Hall sensor offset angle identification with Motor Pilot and oscilloscope

[Mark7]

[VERSION]: MCSDK version 6.3.1
[TOOL]: MC Profiler

[HARDWARE] EVSPIN32G4

Hi everyone,

I have a question regarding the hall sensor placement electrical angle parameter.

I have tried identifying the hall sensor placement electrical angle with 2 different methods

1. Running the hall sensor profiler in the motor pilot, as outlined in the MC workbench documentation chapter Motor Profiler Application Note, section 5. Hall sensor profiling with the Motor Profiler tool
2. Measuring the delay between the Phase U back EMF and the first rising edge of hall sensor H1 with an oscilloscope as outlined in https://community.st.com/t5/stm32-mcus-motor-control/is-this-a-120-or-60-deg-displacement-angle-hall-motor/td-p/263494 and um1052 section 8.3.3 https://www.st.com/resource/en/user_manual/um1052-stm32f-pmsm-singledual-foc-sdk-v43-stmicroelectronics.pdf. Since my motor is wired in a delta configuration, there is no neutral point to reference the phase voltages. So as suggested in um1052, I wired the phases up to a star resistor network to reconstruct a neutral point.

The motor I am using has the sensors placed 60° apart (Sensors displacement parameter = 60).

Now, the motor pilot yields offset angle 270°, while the identification with the scope gives an angle somewhere between 300-310°. I checked this at different rotational speeds, the phase delay is clearly >=300° every time. That is, when I measure the phase voltage against the neutral point, exactly as suggested in um1052. I stumbled upon a documentation from Infineon https://www.infineon.com/dgdl/Infineon-Functional_Reference_Manual_iMOTION_Solution_Designer-UserManual-v01_02-EN.pdf?fileId=8ac78c8c850f4bee01853e4882b37df2 section 2.8.6, where they propose a very similar method of offset angle identification using the back EMF voltage, but instead of measuring the phase U voltage against GND, they measure the line voltage between phase U and V. Using this method instead, I get the same 270° angle offset as with the motor profiler.

All this to ask: is there perhaps an error in the ST documentation, where the line back EMF voltage between phase U and V needs to be used for identification instead of phase U voltage against the neutral point? Or is there another explanation for the >=30° discrepancy between both identification methods?

There is another point which puzzles me about the hall sensor profiling in the motor pilot application. If I connect the terminals for the phases differently, but keeping the order of the phases the same, I get offset angles that are 60° apart. What I mean is, connecting up the phases as

• U, V, W results in offset angle 270°
• V, W, U results in offset angle 150°
• W, U, V results in offset angle 210°

Since the phases are spaced 120° apart, I would expect the offset angles to be spaced 120° apart as well, e.g. 270°, 150°, and 30°. Could you please share some insight into how the profiler identifies the angle offset?

Thank you for your assistance.

[GMA]

Hello @Mark7,

Thank you for your reporting about hall sensor placement electrical angle documentation. This point will be investigated on our side.

About your second point, using a motor with a HALL sensors displacement parameter = 120°, and using your phase rotation scheme I obtain a +120° difference as mentioned.

[Mark7]

Hello @GMA ,

thanks for looking into this. I'm curious what your team will figure out regarding the scope identification method.

---

@GMA wrote:

About your second point, using a motor with a HALL sensors displacement parameter = 120°, and using your phase rotation scheme I obtain a +120° difference as mentioned.

---

I thought the hall sensor placement electrical angle parameter was always the phase difference between Hall sensor 1 and whatever is connected to motor phase U. Since I did not touch the hall sensor connection, but only the motor phases, I would not expect the hall sensor displacement to make a difference. Does the motor profiler change which hall sensor it references the hall sensor placement electrical angle to? Is there some way for me to see what calculations are happening in the background during profiling?

Best regards,

Mark