Part Number: TMS320F28379D
Hi,
I am able to successfully configure PWM global loading registers allowing me to control multiple PWM modules' frequencies by configuring just the master PWM. Great.
However, I have noticed that just immediatly after setting the new Period (TBPRD) to the PWM1, the "salved" PWM's signal is corrupt for about 30us. I have checked the Errata for F28379D and I could not find anything in relation to this, so here I am.
My setup is as follows:
- PWM1 is the master, HIGH on ZERO, LOW on PRD, produces SYNCOUT on ZERO
- PWM4 is the slave, has no phase shift, is configured almost identically to the PWM1. The difference is the LOW is on ZERO, HIGH on PRD (opposite to PWM1). PWM4 is also configured to enable Global Load of TBPRD register from the PWM1 on "EPWM_GL_LOAD_PULSE_SYNC_OR_CNTR_ZERO".
- Every 1ms I just calculate new period and update the the TBRD of the PWM. The Purple trace is the GPIO toggle allowing me to trigger the scope, just before the TBPRD is written to.
In daisy chained PWM configurations, when is it best to updated the registers? If I have PWM1 -> PWM4 chain, should I update the PWM1's TBPRD after the PWM4?
Thanks in advance.
