Motor Control and F28335

Hello

I am using "Simulation of FOC Using PMSM Model" demo to improve my project. It is very useful for me. But there are some questions about it? These are:

1. The model uses two ADC inputs for current regulator. (Selector selects Ia and Ib motor currents) SoI think we can read only two channels instead of three channels. I saw the 3rd current is used to control speed. I won’t control speed. So we can use less current sensor device and there is no need to use Selector block. Isn't it right?
2. What is the function of "Multiply by -1" subsystem after the Shift Arithmetic block?
3. If the input of Park Transform is sinusoidal, shuld the output of it be sinusoidal, shouldn’t it?
4. What is the aim of the desired gain block (Iqref and -1 multiplied by IQNmpl32 block) under Motor Control Algorithm > Generating Raw Space Vectors > Control Subsystem?
5. What is the aim of Saturation block after the PID regulators?
6. In Motor Control Algorithm > PWM Scaling subsystem why are we adding 1 to the duty cycle?
7. Is the qref magnitude of the motor currents?
8. Is Q format used floating point processors, for example TMS320F28335? Because this processor is floating point.

If you could spend time for replying my questions, I will be very glad. Thank you from now…

Meral

Answer 1

[Sabin]

Some answers, with a bit of delay. First of all this example is not anymore in the product in the form it was. The model was in R2017b but updated overtime. Apart from that:

1. For Park transform we need the measurements of two phases if the system is balanced (ia+ib+ic=0). The speed might not be needed but the position is needed for the Park transform.
2. Not sure which part of the model are you reffering.
3. Park transform has AC inputs but the output is DC. Check this page for more insights: https://mathworks.com/help/sps/ref/parktransform.html
4. Not sure which part of the model are you reffering, I don't see it in 17b.
5. dq voltages mjust satisfy the quadratinc voltage constraint. that is actually done there computing the magnitude and giving old voltages or unsaturated new command if feasible.
6. That is a scalling of the modulation waveform.
7. If the d axis current is 0 then the q axis current is the magnitude of the phase currents.
8. The Q format does not require a floating point unit but could be done on such a processor. Not sure I understand the question correctly.