• State Resolved
  • Locked Locked
  • Replies 1 reply
  • Subscribers 82 subscribers
  • Views 465 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

TMS320F28388D: PWM Synchronisation Delay

ROGERG
Mastermind 36180 points
Part Number: TMS320F28388D


Dear Experts, 

could you please help with below ePWM module synchronisation question. 

In customer application it is very important that several PWMs are time-synchronous.

For this purpose, ePWM module 1 is configured as a master, which generates a SYNC-OUT signal for the other three ePWMs.

However, it can be observed that there is a time delay of almost 60 ns between the master module (ePWM 1) and the slave modules (ePWM 2,3,4) which can be measured using an oscilloscope.

However, the slave modules all run synchronously with each other. This problem also occurs in the sample project "epwm_ex3_synchronization" of the C2000 product.

Can it be concluded from this that there is general delay from a master module to a slave module ? In the datasheet for "Time-Base Counter Synchronisation Scheme" there is also a "Pulse-Streched" block mentioned with 8 PLLSYSCLK cycles. Does it mean there is always an 8 clock cycles delay to be expected ?

If so, then the conclusion would be that for perfect synchronization of the ePWM modules with each other, a master module would have to be used, which can not be used for the actual task, but would only serve as an aid to synchronize the slave modules. i.e. you would need five ePWM modules to perfectly synchronize four ePWMs? Is this assumption correct ?

Thanks and best regards,

Gregor

  • +1
    TI__Mastermind 42195 points

    Gregor,

    This should only require 4 PWM modules. The 3 slave modules will all require a small shift in their TBPHS value to align to the master. This delay should not be 8 clock cycles I expect this pulse stretcher was added to ensure that the sync pulse is always reliably detected by the next PWM in the sync chain. 

    I expect a 2 pulse delay between the master PWM and the slave PWM. Since you are seeing 60 nS we should verify if a clock divider is being used.

    Regards,
    Cody