What I can derive from here, is that the bidirectional control seems to confirm the correlation between duty and fan speed, but what was the ECM commanding the fan to do at the time it "quit working"?

[...trimmed text...]Login to read more.