Iq and Id PID oscillating/unstable after some time

[CWern.1]

Hello,

we do have a custom motor control board based upon B-G431B-ESC1.

A custom inverter board is created with the board designer and the project is created with MotorControl Workbench 6.3.1.

 

The system is running fine at higher RPMs (>2500 RPM; max RPM of the motor is 2900).

However when running at lower RPMs the system is becoming instable after some time. I see in the motor pilot that IQ_MEAS and ID_MEAS are increasing till the overcurrent limit is triggered. In this error case the motor RPM is remaining constant, but power consumption is increasing. 

I can also see that a shift between the HALL_EL_ANGLE and the STOPLL_EL_ANGLE (this is the auxiliary sensor) is becoming visible in this case. When the system is working normally the two are aligned.

I can delay this behaviour by decreasing the integral part of the PI parameters for Iq and Id or increasing FOC rate to 5, but I can not get it to stop. 

Does anyone have any tipps how to get the PI regulators to become stable?

 

Attached are the motorcontrol project and the the json of the board.

[Brohanlo]

---

The system is running fine at higher RPMs (>2500 RPM; max RPM of the motor is 2900).

However when running at lower RPMs the system is becoming instable after some time. I see in the motor pilot that IQ_MEAS and ID_MEAS are increasing till the overcurrent limit is triggered. In this error case the motor RPM is remaining constant, but power consumption is increasing.

---

I am having a similar problem and am looking for a solution. I am using a EVSPIN32G4-DUAL evaluation board and trying to spin a motor in speed mode. If the speed reference is increased above 700 RPM's, the I_Q increases until an overcurrent fault occurs, see attached. 

I notice in my case that the speed measured from the encoder starts to differ from the speed reference. Presumably the current increase is caused by the speed controller trying to keep up with the reference. I don't understand why the increase in current doesn't correspond to an increase in speed though.

I have played with the PI values but that hasn't fixed the issue.

[GMA]

Hello @Brohanlo,

As this is a different use case, could you create a new post to describe your issue?
In your previous post, you mentioned that the encoder/Hall signal RC filtering was not adequate for your application. Did you modify the RC filter? Have you plotted the filtered encoder signals in your case?

[CWern.1]

A quick update:

We managed to get the current to stabilize by making the Id PI parameters negative. This seems to work but I have no idea why. Can anyone explain this?

[GMA]

Hello @CWern.1,

Could you decrease the Average speed FIFO depth to 6?

 

[CWern.1]

This is still open, can anyone explain why the PI parameters need to be negative for stable motor control?

[GMA]

Hello @CWern.1,

Could you share a plot of Iq/Id ref and Iq/Id measurements with original PID values and your negative modified one?
Did you plot Ia/Ib signals to check that they are correctly sampled?