What happens if a PWM frequency is applied to an induction motor?

Question

Electronics

If a direct current is applied to an induction motor, then the rotor will become magnetized and will not rotate. And if you apply a PWM frequency and change its duty cycle, then what will happen?

Answer 1

But I still have fear.
Atmega 328 will probably handle it in the Arduino nano package.

Yes, an arduino will do just fine. And fear... To overcome it, first simulate not a sinusoid generator, but a rectangular pulse generator, it is generally elementary simple. Describing the algorithm in words is another quest, but I'll try.
1. Apply log.1 to the desired output of the MK, then start the counter for a count of 10 milliseconds, at the end of the count, change the unit at the output output to 0 and start the counter again. At the end of the count, return one again at the output, and so on. Then repeat these steps cyclically. At the output we get rectangular pulses with a frequency of 50 Hz. This is the digital analogue of a symmetrical 2-transistor multivibrator.
2.Now let's connect PWM here. By the way, usually in advanced industrial MKs there is already a node that can generate PWM, it just needs to set the carrier frequency (usually 10 ... 1000 times higher than the generated 50 Hz) and enter the coefficient value into a special register as needed. filling, from 0 to 100%. We are using it. Everything is as in paragraph 1, but instead of supplying 0 and 1 to the output, we supply there what the PWM driver generates. And according to the signals of the end of the counter, we will change the value of the coefficient in the PWM shaper every 10 ms. filling 0 and 100%. We get exactly the same rectangular pulses, since zero coefficient. padding is the same regular zero, and 100% is the regular 1. So far, nothing has changed.
3.Now let's slowly move on to the formation of a sinusoid. First, instead of two intervals of 10 ms, let's break the entire 50-millisecond period of the sinusoid into 36 intervals of 1.39 ms (1389 µs). Why 36 - it turns out a little later. Instead of one counter, you will have to start three - one will count at 1389 μs, the second counts the number of counts of the first counter, and as soon as it counts to 9, it starts the third one, counting up to 2, then starts counting again. According to the signal for the end of the count of this third, it is necessary, as in clause 2, to set the coefficient in the PWM shaper in turn. filling 0 and 100%. As a result, at the output we get all the same familiar rectangular pulses of 50 Hz. The first and second counters are currently idle (except for the fact that they start the third one).
4.Now, in each cycle of the first counter (let me remind you, it lasts 1389 μs), we check the status of the second one. If its state is -1, then we take the first value from your table of sines, if 2 - the second, if 3 - the third, etc., and enter this value as a coefficient. filling in the PWM shaper. At the output, we get an increase in the filling of the PWM signal along a sinusoid. But these will be quarters of a sinusoid.
5. So we got to the third counter. It turns out that in each cycle of the first counter it is necessary to check the status of not only the second, but also the third. If the state of the third is 0, then the values ​​​​from the table of sines must be taken in ascending order, as described in paragraph 4, and if there is 1, then in descending order: not from 1 to 9, but from 9 to 1. So we get the falling part of the sinusoid.
6.And finally: we add one more output to the MK. The first controlled power transistors that form the positive part of the sinusoid (those that are powered by a positive voltage source), and the added second - those that form the negative and are powered by a negative voltage source. We also add a fourth counter, which will count the cycles of the third, it must also count up to 2, and each time after the end of the count, at its signal, the PWM output from the shaper must be redirected in turn to the first output, then to the second. In this way we get a complete 2-polar sinusoid.
If we want to change the output frequency of the sinusoid, we need to enter a different initial counting interval when starting the first counter - not 1389 μs, but which one is needed. This can be done "on the go" if a special subroutine is provided for this.
If we need to run a 3-phase motor, then the MK will have to provide 4 more outputs, to which to apply PWM, phase-shifted by 120 and 240 degrees (this is not difficult, but it will require one more counter).
Here is something like this. And we must keep in mind that all these programs on the Internet can be found already in finished form.