From a controls point of view, micro electromechanical systems (MEMS) can be driven in an open-loop and closed-loop fashion. Commonly, these devices are driven open-loop by applying simple input signals. If these input signals become more complex by being derived from the system dynamics, we call such control techniques pre-shaped open-loop driving. The ultimate step for improving precision and speed of response is the introduction of feedback, e.g. closed-loop control. Unlike macro mechanical systems, where the implementation of the feedback is relatively simple, in the MEMS case the feedback design is quite problematic, due to the limited availability of sensor data, the presence of sensor dynamics and noise, and the typically fast actuator dynamics. Furthermore, a performance comparison between open-loop and closed-loop control strategies has not been properly explored for MEMS devices. The purpose of this paper is to present experimental results obtained using both open-and closed-loop strategies and to address the comparative issues of driving and control for MEMS devices. An optical MEMS switching device is used for this study. Based on these experimental results, as well as computer simulations, we point out advantages and disadvantages of the different control strategies, address the problems that distinguish MEMS driving systems from their macro counterparts, and discuss criteria to choose a suitable control driving strategy.

No abstract

Comb drives inherently suffer from electromechanical instability called lateral pull-in, side pull-in or, sometimes, lateral instability. Although fabricated to be perfectly symmetrical, the actuator's comb structure is always unbalanced, causing adjacent finger electrodes to contact each other when voltage-deflection conditions are favorable. Lateral instability decreases the active traveling range of the actuator, and the problem is typically approached by improving the mechanical design of the suspension. In this paper, a novel approach to counteracting the pull-in phenomenon is proposed. It is shown that the pull-in problem can be successfully counteracted by introducing active feedback steering of the lateral motion. In order to do this, however, the actuator must be controllable in the lateral direction, and lateral deflection measurements need to be available. It is shown herein how to accomplish this. The experimentally verified dynamic model of the comb drive is extended with a lateral motion model. The lateral part of the model is verified through experimental results and finite element analysis and is hypothetically extended to accommodate both sensor and actuator functionalities for lateral movement. A set of simulations is performed to illustrate the improved traveling range gained by the controller.

No abstract

No abstract