It looks like I messed something up in my graphs when trying to recreate the 2.24 lambda histograms in v4.x. I have the old NGK AFX (not the Ballenger variant) sensor, with a brand new sensor as of a couple of weeks ago that was calibrated in free air for 40 minutes before installation, and a brand new LS3 MAF sensor just a few days old.

In my channel list, I have these related PIDs:
MPVI1.1 -> NGK - AFX [Input / 10.41667)+0.62 = Output, unit = lambda)
Equivalence Ratio Commanded

I had been using a custom Math-User parameter, "Lambda Err", to populate the VE/MAF graphs in the scanner. When I edit that parameter, both variables (40160 is NGK - AFX Transform, and 50118 is EQ Ratio Commanded) have a white background and the formula is:

100*([40160.238]-(1/[50118.238]))/(1/[50118.238])

When I use that custom Lambda Err in the VE/MAF error graphs, it's not working well in the upper RPMs and the shape at the upper RPMs in both MAF & VE does not look normal.

Re-checking my scanner setup today, I built a second graph to try to use the built-in "EQ Ratio Error" .

In a recent MAF log, in MAF Frequency Freq cells above 5600:
* the "EQ Ratio Error" says I'm up to 15% lean.
* the custom "Lambda Err" parameter, I'm generally within 5% error

In a recent VE log, the "Lambda Err" says I'm very rich at WOT, but the "EQ Ratio Error" says I'm lean.

I'm hoping someone can verify the needed corrections. Thanks!