Brushless DC motors

2. Motor control/commutation

The BLDC motor rotates because the permanent magnet on the rotor attempts to align itself in the direction of the magnetic field generated by the electromagnets of the stator. While doing this, the torque is at a maximum when both magnetic fields are perpendicular to one another. The type of control is distinguished in two ways: in the form of the power supply of the windings (block-shaped or sinusoidal), or in the methods for identifying the rotor position. It is necessary to determine the rotor position because the currents in the windings must be timed to switch in such a way that the stator magnetic field is always perpendicular to the rotor magnetic field, i.e. it continues turning with the desired speed.

2.2 Sine commutation

The optimum form of energization is the sine commutation, whereby each winding of the motor is powered with a sine wave shifted by 120°, resulting in a continuously rotating stator magnetic field with constant strength. As a rule, if only Hall sensors are available for determining the rotor position, the sine commutation can also be used by interpolating between the switching points. In most cases, this directly leads to a significant improvement of the performance of the motor. In case of load changes between two Hall sensors, the sine wave cannot be adjusted, however, resulting in incorrect positioning of the magnetic field. This can only be corrected with the next Hall sensor signal.

The sine commutation therefore ideally requires a higher resolution system for determining the rotor position. Normally, this consists of an optical or magnetic encoder that determines the position of the rotor with sufficient precision at all times and correspondingly adjusts the current.

2.3 Field-Oriented Control

Sometimes field oriented control is still distinguished from sine commutation; often, however, both terms are used interchangeably. The terms are used interchangeably because the sine commutation, as described above, optimally controls the stator magnetic field. 

As long as one leaves out how the current controller itself functions, the difference is not apparent in the description of the sine commutation. In the preceding it was assumed that a sinusoidal current value would be generated which, with sufficient speed, would be inputted into the winding by the current controller. Both tasks, the current value (which corresponds to the torque control loop) and the current control of the windings, are thus treated separately in the sine commutation or, respectively, are performed by separate controllers. In this case, however, the current controllers for the windings receive with increasing speed a current value that changes with higher frequency. At the same time, the increasingly strong influence of the back-EMF from the motor must be compensated for. Because the bandwidth of the current controller has an upper limit, at higher speeds there can be phase shifts and distortions of the current flows so that the stator magnetic field is no longer perpendicular to the rotor. 

The field oriented control solves this problem by controlling the current vector directly in the rotating coordinate system of the rotor. For this, the measured current flows of the three phases are transformed through a Clark-Park transformation into the biaxial coordinate system of the rotor. Thus, the torque value will no longer be first converted into current values, as is the case with the sine commutation for the individual windings, which are then each separately controlled; but instead it is simultaneously controlled in a coordinate system of the rotor current level and orientation of the magnetic field. Subsequently, the current flows for the individual windings are calculated (through an inverse Clark-Park transformation). With this method the control is frequency-independent and, even at higher rotational speeds, will always produce optimum sinusoidal current.

2.4 Sensorless Control

Sensorless control is not an additional control method, but rather a term for methods which are capable of determining the rotor position without sensors (i.e. Hall sensors, encoders). These methods can be roughly broken down into two classes:

Simple sensorless control is based on the direct measurement of the back-EMF in the respective non-energized winding. Compared to standard control, however, this method requires special hardware and is instable below approx. 20% of the rated speed of the motor as the measurement signal is too small. Furthermore, this method only functions in combination with block commutation, because with sine commutation, all three coils are always simultaneously energized.

More sophisticated solutions are based on so-called "observer tracking", which reproduces values that are not directly measurable, such as the speed or back-EMF, from other values measured by the current controller. The core of this type of system is an extremely precise model of the motor that, in parallel with the actual motor, calculates from the known input values, such as the set PWM, those values which are also measured, such as the current level in the winding. The calculated values are then compared to the measured values in every cycle. Due to the observation error determined by using this method, the internal values of the motor model are constantly readjusted. With this method a more correct estimate is also obtained for the values that are not actually measured, such as the speed. Although this method only works because the reaction of the winding changes as a function of the speed due to the induced voltage, the directly measured values can be readily measured even at low speeds. The result is a "virtual encoder" that delivers the position and speed information, beginning at a certain minimum speed, with the same precision as an actual optical or magnetic encoder. As with this method, the back-EMF does not have to be directly measured, it can also be combined with sine commutation or the field oriented control.

What both sensorless methods have in common, is that no information on the rotor position is available when idle, thus a special startup method is required. Similar to a stepper motor, the motor is operated in the controlled mode for several commutation cycles until it has attained the required speed and the sensorless measurement can determine a rotor position.

3. Important Parameters