Maximum Torque per Ampere Control of Brushless DC Motors with Large Winding Time Constant and Hall-Sensor Misalignment

Brushless DC (BLDC) motors with Hall-effect sensors are ubiquitous in industrial, robotics, and mobility applications due to their high efficiency, lower cost, and superior power density. A typical BLDC motor is driven by a voltage-source inverter (VSI) using the common six-step 120º commutation logic, where the Hall sensor signals estimate the rotor position at discrete intervals. However, Hall-sensor misalignment in many cost-effective BLDC motors leads to unbalanced phase currents and increased torque ripple. Additionally, BLDC motors with large winding time constants have a prolonged phase current commutation period that deviates the system from optimal maximum torque-per-Ampere (MTPA) operation. This thesis builds on previous research to propose a new method combining a Hall-sensor filter and dynamic MTPA PI controller for realtime correction of the advance firing angle. The proposed method is verified using detailed machine simulation and implemented on a modern microcontroller. Experimental results demonstrate significant improvement on a typical industrial BLDC motor with substantial Hallsensor misalignment and large winding inductance.