Getting started with EVSPIN32G4 6 Step, Workbench 6.3.0 and Motor Pilot 1.2.9

[JFree.1]

I'm looking to upgrade one of my years-old BLDC 6 step designs which used STM32F401 and too many other components! I like the look of STSPIN32G4 and so purchased an EVSPIN32G4 development platform.

It seems I have successfully worked through MC Workbench 6 Step Wizard, including adding my test motor, without error and one warning - pwm generation "Danger: there is no OCP nor Driver Protection enabled" - this despite there being everything required for this on the EVSPIN board but no option to enable it (can we add this to the wish list please).

'Generate the Project' works, I've followed through with CubeMX and IDE, produced code and downloaded to EVSPIN32G4 - all good.

ST Motor Pilot connects ok with board, clicking on 'Start', motor briefly kicks before Pilot reporting 'Speed Feedback' fault. Using 'Ack Faults' to clear the error, 'Start' again produces same effect consistently.

In 'Advanced Configuration' I notice a 'Speed Kp' value of 2038.  Changing this to -2038, I can now run the motor in either direction by moving the 'Speed Reference' slider, but with Ref 3600RPM producing 300RPM, and Ref 180RPM producing 2400RPM !  Something daft surely ?

[ADAVE.1]

Hi @JFree.1 ,

Glad that you liked the STSPIN32G4 😀

Coming to your questions:

- Automatic configuration of the internal peripheral of the STSPIN32G4 with the OCP feature is not unfortunately yet supported. But it will be soon...

- Issues with starting the motor usually relies in not optimized BEMF sensing network and/or improper open loop (acceleration) procedure. EVSPIN32G4 has a wide operating voltage rating and a huge BEMF sensing divider value that will not surely be suitable for all kind of motors. What is your bus voltage value? And did you try to optimize the start-up sequence? For additional help and suggestions, I recommend you to give a look at the following manual UM3259: 6step sensorless parameter optimization 

- Kp is the proportional gain of the speed loop. Since the speed error is proportional to applied phase voltage, this value is supposed to be positive. I suggest you to keep it low and positive in your trials and make the system more reactive when you manage to properly close the speed loop.

Regards

Alessio 

[JFree.1]

Thanks Alessio,

I'm upgrading an old design where I'm not allowed to affect any existing functionality, while free, of course, to add any new and exciting features!  To this end I'm not sure WB or Motor Pilot will be of much help.  If I can work out how to include a DAC to control cycle-by-cycle current limiting (RM0440 28.x - read it all several times but the fog has yet to clear!) into a minimal Hall sensor feedback 6-Step scheme then I will be on track to deliver a prototype to perform 'no worse than' the old design with relatively little time or effort.  New features and goodies may be built in and added from here.