• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 81 subscribers
  • Views 377 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

TMS320F28335: Synchronization Signal of TMS320F28335 Control Card (EPWMxSYNCO)

TI-CSC
Genius 9540 points
Part Number: TMS320F28335

Dear TI Experts,

I am working on my battery management project with F28335 MCU. In my case, I need to use two MCU control two individual converters to supply a DC bus. If I do not synchronize this two MCU, there will be a huge fluctuation at the output current.

What I am doing now is using the EPWMxSYNCO signal generated by ePWM module, and send it out via GPIO-33. The receiving MCU will receive the signal via EPWMxSYNCI at GPIO-32. But the receiving device has a issue on the synchronization - it cannot sync at each cycle of TBCTR generate the signal. Hence it causes the jittering of the synced device, and the converters output became uncontrollable.
The attached picture shows the EWPMxSYNCO from the master device. The sync pulses disappeared in almost half of the scope.

sn_customerservice_case_d2c9394b1b2be090602cf715464bcb37_attachments.zip

What causes such issue in my case? 

Regards,

Jonard Rico

  • 0
    TI__Mastermind 41365 points

    Jonard,

    I can't tell what you are trying to show in your scope image, it looks like some aliasing due to a low sampling rate on the oscilloscope. But maybe the pulses really aren't there. If the sync pulses aren't coming out of the device go look at your code and try to find anything that could be modifying the GPIO mux configuration or the PWM configuration. The Sync pulses shouldn't stop as long as the counter value generating such a pulse is reached. Usually CTR = 0 or PRD.

    Please also not the following:

    If you are syncing two devices on a period-by-period basis with separate clock sources it is recommended to used a PWM period of TBPRD-1 on the secondary device. the following post and associated thread talk about why such a trick is a good idea and how much margin to clock frequency error you get by reducing TBPRD-1 or a larger value like TBPRD-2.

    https://e2e.ti.com/support/microcontrollers/c2000/f/c2000-microcontrollers-forum/948337/ccs-tms320f28377s-28377s-sync-question/3534473#3534473 

    Regards,
    Cody 

  • 0 in reply to Cody Watkins
    TI__Genius 9540 points

    Hi Cody,

    These weeks I worked with the solution you provided. But I met another problem.

    As I am working on two boost converters controlled by two separate F28335 control cards. Since I am trying to control the shared output current of these two boost converters, I meet trouble - if I do not sync the EPWM TBCTR, there will be obvious fluctuation in output currents. If I sync the time base, there will huge jitters and cause the instability of these two boost converters. And the output voltage may become unreasonably high.

    TBCTR of the master MCU is 2995 (50.1kHz). The synced MCU TBCTR is 2999(50kHz).

    I tried the twisted wires connect two MCUs, it did not work fine. The coax cable did not as well. Even more, I tried the RS-485 diff signal with coax cable, it neither worked.

    I wonder if the EWPMxSYNCI is the pin that only suitable for two MCU chips close to each other?

    Thank you and have a great day.

    Regards,

    JOnard Rico

     

  • 0 in reply to TI-CSC
    TI__Mastermind 41365 points

    Jonard,

    The EPWMxSYNCI should work over a variety of distances and physical layers. The longer the distance and more devices you put in the sync chain would cause a larger phase offset, but this can be easily accounted for in the TBPHS register.

    If you are seeing jitter then perhaps some noise is being coupled on to the SYNCI pin. Have you observed the pin at all? This noise is very depend on the system and event the board layout, but I have seen two microcontrollers be synchronized this way many times. 

    Regards,
    Cody