Hi guys, I made two small logs while reving stationary with the instant and average map cycling methods. I also did some research/thinking about the cycle average method...
Correct me if I am wrong a 4 cylinder engine needs two crank rotations for a full cycle which means that at 1000 rpm a cycle takes 1000/2 = 500 cpm, 1 min = 60s, 60/500 = 0.12 seconds per cycle. So the update rate of the map value would be around 0.1 - 0.15s at idle. This delay will show in the logs I took with the avarage sample method and might be the cause of the dip in rpm becuase the engine is getting no extra fuel in that period of time while the air is rushing in.
I tried to componsate for the delay with high AE and a low tps(dot) threshold but it showed no real improvement and the enige would still stall when engagng the clutch (which drops the rpm to 700) and giving it a bit of gas to drive away normally. The engine wont stall when you accelerate with releasing the clutch slowly without any gas because of the steady idle airflow where the map delay doesnt apply.
I Would love to have the map reading on the avarage setting for smoothing readings, on the instant setting Iget great response but too much fluctiation in the map reading that gives a bad tuning accuracy. I would love to hear your guys input about this and maybe a potential fix
. I wonder how ecu's like me221 and megasquirt deal with the map readings
Link to logfiles and tune:
https://drive.google.com/drive/folders/ ... sp=sharing
Log with avarage sampling method:
Log with instantanious sampling method: