How to master permanent magnet brushless DC motor control?

The permanent magnet brushless DC motor adheres the permanent magnet to the surface of the rotor core and adopts the concentrated winding method to form a hidden pole rotor structure. The main magnetic field generated by the rotor permanent magnet is close to a trapezoidal wave. When the rotor rotates at a constant speed, the stator winding is cut by the main magnetic field. The waveform of the induced potential in each phase of the stator winding is basically consistent with the main magnetic field. To simplify the analysis, it can be approximately regarded as a trapezoidal wave with a flat top width of 120° electrical angle, as shown in the figure below. In order to output constant electromagnetic power or torque, the three-phase stator winding must add six-step trapezoidal wave or square wave current:

Normally, permanent magnet brushless DC motors have three position sensors (usually halls) fixed on the stator to detect the magnetic pole position of the rotor relative to the stator. The figure shows the position signal detected by the hall signal. In the figure, the hall is installed at 120°:

When the motor is running, the three hall signals are sent to the MCU after some processing. The MCU then drives the on and off of the switch devices in the three-phase full-bridge inverter in a certain order according to the obtained hall position information, so that the stator current of the six-step trapezoidal wave can be obtained, so that the stator winding generates a rotating magnetic field. The average rotation speed of the stator magnetic field is synchronized with the speed of the rotor permanent magnet, so that an effective electromagnetic torque will be generated and the rotor will rotate continuously. It can also be seen from the principle that the running commutation moment of the brushless DC motor is closely related to the rotor position. It can be seen that how to control the hall signal and the position information transmission of the MCU to drive the on and off of the switch devices in the inverter is the key to the control of the permanent magnet brushless DC motor with hall.

However, since the installation of position sensors will increase the extra cost and volume, and will be affected by the ambient temperature and humidity, the control method without position sensors (SensorLess) is becoming more and more popular. For the control of permanent magnet brushless DC motors without position sensors, the back-EMF detection method is currently widely used. However, the back-EMF method has a fatal flaw, that is, it cannot accurately obtain the relative position of the rotor in the motor starting state and the low-speed state with very small back-EMF, so it is impossible to accurately obtain the relative position of the rotor, so the starting of permanent magnet brushless DC motors without position sensors has always been a difficult point.

The most widely used permanent magnet brushless DC motor position sensorless control method is the traditional three-stage method: first, give the inverter a fixed drive state, pull the stator winding to a fixed position, and then use forced commutation to perform open-loop acceleration. When the acceleration reaches the point where the back EMF can be accurately obtained, switch to the closed-loop operation stage that uses the back EMF to estimate the position. The three-stage method is suitable for operating occasions with small moment of inertia and light loads, but not for heavy loads. Long-term open loop operation will cause the motor to lose step, the stator current will increase, and the motor will not be able to operate correctly and effectively (such as vacuum cleaners, cars, fans, etc.).

So how can we better realize the control of position sensorless permanent magnet brushless DC motor?