Phase-Locked Loop Bandwidth Design and Partial Compensation Strategy for PMSM Position Estimation Based on Embedded Linear Hall Sensors

Embedded magnetic encoders (EMEs) using linear Hall sensors offer a compact and low-cost solution for rotor position estimation in permanent magnet synchronous motor (PMSM) systems. However, the accuracy of EMEs is compromised by sensor non-idealities and magnetic field harmonics. The phase-locked loop (PLL) is commonly used to filter these disturbances, but selecting its bandwidth involves an inherent trade-off between harmonic suppression and dynamic tracking performance, which is critical in spacecraft applications such as the gyrowheel. To address this conflict, this paper develops a frequency-domain analytical framework that unifies steady-state harmonic-induced errors and dynamic tracking errors into a single total error function. The optimal PLL bandwidth is derived by minimizing this function, and a partial compensation strategy based on prior information is further proposed to reduce the total error. Theoretical analysis demonstrates that the proposed compensation mechanism shifts the optimal bandwidth toward higher frequencies and reduces the lower bound of the achievable total position estimation error. Experimental results validate the opposite dependence of the PLL method’s dynamic and static errors. On this basis, the partial compensation strategy achieves a 71% accuracy improvement at the shifted optimal bandwidth under the dynamic operating conditions.