Stability analysis of a Scanning Tunneling Microscope control system

Control system of a Scanning Tunneling Microscope is analyzed. For the aim of model-based control system analysis and design, open-loop model of the plant is obtained based on the closed-loop system identification tests. We conduct identification tests in the frequency-domain, derive transfer function models of the open-loop system, and discuss the main sources that lead to model uncertainties. We define appropriate closed-loop stability and performance measures, and evaluate performance of a conventional PI controller in the framework of the defined criteria. We show that uncertainties arising from a quantum mechanical property of the tunneling junction, known as the Work Function, can have a destabilizing effect on the control system. This increases the risk of tip-sample crash which is a prevalent operational problem in STM. We experimentally evaluate the presented analysis, and suggest guidelines for PI gain tuning.