[Objective] Hall position sensors are prone to faults under harsh environments, mechanical vibrations, and electrical stress. In addition, installation deviations can significantly affect the motor’s control accuracy and system stability. To enhance the reliability and fault tolerance of permanent magnet synchronous motor (PMSM) drive systems based on Hall position sensors, this study investigates fault diagnosis methods and fault-tolerant control strategies, aiming to address the limitations of traditional approaches in fault diagnostic efficiency and fault-tolerant control accuracy. [Methods] To overcome the low efficiency and misjudgment issues of traditional feature-sequence-based fault detection methods, a novel fast fault diagnosis method based on pseudo-acceleration variation thresholds was proposed. This method compared the change in pseudo-acceleration method with a preset threshold at the Hall transition signal to rapidly detect faults. Furthermore, an improved fault-tolerant control method that integrated traditional fault-tolerant interpolation method with the fast diagnosis method was proposed to minimize the effect of Hall installation deviation. The method was combined with the designed adaptive notch angle observer to reduce the second-harmonic errors caused by Hall sensor misalignment, thereby improving fault-tolerant performance. [Results] Experimental results showed that the proposed fast fault diagnosis method accurately identified both single-phase and dual-phase Hall sensor faults, significantly reducing diagnostic time and avoiding the delays and misjudgments of traditional detection methods. The improved fault-tolerant interpolation method, combined with the adaptive notch angle observer, accurately estimated rotor position even in the presence of Hall sensor faults, effectively enhancing system stability and control precision. The method maintained excellent control performance particularly under different operating conditions. [Conclusion] The proposed fast fault diagnosis method based on pseudo-acceleration variation thresholds, along with the improved fault-tolerant control strategy, outperforms traditional methods in the case of single-phase Hall sensor faults, dual-phase Hall sensor faults, and under different operating conditions. It provides reliable technical support for the stable operation of motors under Hall sensor faults.