This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Center aligned PWM for Space Vector Modulation

Hello, 

 I have just started with programming on micro controller . I am working on variable frequency drives . I need to implement SVPWM on TMS570 LS3137. I am having some  basic questions as follows   

Problems: 

1. How to create center aligned PWM?

2. Out of 6 PWM signals - 3  non inverted center aligned PWM + 3 inverted PWM , how to implement inverted PWM with respect  to non inverted in case of center aligned?

3. How to define variable duty cycle ?

Thank you in advance.

Regards ,

Suyog

  • Hi Suyog,

     Attached is an example of creating symmetrical PWM. The example should answer your question 1 and 3. You may use the example to extend to question 2. In the meantime I will talk to my colleagues to see if there is a better example for question 2.

    https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/312/1738.Symmetrical_5F00_PWM.het

  • Hello Charles,

    Thank you very much speedy reply.

    REgards,
    Suyog
  • Hello Charles ,

              I am having few  doubts as follows :

    1. Why havn't you used only XOR logic for symmetrical PWM  (explained in attachment ) ? Is it necessary to set up with up-down counter ?  

    2. Suppose, we make XOR logic on 2 pins. What is the expected output waveform - edge aligned / center aligned ?  

    Regards,

    Suyog  

  • What do you meant I 'have not' use the XOR logic? If you look at the comment in the file it says " The output of the two pins are XORed together will produce a symmetrical PWM. In order to use XOR feature, the two pins to be selected must be an even/odd pair adjacent to each other."

    The NHET natively can not produce symmetrical PWM because it does not implement a up/down counter. Instead the counter is only a up counter resulting an asymmetrical PWM. In order to produce a symmetrical PWM, you generate two asymmetrical PWM signals and XOR them together.

     Below is an illustration of symmetrical PWM generation using XOR feature base on an earlier generation of the HET in the R1x family of MCUs. The concept is similar for the NHET implemented in the LS3137.

  • Hello, 

    Sorry, I misunderstood few concepts.  Yes, I had already gone through this program. I've got it clear now. 

    Thank you for helping me .

    Regards,

    Suyog 

  • Glad it is clear to you now.
  • You can try the attached project for HET IDE for an example of HET doing this:

    0724.pwm3ph.zip

  • Hello Anthony,

     Thank you very much for your reply.
             I am going through your code. I'll get back to you if I get stuck on a problem .

    REgards,

    Suyog

  • Hi, Anthony F. Seely

    Why define the following pins:

    ; AND-Share Hi-Res Pins are even pin + 1
    PHAHI_SH .equ (PHAHI + 1)
    PHALO_SH .equ (PHALO + 1)
    PHBHI_SH .equ (PHBHI + 1)
    PHBLO_SH .equ (PHBLO + 1)
    PHCHI_SH .equ (PHCHI + 1)
    PHCLO_SH .equ (PHCLO + 1)

    Are these necessary?

    In the same low cost devices such as TMS570LS 02x/03x/04x, there are only four odd pins.

  • So this is a general case; where all 6 pwms are generated for the output.

    I'm not sure about the low cost kit you are referring to but some driver chips can take 3 PWMs and generate deadband and complimentary pairs

    from only 3 signals.   So maybe that's what is going on w. the 4th PWM possibly there to drive the high side output voltage ... (the voltage regulation is not

    comprehended by the example program we are talking about here ..)


    The good thing is once you have your head around the N2HET you can customize the program to do exactly what you want. ...

    So don't get discouraged.

    Best Regards,
    Anthony