N55e Selftuning assistance

Sava666

Corporal
Dec 14, 2020
174
36
0
Good afternoon everyone.

I came from the N54 world, have a little experience in tuning, already touched hybrids and ST and so far all successful and easy because MHD+. Now I'm tuning a friend's car on N55e FBO PS2 + new OEM boost solenoid (yes, it arrived from America). Had to get involved because in my country there are too many scammers (JDM tuners) who just charge money and do nothing, pretending to charge for time and dyno, lol. Please don't criticize and I'm not going to respected tuners like Wedge or Twisted, I'll be honest, money is tight, I don't live in Europe or the States and 500 bucks is a lot of money for me. And it's DIY =) Already read the threads from Sbrach, Anjuna and another guy on E90post. Also have Cstokes XDF and the latest from MHD on github. So @JPuehl @carabuser @Sbrach @houtan @wheela share some knowledge be so kind to help me. English is not my first language.

1. Does it make sense to go straight to Commanded WG? And not to beat my head into stock BMW model.

2. Attached is my .bin and log (rev1).

3. I also have a few questions about the maps below. If you can, please tell me if I am doing everything right and what I need to watch out for.

WGDC limit - 80%.
All PIDs set to 0 including ceilings/floors
WG Position fur Vorsteuerung B1/B2 - stock
WG Position Model - Stock
Revised compressor curve / turbine power - you can see it in a file.

ATL Regler - What are these 3 for?
Additional WG due to spring - 0ed.
Boost multiplier (Kennfeld zum Invertieren von F_pssolber und Wichten über der Motordrehzahl (KF_PSRF_GRD) - should I revise this table too?
Boost multiplier - set to 1 after 4000rpm
Boost set limit - 4. Should I revise the X-axis? it says 50 all over the end (kg/h)
Target pressure ratios (Norm/Dyn) set to 1.
Target boost offset in Sport - Was left stock but now I want to try with 0. Is this the reason my target dips in the middle?

Now for the torque limiting maps.

Torque Eff Divisor - Stock
Load to torque all 3 - Stock
TR temp - stock
TR RPM - set to 1.
Load Diff - set to 65
Load limiting factor - set to 1
BLM - should I revise axises up to 420maf/200 load?
Load limiting factor - set to 6. Why? %/hpa
Load target offset - set to 1.
Modeled torques - set to 900
Torque demand ceilings - 900
Toggles - all deactivated
KL_MDMAX - set to 900

 

Attachments

  • N55_v1_98.bin
    4 MB · Views: 36
  • Like
Reactions: wheela

Sava666

Corporal
Dec 14, 2020
174
36
0
Shall I use "Boost Setpoint Factor Commanded" channel instead of the one in the log and rescale the whole setpoint up to 25psi?
 

wheela

Captain
Jun 4, 2021
1,170
660
0
Twin Cities, MN
Ride
2015 e84 X1 35i Msport
I see load actual is continually dropping until this point, where wgdc% spikes up. That seems very odd, I'd take note of rpm, maf, and load at that point, and check your tables for any abrupt changes in those areas that may be responsible.

Also, that flat load request across the whole pull seems a little curious? I think something is maxing out to cause that? You'll probably need to extend the break points for load and maf on your tables at some point with that ps2. I made a spreadsheet that does a pretty good job of converting back and forth between load and maf vs rpm, it may be helpful. I'll upload it here when I get a chance👍
 

Attachments

  • Screenshot_20230922_095330_Chrome.jpg
    Screenshot_20230922_095330_Chrome.jpg
    110.1 KB · Views: 78

Sava666

Corporal
Dec 14, 2020
174
36
0
I see load actual is continually dropping until this point, where wgdc% spikes up. That seems very odd, I'd take note of rpm, maf, and load at that point, and check your tables for any abrupt changes in those areas that may be responsible.

Also, that flat load request across the whole pull seems a little curious? I think something is maxing out to cause that? You'll probably need to extend the break points for load and maf on your tables at some point with that ps2. I made a spreadsheet that does a pretty good job of converting back and forth between load and maf vs rpm, it may be helpful. I'll upload it here when I get a chance👍
Yes, I have seen it. But I started from scratch, just removed limiters, stock rom from the github, load ceiling at 0 tim. corr set to 180. That's why I made it flat. Raised the high parts in the Compressor table. Adjusted all Load related tables, 3.5 tmap adjustments.

Figuring out exactly which logging channels to monitor because there are a lot of them.
 

wheela

Captain
Jun 4, 2021
1,170
660
0
Twin Cities, MN
Ride
2015 e84 X1 35i Msport
When you increased values in the high parts of the compressor table, did you try to blend them in? If it was an abrupt change, that may help explain the sudden jump in wgdc.
You said you zeroed PID's, assuming that's temporary to see what effect these other tables are having?
 

wheela

Captain
Jun 4, 2021
1,170
660
0
Twin Cities, MN
Ride
2015 e84 X1 35i Msport
Just thinking- you may want to bring wgdc due to spring back to stock values since none of the other wastegate position models were adjusted.
 
  • Like
Reactions: Sava666

wheela

Captain
Jun 4, 2021
1,170
660
0
Twin Cities, MN
Ride
2015 e84 X1 35i Msport
My current thinking on how n55e boost control works is roughly like this:

The dme has a target load, it calculates the boost and airflow needed to hit that load at the current engine conditions.

It uses compressor characteristics to lookup how much turbine power is needed for the turbo to produce that boost and airflow at the given engine conditions (turbine power is roughly pressure and temp drop across turbine × exhaust mass flow rate through the turbine)

Then it figures out what percentage of exhaust flow needs to go through the turbine (vs. the wastegate) to hit the target power at the current engine conditions.

Then it looks up what wastegate position is needed to hit that exhaust percentage through the turbine at the given engine conditions.

Then it looks up what wgdc% is needed to hit the targeted wastegate position at the current engine conditions.

Then it applies any "adders" etc.

Then it looks at actuals vs. targets and applies PID (based on turbine power) to close the gaps.
 
  • Like
Reactions: WhyAnN20

wheela

Captain
Jun 4, 2021
1,170
660
0
Twin Cities, MN
Ride
2015 e84 X1 35i Msport
I think this is why the dme has an exhaust gas temperature model - it needs to know exhaust temp upstream of the turbine to calculate turbine power. Since it has the exhaust temp model, it also has the component protection mode if calculated exhaust gas temp gets too high.
 
  • Like
Reactions: WhyAnN20

wheela

Captain
Jun 4, 2021
1,170
660
0
Twin Cities, MN
Ride
2015 e84 X1 35i Msport
Same for all the pressure drop models across various intake and exhaust components, it needs these to determine actual pressure at the compressor inlet (for calculating pressure ratio) and the pressures & temps upstream and downstream of the turbine (along with mass flow through the turbine) to calculate turbine power.
 
  • Like
Reactions: WhyAnN20

wheela

Captain
Jun 4, 2021
1,170
660
0
Twin Cities, MN
Ride
2015 e84 X1 35i Msport
About commanded wg, I believe that approach is easier, but my understanding is that PID is not fully active with commanded, so my preference is to beat my head against the bmw stock model😁
Not sure if you've come across this thread yet - it's pretty old, but since you're interested in commanded wastegate, they get into that quite a bit here. It's been a whole since I read through it, but I'm going through it again to see what else I can pick up:

 

JPuehl

Corporal
Jul 24, 2019
105
142
0
Greater Seattle area (a.k.a. Zombie Land)
Ride
2011 135i 1M clone N55e DCT
  • Like
Reactions: wheela

Phack79

New Member
Sep 28, 2023
6
1
0
Can you share a log showing the overboost? Is it only right after shift? The table you reference IS in the 98G0B BIN/XDF but I'm not sure for the F20 N55 EWG, but I'll take a look.
You can check on log. When the 6th gear engaged the WGDC pops up to 97% from 90% and got overboost up to 2bar (requested 1.75bar ) The Compressor base KW is not altered. So that cause additional WGDC maps out of PIDs



1705457064515.png