• State Resolved
  • Locked Locked
  • Replies 2 replies
  • Subscribers 31 subscribers
  • Views 285 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.

AM2634-Q1: EPWM output clock accuracy

Francois Charlot
Mastermind 29297 points

Part Number: AM2634-Q1

Hi Champs,

My customer considers using an EPWM output as a clock source for an external circuit that requires a 25 MHz ±100 ppm clock.

The AM263x datasheet states that the minimum half-period of the EPWM outputs is 20 ns, which fits exactly the 25 MHz need.

However it is unclear what the accuracy of such a clock could be. Besides that of the crystal that drives the MCU clock, could the internal logic affect the clock accuracy?

Thank you.


Best regards,
François.

  • +1
    TI__Guru*** 135355 points

    Hello Francois,

    The internal clock will have further variance and even with just 100 ps variance the difference in frequency would be significantly higher than 100ppm. In fact for the PPM to be met, you'd require the internal clock variance to be 2pS or less. That would enable the potential range of the signal generation to be at lowest 24997500 which is the minimum allowed for 100ppm.

    Based on this I cannot reasonably believe that between the internal clock circuitry, the pathing to the PWM module, and the external crystal specs that it would be possible to generate a PWM signal with 100ppm accuracy. I also looked in general for if similar has been done and the resounding answer I saw was 'no, you have to use a crystal' to get that level of ppm accuracy.

    Best Regards,

    Ralph Jacobi

  • 0 in reply to Ralph Jacobi
    TI__Mastermind 29297 points

    Hello Ralph,

    Thank you. This answers my inquiry.


    Best regards,
    François.