SVPWM and BEMF Wareform type

[Vincent7]

Hello Everybody,

MCSDK:6.3.0

CubeMX: 6.3.0

Workbench : 6.3.0

Motor Pilot : 1.2.9

Power Board : IHM08M1

Control Board : NUCLEO-G431RB

Motor : GBM2804H-100T

I am currently testing FOC-OpenLoop-Voltage. I am using Motor Pilot for tuning and measuring SVPWM (at OUT1~OUT3) and BEMF (at D6, D8, D10) with an oscilloscope. However, both waveforms appear as square waves. Could this be related to the specific motor model I am using?

[egarcialopez]

You are trying to read directly the SVPWM signal, which is using Pulse Width Modulation to represent the three sinusoidal voltages that you intend to inject at each phase. If you want to reconsrtuct the three sinusoidal voltages you need to insert a Low Pass Filter between any of the three BEMF and the oscilloscope. You can build a very simple Low Pass Filter (LPF) with just a resistor and a capacitor.

[Vincent7]

Hello egarcialopez, 

Thank you for your reply.

I just searched online for how to create a low-pass filter using a resistor and a capacitor and connect it to D6, as well as how to calculate the appropriate resistor and capacitor values. I managed to test it successfully. I'll share the waveform later. Thank you again!

[Vincent7 · Accepted solution]

Hello egarcialopez,

Thank you for guiding me to use a low-pass filter to successfully display the saddle waveform.

Although the attached waveform still requires some further adjustments, it already shows the result I was aiming for.

Once again, I sincerely appreciate your guidance!

 

[egarcialopez]

Hi Vincent7,

you are welcome. Nice traces :).

It looks like there is a little ripple, maybe you can adjust the capacitor to a higher value so the cut frequency of the LPF is lower, closer to the frequency of the sinusoidal wave. Or add a second pole with a series inductor. But if you only need to have a general idea of what the phase voltage looks like, the filter that you are using is good for you.

[Vincent7]

Hello egarcialopez,

Sorry for the late reply.Thank you for providing a method to reduce part of the filtering effect. Currently, the components I have on hand are resistors (1kΩ / 4.7kΩ / 10kΩ) and a 154k capacitor.I can also manually adjust the PWM value.

For now, I would like to first confirm whether the saddle-shaped waveform appears, and after resolving the torque control issue, I will revisit the filtering problem.
Anyway, I really appreciate your suggestion~