open loop control in MCSDK 6.4.1, CW and CCW have huge difference speed

[VictorChiou]

[VERSION]: MCSDK 6.4.1

[TOOL]:

• 6-Step Firmware

• Control by MC Pilot

[DETAILS]:
In open loop mode, the same duty causes a big speed difference between CW and CCW.
It doesn’t make sense.

[EXPECTED BEHAVIOR]:

• In open loop control, CW and CCW should not have a big speed difference.
• I guess the ".PhaseShift" parameter does not consider CW and CCW.
  The ".PhaseShift" of CW and CCW should be complementary,
  and the sum should be 60° electrical angle.

[HOW TO REPRODUCE]:

1. Control the motor with 6-step algorithm, current mode.

2. Use Hall signal for commutation and speed calculation.

3. Control by MC Pilot.

4. Set CW or CCW direction.

5. Enable open loop mode and set open loop duty.

6. Start the motor.

7. Get the CW or CCW speed.

[2025-10-07 update]

I tried using v6.3.2.

still In open loop mode
I can get the same speed with the same duty when I set the "Placement electrical angle" to 124.

so I think my issue is the same issue reported by @jovijuan.

 

[Fabrice LOUBEYRE]

Hi @VictorChiou, first, I apologize for the delay.

In the six-step algorithm, there is no relationship between the PhaseShift parameter and the counterclockwise (CCW) or clockwise (CW) rotation of the motor. As you must know, we define PhaseShift as the electrical angle shift between the H1 Hall sensor and the actual 0-degree electrical angle. Regardless of the rotation direction, this parameter remains unchanged. Only the six steps ordering and Hall states vary.

However, when performing the Hall profiling of a motor, for example, using the Motor Profiler feature of the MCSDK, the PhaseShift parameter computation may vary between CW and CCW. Therefore, a mean value of both values is computed and used as the PhaseShift parameter. 

For example, the motor profiling of the SHINANO motor embedded into the MCSDK provides the following values:

mean value 297 = ((304 + 291) / 2). You can use 304 for CCW and 291 for clockwise CW for greater precision.

 

 

At each six-step transition, the phase shift information is added to the Hall timer event using the code you referenced, regardless of the rotation direction.

Regarding the problem of CCW/CW initial speed in open-loop mode, I do not currently see an issue with my setup.

I will investigate this point further

Best regards

Fabrice

[jovijuan]

I have also experienced important issues with Hall-effect sensor as main sensor in 6-step commutation mode when Hall sensors placement is not aligned with 60° steps. And those issues are related to how that delay is handled differently between CW and CCW.

It's described in this other forum post: Wrong 6-step commutation from Hall sensors not 60° aligned since v6.4.0