[cjperformance]

In the meantime, while i sort out the belt issue... something i came up with to help with MAF adjustments:

I found out the reflex log contains a lot more data than what you choose to put on the graph. If you export the log from the reflex to a csv file, you can selectively export MAF values paired with Lambda Corrected % values. Using some spreadsheet magic i was able to sort and average the correction values by the MAF x axis values defined in both the E92 and Reflex. It spit out a set of percent corrected by MAF x axis increments very close to what youd see on a MAF EQ Error Ratio chart in HPT.

Interestingly, the values it spit out for average adjustment at specific MAF ranges were very close to my custom PI-MAF table in HPT... which is essentially an EQ Error % table, but hardcoded to get the pct difference between actual lambda and my desired .82 lambda. Needs to be tweaked still to discard non-WOT data, but i can manually copy the ranges I need. Since commanded lambda in the E92 is skewed due to pulling PE to supplent the MAF for tq mgmt reasons, i can no longer use commanded EQR vs Actual EQR, thus the hardcoding to .82.

I think the fact that the wideband has to read downstream before the reflex can make adjustments still allows the EQ Error Ratio process to work. Ill have to try few iterations of adjustments this way and see if it allows things to get dialed in in a reasonable way, and see if the correlation holds as i get closer to ideal.

I did my last round of adjustments this way, but the belt issue really invalidated any data i collected.