Hi XiangGuo,
Here is the feedback that @xiangjun_rong kindly provide me:
1 Is the midpoint voltage of the three-phase winding constant?for UDCB/2?
You can check the AN12435 equation[6].
Where you can see the expression for the voltage of the unpowered phase being equal to the (UDBC/2) + (3/2*BEMF voltage of the unpowered phase)
2 After Q5 is on and Q1 is off, the B phase current is twice normal?
We do not think the current will be double in the branch switch, because the motor winding is an inductor, the current flowing an inductor can NOT increase immediately, it needs time.
Please check the following equation:
The inductor current increment= (V) * (Ton / L)
Where: V is voltage to inductor, Ton is turning on time and L is winding inductor.
3. The current at Rshunt is bidirectional, right?
Yes, the current at the Rshunt is bi-directional.
But it is dependent on the layout and PWM waveform, based on the circuit that you provided us,
Lets take this scenario :
Assume that Q1/Q2 PWM signals are complementary, Q3/Q4 PWM signals are complementary,
Assume that at an instant, Q1/Q4 turn on, the current flows from DC bus to A winding via Q1, to B winding, to Q4, to GND to Rshunt.
At next instant, Q1/Q4 turn off, Q2/Q3 turn on, because of winding inductor, the current will keep from A winding to B Winding, this is the current flowing direction: A winding to B winding, to Q3 parallel diode, to DC bus(charging the power supply capacitor), to Rshunt, to Q2 parallel diode, to A winding.
So in ONE PWM cycle, the current flowing into Rshunt is in two directions.
I hope this helps !
Best regards,
Diego.