Electronic Thesis and Dissertation Repository

Degree

Master of Engineering Science

Program

Electrical and Computer Engineering

Supervisor

Dr. Lyndon J. Brown

Abstract

Permanent magnet BLDC motors are characterized by a central magnetic core, called the rotor, and fixed electric coils (usually six) equally spaced in a ring around the core, called the stator. Motor movement is controlled by alternately energizing and de-energizing the stator coils to create a rotating magnetic field that propels the rotor. In order for this process to work correctly, BLDC motors required a technology called electronic commutation, in which the coil currents must be very carefully synchronized to rotor position to ensure that the rotating field is correctly aligned with the permanent magnetic field in the rotor. Usually rotor position is measured by external sensors such as Hall-effect sensors and optical encoders and these external sensors increase the system cost and reduces reliability. In order to control the price and make it more reliable this thesis propose to infer the rotor position from voltage and current measurement of motor.

The most common approaches to sensorless control are based on the measurement of the electromotive force (back-EMF), that is induced by the rotor motion. As the back-EMF is nearly zero at very low speed and at stationary position, and can not be measured. Therefore a separate algorithm is required for start-up and control at low speed. The other method of sensorless control involves the inference of rotor position from the variation in inductance caused by rotor position. This thesis presents a prototype system for sensorless control of BLDC motors over the entire speed range of the motor, including stall (zero speed) conditions using the voltage and current signals from the motor.


Share

COinS