Minimizing Scanning Errors in Piezoelectric Stack-Actuated Nanopositioning Platforms

Abstract

Piezoelectric stack-actuated parallel-kinematic nanopositioning platforms are widely used in nanopositioning applications. These platforms have a dominant first resonant mode at relatively low frequencies, typically in the hundreds of hertz. Furthermore, piezoelectric stacks used for actuation have inherent nonlinearities such as hysteresis and creep. These problems result in a typically low-grade positioning performance. Closed-loop control algorithms have shown the potential to eliminate these problems and achieve robust, repeatable nanopositioning. Using closed-loop noise profile as a performance criterion, three commonly used damping controllers, positive position feedback, polynomial-based pole placement, and resonant control are compared for their suitability in nanopositioning applications. The polynomial-based pole placement controller is chosen as the most suitable of the three. Consequently, the polynomial-based control design to damp the resonant mode of the platform is combined with an integrator to produce raster scans of large areas. A scanning resolution of approximately 8 nm, over a 100 mum times 100 mum area is achieved.