Proceedings Article | 2 March 2022
KEYWORDS: Mirrors, Microelectromechanical systems, Field programmable gate arrays, Control systems, Computer architecture, Laser applications, Electronic filtering, Temperature metrology, Signal to noise ratio, Ferroelectric materials
The present paper on regards the improvement of the way one resonant MEMS (Micro Electro-Mechanical System) micro-mirror is usually controlled to project the wanted laser beam scanning image/pattern, with the goal of creating a fast-response application able to reach as fast as possible the wanted projection aperture and to maintain it with discrete voltage pulses, minimizing the angle amplitude error and with a look to the power consumption of the architecture. On top of that a self-adapting actuation system was realized, able to comply to any mirror design obtaining, thus, a driving system which no longer requires to be initialized according to the mirror parameters. The previous architecture, in fact, leveraged the well-known resonant frequency of the driven mirror, to set two square-wave signals with the same frequency and shifted by 180 degrees for the two mirror’s electrodes, and with a certain voltage amplitude to reach a specific aperture. This technique, otherwise, fails to ensure a fast initial opening of the mirror to get to the target angle and requires a continuous excitation during its operation time to hold the same oscillation amplitude, even more if we consider the low damping coefficient that characterize these high-Q mirrors and that, despite the benefits, makes the device less stable due to its high responsivity. For these reasons the project was born to overcome these limitations and to create a design based on larger but sporadic pulses, thinking about boosting the initial opening and controlling it without a continuous change in driving amplitude voltage. The paper will detail a first modelling of the architecture, designing with Simulink the features of the new driving system and testing them with the model of a generic resonant mirror, a various Q factor. Once the simulations results were satisfying and some design solutions were identified, will be described a RTL implementation of an architecture on an FPGA (Field Programmable Gate Array), mounted on an custom evaluation board. A discussion of experimental results will follow, to verify experimentally the stability of the architecture designs identified on a board that manages the full control of a mirror, actuating and sensing its oscillation, especially in comparison with the legacy driving mode.
