Self-learning current optimizing control for conventional stepping motor drive technique based on step pulses

Abstract

The essential advantage of the conventional stepping motor drive technique bases on step command pulses is the ability of open-loop positioning. By ruling out the cost of a position sensor, stepping motors are preferred in low power positioning applications. However, machine developers also want to obtain high dynamics with these small and cheap stepping motors. For that reason, stepping motors are used at its limits as much as possible. A drawback of the open-loop control is the continuous risk of missing a step due to overload. Due to this uncertainty, robustness is a major issue in stepping motor applications. Until today, to reduce the possibility of step loss, the motor is typically driven at maximum current level or is over-dimensioned with results in low-efficiency. Therefore in this paper, a self-learning $\cos \left(\varphi \right)$ -controller optimizing the current is presented. Moreover, to allow broad industrial applicability, this technique is computationally simple, needs no mechanical or electrical parameter knowledge and take into account the unique character of stepping motors and their conventional drive technique based on step command pulses. The proposed algorithm is validated through measurements on a hybrid stepping motor.