The trigger should be off of the command signal for an accurate measurement, not off of the
position signal. The step response measurement should start at the trigger point, and stop
when the system is within 99% of its final position. Another option would be to create a mask
on the oscilloscope and compare that to the system you are testing.