Reinventing the wheel: Fuel Trims on 2016 FI 5.0L Mustang

[Ranger Coyote]

I have read every single post about techniques to correct fuel trims and following the instructions EXACTLY, I was not getting positive results. I came up with a completely new way to correct the MAF and I'm not sure if I just got lucky or what but I FINALLY got it!!! For best results only log the minimum amount of items and make sure the polling interval is 10ms on PIDs "Mass Airflow", "Short Term Fuel Trim Bank 1", "Short Term Fuel Trim Bank 2", "Mass Airflow Period"

#1 Disable LTFT in the tune
#2 Create a new math and Graph in the VCM Scanner: MAFSTFT.xml and MAF.Table.xml
Use the mass air period values from your tune in the graph - row axis values.
#3 Get a good long log after the car is 100% warmed up and been running for a several minutes. Copy the values that have a good number of hit counts from the graph and "Paste Special-Average" directly to your MAF vs Period table. This took me around 7 applications to get my error less than 1%. Before you flash it, go in to your MAF 2D Chart and make smooth out any parts of the curve that don't look right.
maf2.jpg
maf.jpg
mathgraph.jpg

I hope this helps other people as I fought with my fuel trims FOREVER!!! If anyone sees a problem with the information I posted please let me know and I will remove/revise it as to not lead people wrong.

[blackbolt22]

I'm confused because this looks like the normal way of doing it. Which part is re-invented? It looks like what I've always done for Coyotes...

Though I do 'multiply by % half' instead of average and this gets me very close after 1 adjustment.

[Ranger Coyote]

The normal method is to log MAF Period and populate it with your average trim error, then multiply % -half to the MAF vs Period table. In this method you are applying trim errors to the Mass Airflow during logging and getting "corrected" MAF vs Period values instead of a % error. The end result is more or less the same but getting there in a slightly different way. I had no luck what-so-ever attempting to apply trim errors to my curve like everyone else seems to be successful doing. What I was seeing is something like 12% rich and when I would apply -6% to that period the next time I would log it would be way lean... I have no clue why I could not make any ground by using the normal method.

[blackbolt22]

Hah that's an interesting way of doing it. Thanks for the clarification. I've seen the swing in the wrong direction when multiplying by 100% instead of 50% .

I might try it out and see if I can fine tune my MAF Temp compensation table which I still need to do. She's super rich at start and for a while until everything is at full normal temp. Then it's near perfect

[Ranger Coyote]

I zero out my MAF temp comp while messing with the curve then just put back the Roush values when I'm done. I'm 100% guessing but you might want to consider checking the Rail Temp modifier and slope in your injector settings. In the ID1000 return fuel cheat sheet they have you put 1's in the Offset/Rail Temp modifier.

[Grim5.0]

something ive come to notice, it helps to build walls or boundaries around the data points you want to be accurate, and not averages from the gap in between to the next data point. if at 615uS we want to know the trim error or maf averaged, try putting a data point at 600 and 630us in the scanner axis label. dont copy this into the maf curve, but now you have reduced the averaging range from which the scanner is saying 615 = xx% errror etc.. youll have to manually adjust the maf curve doing this. or any table for that matter

for the maf curve, this will be a pain due to the number of data points. using the idle tb airflow/tb angle table as an example, some areas change quite a bit and some a small amount. but we are into splitting hairs here arent we?

no walls.pngwalls around wanted points.png