MAF Calibration: huge difference between old method and new methods

[marcoturbo]

Hi,

Looking to improve my own car (stock C6 ZR1), I've built several histograms in order to calibrate the MAF and notice a huge difference in the errors reported between the "old" method and the "new" methods (see smokeshow's and cringer's topics) doing MAF and VVE in the same time.

I've checked the maths formulas at several times and hope - as a noob - there's no mistake.

Any ideas or suggestions ?

The "old" method: errors are not very important (in my view...) and positive



The "new" methods. I know they differ between them but the errors are very important and negative !





Thanks !

[horsepowerguru]

i didn't have very good luck trying to tune both VVE and MAF at the same time. there's too many variables to filter IMO. I went OL w/ no LTFT and tuned MAF using STFT and had great luck with it. not talking about WOT of course....

[smokeshow]

GMVE has no business informing MAF calibration error...so not sure what that 'GMVE LTFT+STFT' graph is doing. If it involves GMVE at all, the math parameter cannot be correct.

[marcoturbo]

GMVE has no business informing MAF calibration error...so not sure what that 'GMVE LTFT+STFT' graph is doing. If it involves GMVE at all, the math parameter cannot be correct.

I'll check tonight.

[marcoturbo]

Here is the math:

(([16.71.avg(200)]*(1+(.01*([6.156.avg(200)]+[7.156.avg(200)])+.01*([8.156.avg(200)]+[9.156.avg(200)]))/2)-[50070.56.avg(200)]/60*4*[11.92.avg(200)]*[2312.avg(200)]/[2126.240.avg(200)])/([50070.56.avg(200)]/60*4*[11.92.avg(200)]*[2312.avg(200)]/[2126.240.avg(200)]))*100

I don't remember where I found it but it was probably here

[Alvin]

You might have better results calibrating VVE and MAF completely independent of each other.

This is my preferred method.