Identification and compensation of a non-contact servowriter positioning system

The paper presents an auto-tuning servo controller for a non-contact servowriter positioning system. Position of the drive arm is obtained through a retroreflector mounted directly on the actuator. The actuator dynamics are identified through an in-situ closed loop system identification routine using maximal length Pseudo Random Binary Sequences. The identified results are used to fine tune the controller consisting of a PI+lag-lead compensator, and a notch filter. The optimized controller parameters are down loaded into a PC mounted servo-axis card to establish closed-loop position control of the arm. A positioning accuracy of 1 /spl mu/inch is demonstrated at the retroreflector. Both simulation and implementation results are provided to validate the system performance.