Abstract
Although continuous-time (CT) analyzes offer valuable theoretical insights into improving robustness, they often fail to incorporate critical sampling effects and other discrete-time (DT) dynamics essential for practical implementations. Consequently, DT analyzes are required to comprehensively evaluate how various discretization approaches influence system performance and robustness in servo-drive applications. This study investigates a discrete disturbance observer (DOB) framework for servo-drive systems, examining both differentiator- and estimator-based DOB designs implemented via Euler and Tustin discretization methods. By analyzing the impact of these designs on sensitivity peaks, noise sensitivity, and disturbance rejection, the study derives design constraints for the nominal plant model and observer bandwidth using DT sensitivity analyses and the Bode integral theorem. Theoretical evaluations and experimental results consistently demonstrate that the Tustin method, particularly when applied to differentiator-based DOB, delivers better disturbance attenuation and control performance. This systematic approach not only provides practical guidelines for tuning DOB parameters, but also assists in choosing discretization techniques most suitable for optimizing practical servo drive systems in real-world applications.