT_FRUEH_EINH = ms
AVAN_VERSTELLZEIT_SPAET_EINH = ms
Nr 8( 1) : 02.07.2023 15:39:41.4
apiJob("mss54ds0","STEUERN_AVANOS1_DICHTHEIT",""," ")
JOB_STATUS = OKAY
AVAN_STATUS = AVANOS-1-Einheit vorzeitig weggedriftet
Stop : 02.07.2023 15:39:56.7
Dauer : 64.9 s

[S54B32]

This happens more intense without the flap that restricts the intake diameter in lower rpms. I recommend to zero out that table that is added for the open flap.
And for jerking, correct it in your VE table, until you match your target AFR you wish. I would also copy changes over to the emergency table.

[Tomba]

Copy VANOS tables from stock to CSL calibration as your camshafts are still stock. Than correct fuel and ignition to stock.

[Bry5on]

Sending from my phone, but I found that in my interpolation map, I had to change RPM targets to get a flat AFR curve due to the added resonances of the airbox and SS headers. I’m running the flap, so don’t copy these numbers, but here’s where I ended up for WOT with multiple pulls overlaid (v5):

[Bry5on]

This happens more intense without the flap that restricts the intake diameter in lower rpms. I recommend to zero out that table that is added for the open flap.
And for jerking, correct it in your VE table, until you match your target AFR you wish. I would also copy changes over to the emergency table.

Which table is specific to the open flap? I didn’t realize there was one..

[mushitaro]

This happens more intense without the flap that restricts the intake diameter in lower rpms. I recommend to zero out that table that is added for the open flap.
And for jerking, correct it in your VE table, until you match your target AFR you wish. I would also copy changes over to the emergency table.

Thanks for the advice. I made some adjustments based on your advice and made a big improvement.

1.Does open flap table mean KF_DKBA_RF_N_HERD/SOFT? When all of these were set to 0, a slight improvement was seen.



2.After adjusting the KF_TI_N_RF_VL (WOT injection) table, the jerks were almost resolved. It showed an AFR around 12.5 with the following injection factor settings. (I can't show you because I haven't been able to integrate the engine speed and AFR..)



My blind spot in this case was that I didn't think the injection factor would need to be this low. What I noticed this time is that the effect of securing the intake air volume of anteater nose (stock CSL) is amazing. Hats off to the BMW M engineers 20 years ago.

[mushitaro]

Copy VANOS tables from stock to CSL calibration as your camshafts are still stock. Than correct fuel and ignition to stock.

Thanks for your comment.
The VANOS table was already in CSL calibration. Did you find any concerns in my VANOS test results?

[mushitaro]

Sending from my phone, but I found that in my interpolation map, I had to change RPM targets to get a flat AFR curve due to the added resonances of the airbox and SS headers. I’m running the flap, so don’t copy these numbers, but here’s where I ended up for WOT with multiple pulls overlaid (v5):

Thank you for such information. It was a necessary information for a beginner like me.

How do you integrate engine speed and air-fuel ratio?
I'm using PLX's SM AFR and I'm having trouble integrating it with Testo. Initially, I chose the PLX in anticipation of using the PLX's Bluetooth integrated device, but later realized that it was no longer available as it was discontinued.. (I should have bought the innovate LM-2 from the beginning..)

I'm looking for ideas on how this can be achieved most easily.
I am a beginner and studying electronic devices and serial communication standards.

[Bry5on]

Thank you for such information. It was a necessary information for a beginner like me.

How do you integrate engine speed and air-fuel ratio?
I'm using PLX's SM AFR and I'm having trouble integrating it with Testo. Initially, I chose the PLX in anticipation of using the PLX's Bluetooth integrated device, but later realized that it was no longer available as it was discontinued.. (I should have bought the innovate LM-2 from the beginning..)

I'm looking for ideas on how this can be achieved most easily.
I am a beginner and studying electronic devices and serial communication standards.

I’m using testo and an innovative LC-2. It’s a little clunky to set up but it works pretty well once it’s initially set up.

[S54B32]

Which table is specific to the open flap? I didn’t realize there was one..

Sorry for late response…there are 3 of the „famous AlphaN“ maps. One is normal operation, one is emergency/limp and one is added if flap is open (not sure if it’s always added when flap is missing, best is to zero out). All these are 20x24, so look out on the public XDF for these 3 maps, I’m using my own xdf but these should be included. I have no notebook with me on vacation, but when I’m home I can look up the addresses if you don’t find them. The map for the flap starts in the lower areas with many zeros, easy to find out which is the correct one to zero completely out when flap is deleted.

[mushitaro]

Sorry for late response…there are 3 of the „famous AlphaN“ maps. One is normal operation, one is emergency/limp and one is added if flap is open (not sure if it’s always added when flap is missing, best is to zero out). All these are 20x24, so look out on the public XDF for these 3 maps, I’m using my own xdf but these should be included. I have no notebook with me on vacation, but when I’m home I can look up the addresses if you don’t find them. The map for the flap starts in the lower areas with many zeros, easy to find out which is the correct one to zero completely out when flap is deleted.

Thank you for contacting us during your vacation.
I would like you to tell me the name of the map when you have time.
According to public XDF (000401PD11) the ”famous AlphaN” is known to be 18*12.



20*24 maps are not found by looking around public complete XDF.

Is what I'm seeing radically different? I'm a little confused.

[S54B32]

This 18*12 is not used per my knowledge to calculate RF in 0401 SW. The 18*12 is the limp mode base map to calculate RF (if Maf is faulty or not present) you would use in non CSL „AlphaN“ tunes. You must use the 20*24 maps in 0401, some time ago I saw a public xdf with these included, but the conversion inside the cells was not correct if I remember correct. But anyway the size and location was correct.

When I’m back I will let you know the map names.

[Bry5on]

This 18*12 is not used per my knowledge to calculate RF in 0401 SW. The 18*12 is the limp mode base map to calculate RF (if Maf is faulty or not present) you would use in non CSL „AlphaN“ tunes. You must use the 20*24 maps in 0401, some time ago I saw a public xdf with these included, but the conversion inside the cells was not correct if I remember correct. But anyway the size and location was correct.

When I’m back I will let you know the map names.

Appreciate all the help so far! I also haven’t been able to find the maps specified in the public XDF as the table sizes aren’t labeled. Looking forward to your guidance on this one. Thanks again