The hardware of this routine is based on S32K142EVB, the IDE is S32_Design_Studio for ARM 2018.R1, SDK version is S32K1xx_RTM_3.0.0, PTB12 is used to simulate Hall pulse output,PTC12 and PTC13 are buttons to change the flip frequency of PTB12 port, and PTB13 is used as the input capture port. When using the demo program in this article, you need to connect PTB12 and PTB13 ports.
Here we assume that we are using a brushed DC motor!
1.The Hall sensor
The Hall sensor is a magnetic induction sensor. The magnetic ring and the Hall element form an induction combination. The magnetic ring rotates with the rotor. The Hall induction magnetic ring rotates with the rotor. , 3-pole pairs, 4-pole pairs, etc., each pair of poles is divided into two levels of N.S. A pair of magnetic poles outputs one pulse signal, and multiple magnetic poles output multiple pulse signals. The number of magnetic pole stages determines the number of pulse signals. , the higher the accuracy.
Hall sensor
2.The relationship between the motor magnetic ring series and the output Hall waveform
5 pole pairs
3.Determination of motor rotation direction
The direction of the motor is judged by the phase difference of the two Hall signals. As shown in the figure below, the phase of Sensor A is ahead of Sensor B, so it can be considered that the current rotation direction of the motor is clockwise.
4.Calculation of motor speed
The speed of the motor can be calculated by the pulse width of the pulse, and the number of revolutions of the motor can be calculated by the number of pulses. Assuming that the Hall magnetic ring of the motor has 5 pairs of poles, it means that there are five pulses in one revolution of the motor, and the speed of the motor = 60 / (t1 * 5) rev/min. The number of pulses can be obtained by the edge capture function of the FTM.
Motor speed and stroke
Assuming that the clock of the FTM is 2MHz, then it takes 1/2000000 seconds for the counter to add 1. Since the unit of the motor speed is rpm, the calculation formula of the motor speed is : -> Motor Speed = 60 / (5 * a* (1 / 2000000))
In this formula, '5' is the number of pole pairs of the magnetic ring, and 'a' is the difference of the counter corresponding to the falling edge of two consecutive pules.
Let’s do a test, the square wave in the below figure is the outputs of PTB12, and the output pulse period is 32.1ms. Then the time required for the motor to rotate once should be:32.1ms *5 = 160.5ms, then the speed of the motor should be: 60 * 1000 / 160.5 = 373.83rpm.
PTB2 output square wave
The below picture is directly obtained by the debugger. It can be seen that the speed of the motor at this time is 373, which is not much different from the value measured by the oscilloscope, which is 373.83. This is because I did not use the floating-point calculation result in the program. In summary, we use the input capture function of the FTM module completes the calculation of the motor speed.
debuger monitor results
5.How to calculate the direction of rotation of the motor
Above we calculated the speed of the motor, but did not make judgement on the direction of the rotation of the motor. As mentioned above, the rotation direction of the motor is judged by the phase difference of the two Hall pulse waveforms. Usually, we think of using the timestamp to judge the current state of the phase, so we will enable the two input captures, and then calculate the two Halls timestamp of the falling edge of the pulse.
In fact, there is a simpler method, it only needs to read the high and low state of the other Hall pulse level when the falling edge of one hall pulse is interrupted. In short, we only need to enable one input capture, and the other to be used as a GPIO port.
