Driving 3-Phase Full-Wave Brushless Motors: Sensorless 120° Commutation Driving

From this article, sensorless driving of a 3-phase full-wave brushless motor is explained. As an example, we consider basic 120° commutation driving.

Driving 3-Phase Full-Wave Brushless Motors: Sensorless 120° Commutation Driving

The diagram below is an example of a general circuit for sensorless 120° commutation driving of a 3-phase full-wave brushless motor.

In sensorless driving, there is a method in which the induced voltages in the coils are used in place of Hall elements for motor position detection. In this method, the signal CT at the midpoint of the three coils is used. This CT signal and the signals A1, A2, A3 are input to the driving circuit, and comparators are used to perform comparison processing, and outputs are generated. This process is essentially the same as when using sensors, except that induced voltages are used instead.

Induced voltages occur when the permanent magnet rotates and the magnetic flux passing through a coil that is not passing a current changes; it cannot be detected in coils in which current is flowing. In 120° commutation driving, current is flowing in two out of the three phases, and is not flowing in the third phase. An induced voltage appears across the terminal of the phase in which current is not flowing; by using the voltage CT at the midpoint to detect the zero crossing point of the induced voltage, the motor position is detected.

In one phase of a 3-phase motor, there are two off-intervals in which current does not flow during one motor rotation (through 360°) (two 120° intervals with current and two 60° off intervals), and so there are six off-intervals for the three phases. That is, position detection is possible once every 60°. Output signals are generated on the basis of 60° signals between zero crossing points for each phase.

Below is a detailed explanation using a driving waveform example.

The current waveforms in coils 1, 2, and 3 indicate that driving is by 120° commutation; that is, current enters into a coil over 120° and is then off for 60°, current flows out over the next 120°, and is then turned off for another 60°. The resulting output voltage waveforms are A1, A2, and A3.

Induced voltages used in detecting the motor position can be detected during 60° off-intervals for each phase.

Taking coil 1 as an example, during the off intervals indicated in red, an induced voltage appears in the output A1, and is detected. At these times, currents are flowing in coils 2 and 3. Upon switching to off from a state in which current is flowing into the coil, after spike noise (a transient voltage) occurs upon switching, the induced voltage declines. During the transition from a current outflowing state to off, after spike noise has appeared, the induced voltage rises.

The comparator output BE is a signal obtained by comparing the induced voltage signal occurring in the output and the coil midpoint signal CT (not shown); at the induced voltage zero crossing point, there is a H→L transition when falling, and a L→H transition when rising.

The comparator outputs BE1-3 are used to synthesize signals shifted 60° by a process called waveform section selection and synthesis, in which the additions are performed. Because these signals include spike noise, which was not in the original induced voltages, a mask signal (H parts) is used to mask the spike noise to synthesize induced-voltage 60° signal waveforms, and from waveforms which have been delayed in phase by 30°, the synthesized waveforms and output gate waveforms are generated.

The diagram below shows in enlargement the behavior of the outputs A1-3 during coil-off intervals. This is an example of a 60° interval in which coil 3 goes from a state in which current is flowing out to the off state. The circuit diagram shows the state of the circuit at this time.

In the case of A2, the high side MOSFET is on and the low side MOSFET is off, so that current is flowing out and the voltage is at high level. In the case of A1, the low side MOSFET is turned on, so that current flows from A2 and the voltage is at low level. A3 switches from the state of inflowing current to the off state, so that, as the waveform indicates, after a momentary jump in voltage, the induced voltage appears gradually and linearly. The voltage difference between A3 and CT at this time is shown as a broken line together with the A3 voltage waveform. The CT voltage is approximately one-half the voltages of A2 and A1.

On taking the difference between the voltages A3 and CT, the zero crossing point of the induced voltage can be detected, as shown on the right side, and BE3 in the preceding timing chart transitions from L to H.

We have explained sensorless 120 commutation driving using a simple example. Compared with 120° commutation driving using sensors, the broad outline is much the same, with the difference that induced voltages are detected instead of relying on sensor signals.