Optimizing the feedback control of Galvo scanners for laser manufacturing systems

Abstract

This paper summarizes the factors that limit the performance of moving-magnet galvo scanners driven by closed-loop digital servo amplifiers: torsional resonances, drifts, nonlinearities, feedback noise and friction.Then it describes a detailed Simulink ® simulator that takes into account these factors and can be used to automatically tune the controller for best results with given galvo type and trajectory patterns. It allows for rapid testing of different control schemes, for instance combined position/velocity PID loops and displays the corresponding output in terms of torque, angular position and feedback sensor signal. The tool is configurable and can either use a dynamical state-space model of galvo's open-loop response, or can import the experimentally measured frequency domain transfer function.Next a drive signal digital pre-filtering technique is discussed. By performing a real-time Fourier analysis of the raw command signal it can be pre-warped to minimize all harmonics around the torsional resonances while boosting other non-resonant high frequencies. The optimized waveform results in much smaller overshoot and better settling time. Similar performance gain cannot be extracted from the servo controller alone.

Galvanometer scanner modeling for Selective Laser Melting deflection system simulation

“…The galvanometer motor has many application fields as: optical scanning for microscopy [12], precise laser drilling [3], [10], laser marking [5], laser trimming [11]. A first level of modeling considers the galvanometer motor as a rigid body, either in its simplest form as a second-order transfer function [5] or as a third order transfer function as a Direct Current (DC) motor with the two mechanical and electrical time constants [3], [12].…”

“…A first level of modeling considers the galvanometer motor as a rigid body, either in its simplest form as a second-order transfer function [5] or as a third order transfer function as a Direct Current (DC) motor with the two mechanical and electrical time constants [3], [12]. A second level of modeling exists which takes into account the torsion elasticity phenomenon at the mechanical link between the motor and its driven load [10], [11]. The proposed galvanometer modeling in this paper depicts a new way to approach the flexible torsion modes.…”

A tutorial on loop-shaping control methodologies for precision positioning systems

“…In the electrical model ( Figure 13) of a generalized galvo scanner platform, the open-loop transfer function 26,27 relating the input drive current I(s) to the output position Y(s) is…”

Generic software tool for counteracting the dynamics effects of optical beam delivery systems

“…In particular, the proposed approaches apply ‘a drive signal digital pre-filtering’ technique to improve the dynamics performance of optical beam deflection systems by performing a real-time Fourier analysis of the raw command signals. 16–18 However, this technique is not sufficiently effective when it is required micro-engineering components to be produced with high ARR, 3,4 and therefore, other MMPs have to be employed together with specially developed software tools to compensate process limitations. 19,20 Such software tool for layer-based micro-machining was also reported to improve the resulting surface topography following laser-milling 21 by optimising the slicing procedure and vector orientations in each layer for 3D geometries.…”

“…17 Other factors, which limit the performance of optical beam deflection systems, include torsional resonance, heat dissipation, drift, nonlinearities and noise. 18 The dynamics effects due to the system’s inertia increase with the increase in the beam deflectors’ rotary speed that ultimately affect the machining results. The dynamics effects have a direct impact on the resulting machining accuracy and quality and they are discussed in the two sub-sections below.…”