It turns out that while preparing my
motor power requirements graph, I didn't correctly strip data recorded while the motor was accelerating or decelerating. I've fixed this and updated my previous blog to report a real reduction in power used after running the motor in.