Hello,
Just bumping this thread up from the dead because I have only recently started tuning the Ford platform, and after fighting the same torque request drop, I believe I have found the issue... The torque limitation maps that end at 3360rpm are actually slightly larger in the binary data, and as such VCM Editor isn't getting the final row and it's reducing torque request at 3390rpm
I opened the stock file in VCM Editor, changed the maps, loaded it into the car, then pulled the binary with a different tool which then allowed me to open it in my binary editor software (WinOLS) to inspect the maps. Interestingly, as a side note, my logging is reporting a 388Nm torque request when over 3400 (regardless of my driver demand) - not 390Nm - I just changed my scaling to show a 500Nm number where HP Tuners is showing 500Nm, however given this discrepancy, I'm thinking that the actual scaling on the map is different too (out by <1%, meaning a 500Nm torque request is actually only 497Nm... not that it really matters in the scheme of things)
Anyway, here is the map shown in Nm, as imported from the ECU that had been modified by VCM Editor:
Ford_Torque_Error.png
Here is is in ft.lbs (I just changed my map scaling to convert the binary to show ft.lbs equivalent):
Ford_Torque_Error_ftlbs.png
After spending hundreds upon hundreds of hours working specifically on the torque modelling for the Hyundai i30N (I have one of our shop cars up to 535BHP with a completely standard engine), I know how insanely beneficial it is to have correctly modelled torque - and the fact that these maps are wrong means that anyone who is tuning these cars with VCM Editor is either turning off the torque monitoring, or is incorrectly scaling torque/load maps to give higher load requests at lower torque numbers, neither of which are the ideal way to tune any torque based ECU.
I plan on devoting a lot more time into the Ford Focus/Fiesta platforms in the coming months, so anything that I find I'll be happy to share.
Stay tuned!