Extended Digital Programmable Low-Pass Filter for Direct Noise Filtering of Three-Phase Variables in Low-Cost AC Drives

This paper proposes an extended structure of a programmable low-pass filter (PLPF) for direct filtering of three-phase variables in automotive AC drives. The PLPF adjusts its cut-off frequency according to the synchronous frequency of the AC motors. Moreover, it compensates gain attenuation and phase delay of the fundamental component signal. The conventional PLPF can only be applied to variables in the stationary reference frame. In order to obtain filtered three-phase variables, an inverse reference frame transformation is required. However, the extended structure, which includes reference frame transformations, can be directly applied to three-phase variables and remove the noise in the earlier stage before transformation. The extended PLPF is more useful for making signal flows clear and intuitive when organizing an algorithm that directly requires three-phase variables. The extended PLPF requires a moderate computational load corresponding with that of the conventional PLPF because its formula is simplified by use of Kirchhoff’s law. The reduction in the execution time is critical for implementing the filter based on low-cost microcontrollers. Using case studies, dead-time compensations and open-phase fault detection for AC drives show the effectiveness of the extended PLPF.