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.

question about epwm1b jitter on f28069

Expert 1215 points


Hi,

On my board, I use ePWM1B to generate a simple 50% duty cycle PWM at exactly 1/7 of the SYSCLKOUT. But it seems there is a bit of jitter on the generated PWM. So my questions are

- Do I need HRPWM for this because the frequency of the PWM is quite high?

- Given the jitter issue quoted from TRM, "When dual-edge high-resolution is enabled (high-resolution period mode), the PWMxB channel will have ±1–2 TBCLK cycles of jitter on the output." The only way to get HRPWM on ePWM1B is to use SWAPAB or SELOUTB. Will ePWMxB have jitter in the case where SWAPAB or SELOUTB is enabled? 

Thanks

  • Yan,

    Can you define what jitter you are seeing and between which PWM edges?  Can you also mention what frequency you've configured SYSCLK to be?

    Generally speaking, the ePWM module will not have much jitter and therefore will be precise*.  However, it won't be accurate at high frequencies.  This is where the HRPWM can help.

    For example, if you're configured an 80MHz F28069 device's ePWM module to generate a PWM with a frequency of 11.428MHz, the ePWM module can only generate 7 different potential duty cycles.  It also won't be able to accurately generate specifically 50% duty cycle.  The closest to 50% duty cycle you could generate is something like 42.857% or 57.143% duty cycles.  Whatever duty cycle is generated should have very little jitter though*.

    If jitter (not accuracy) is truly your concern than using the HRPWM will likely not help.

    *any jitter that occurs will mostly be a result of oscillator/crystal/PLL jitter.  This also affects SYSCLKOUT.  You should also see this jitter on SYSCLKOUT if output.  If low jitter is needed in your application, you may need to use a highly precise external crystal.


    Thank you,
    Brett