Findings on Knock removal with SEPST

[jty]

For background, I have been tinkering a custom electronic speedo and reverse engineered the details of the Harley Delphi J1850 protocol.
When street tuning with SEPST, the speedo displays TPS/RPM table hits as color hue in real time and when knock retard occurs, mark the cell red.
So, I got retard hits when opening throttle around 1300-1400 rpm, map 80-100.
However, when looking at the VCI log in SEPST Toolbox->Date, it shows way higher RPMs, like 3000.

My speedo logs the exact J1850 data that ECM reports and VCI records. When I analyse the logs, the data packet that tells current RPM, MAP and TPS% comes exactly 16ms after the knock indicator packet, thus the MAP and RPM around 1300-1400 are very exact and the one in the SEPST log is way off.

member ToBeFrank, I wonder if you have seen this?

[rigidthumper]

Not Frank, to be sure, but I believe it's because the SEPST views transient issues, such as things occurring during DE/AE activation, as irrelevant/not useful for VE table correction. It's really useful for steady state, closed loop, repetitive data.

[jty]

Well, turns out that I did not identify the knock packet correctly. Sorry for the confusion.