MAP sense algorithm.
Posted: Thu Feb 06, 2014 2:55 am
Thought I'd post something I'd mentioned elsewhere.
During first cranking, i.e. after a reset / power outage or whatever, over two known 'complete' revolutions, the ECU samples the map sensor at every tooth detect event on the flywheel. It then determines where the MAP reading is highest, as in most vacuum and stores that crankshaft angle. It then sets a "window" of 15-20 degrees before and after that crankshaft position where it performs a sample and smooth function of those 30-40 samples to arrive at a close approximation of what the ECU should use for calculations. These windows would be stored in eeprom for future reference so this would only have to be done once, or if the ECU detects the values are out of the range previously set. Following the initial setup, the ECU would not perform this test again, unless it is hard reset, as in after a firmware upgrade or forced by the user.
I haven't thought much about it since then, but it seemed plausible at the time, and in context.
During first cranking, i.e. after a reset / power outage or whatever, over two known 'complete' revolutions, the ECU samples the map sensor at every tooth detect event on the flywheel. It then determines where the MAP reading is highest, as in most vacuum and stores that crankshaft angle. It then sets a "window" of 15-20 degrees before and after that crankshaft position where it performs a sample and smooth function of those 30-40 samples to arrive at a close approximation of what the ECU should use for calculations. These windows would be stored in eeprom for future reference so this would only have to be done once, or if the ECU detects the values are out of the range previously set. Following the initial setup, the ECU would not perform this test again, unless it is hard reset, as in after a firmware upgrade or forced by the user.
I haven't thought much about it since then, but it seemed plausible at the time, and in context.
