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.

TMS320F280038C-Q1: TBPHS sync chain and delay

Part Number: TMS320F280038C-Q1


Hello,

We are migrating from a f280049 DSP to a f280038 and I noticed in the reference manual there isn't a disclaimer saying there's a 2 cycle delay between SYNC Pulse and slaves updating their TBCTR values like there was in the 280049. Can you confirm the 28003 do not have this inherent 2 cycle delay that requires TBPHS = TBPHS_DESIRED+2. Also the f280049 (as well other models we've worked with like the f28379) had synchronization chains (see second picture) where only certain PWMs could sync downstream ( e.g PWM4 could sync off of PWM1, but not vice versa). Is this limitation also gone on the 28003 in that PWM4 could sync off of PWM1, but also, with a different setup, PWM1 could sync off of PWM4? 


  • Hello,

    Yes, in the f28003x, any ePWM sync-out can drive any epwm sync-in.

    A good reference point to determine the changes to a peripheral between devices is the C2000 Real-Time Control Peripheral Reference guide. In the peripheral descriptions section, it describes the differences between each devices' iteration of a given peripheral- for the ePWM, it specifically mentions that the second version of the type-4 ePWM includes a simplified SYNC scheme. The f28003x uses that version of the type-4 ePWM.

    In order to determine the sync delay, refer to the device datasheet section 6.14.1.1 ePWM Electrical Data and Timing. The tables shown below describe the exact timing requirements of many ePWM connections, including sync-in, in terms of ePWM clock cycles:

    For the F28003x, that is 2 ePWM clock cycles.

    I hope all of that helps, Jason!

    Regards,
    Jason
    Note: Key takeaways from this post are bolded.