Why am I not getting proper gate signal and voltage

Hello Everyone,

I have been trying to follow the example model "PMSM Velocity Control" where my final objective is to replace the idela converter with parameterized discrete devices. For now, I am using ideal converter, with different reference speed (Drive Cycle, constant, signal editor) and I have tuned the speed and current loops accordingly.

The Sample Time

1. Control Sample time (TSc) = Tsw/250 (Switching time over 250 samples per switching time)
2. PWM Sample time (TS) = 1/(10*fsw)

• The motor is seem to follow the reference speed correctly, however the generated agte signals are seem to sampled sine wave, between 1 and -1.
• The voltage (Phase-phase) and (Phase-to-Nuetral) waveform is quite unusual (my input is 48V)
• Although the control loops are tuned and there is very low oscillation in the controlled output, however the phase current has hige distortion which I tried to suppress via low-pass filter.
• When I put the parameterized semicondcutor instead of idela converter, the gate are signal are not even generating. A gate driver is used to provide PWM signal form the controller to the N-channel MOSFET.

I am attaching few screenshots and my model, if any could validate the model and provide the guidance regarding aforementioned.

Thank You!

Answer 1

[Sabin]

Even though it appears that is working the controller is not working properly. The inner current control loops are not doing what are supposed to do and the d and q reference voltages are saturated. The outer speed loop is also not working properly because the unsaturated torque (output from the controller) is approx. 1400 Nm but this is saturated to 81 Nm. The modulation waveforms are saturated and are not looking like sinusoidal signals anymore. You should double check if the system parameters are making sense and if you can achieve the desired power (28 kW) with this 48V system because this may not be physically possible. Then you should tune the controllers.

[Hassan Ali]

Thank you for pointing out the mistake, it was an error in the desired power which was supposed to be 2.8kW, not the 28kW. However, when I am parameterizing the N-channel MOSFET and uses gate driver to provide the gate signal to the MOSFETS, where the outpu gate voltage is Vth.

There is no reponse in that configuration, I am attaching the model hereby, could you kindly have a look, what could be the issue.

Thank You!

[Sabin]

It appears that the parameterization of the gate driver may be the issue. Try setting the Logic 0 input value to 0.3 and On-state gate-source voltage 3 V and it should work. However, the N-Channel MOSFET is a high-fidelity representation of the physical device and is probably unrealistic for longer simulations. In this case it is better to use the MOSFET (Ideal, Switching) block instead.

[Hassan Ali]

Thank you! it is showing response now.

Of course, it takes alot of computation power to run, and would definitely use MOSFET (Ideal, Switching), but my works involve working with Silicon and Gallium Nitride Semiconductors and there power losses, and i am not sure how to parameterize the MOSFET (Ideal, Switching) to match the characteristics of both semiconductor material as the parameterization options are very limited in MOSFET (Ideal, Switching).

Is there any way to achieve this objective?