Pulse Width Modulator (ePWM): Force a Continuous High Synchronously

Nick,

Which C2000 device are you using? There's not a straightforward way to do what you're looking for with the ePWM, but I'm pretty sure it can be done.

What does your trigger event look like? Is it a short pulse or a long event? Is there an expected range?

Can you explain why you want this to be a software triggered event? Most users prefer for this to be hardware as it is generally a fail event causing the PWM shutdown.

Regards,
Kris

Hi Kris,

Thanks for your reply. I am using the F28M36P63C2. The trigger event can be anything really, as it does not relate to any external process state, fail event, or anything like that. It's something I want to purposely set on startup that will serve to "gate" a one-time sync to an external ADC (www.ti.com/.../ads1274.pdf, Fig 73). I'd like the low-going ADC sync pulse to be on the order of 1us in duration at least, but what's equally critical is the timing of the low-to-high edge of the final EPWM pulse, which I am currently controlling using the up-down counter and CAU, PRD settings. This edge must always occur at the PRD, as I show in my diagram, and which is currently running fine.

But the timing of the final ADC sync pulse (more precisely, the low-to-high edge) needs to be under the control of the C program running in the C28 so as to coordinate with the startup sequence of other peripherals. The trigger event that is needed to turn the EPWM to a continous HIGH value can be as short or as long as necessary, sort of why I tried using EPWM1.AQCSFRC.bit.CSFA = 2, which works, but is asynchronous with the pulse I am trying to gate, and the concern is that there's a chance this could occur in the middle of the sync pulse, thus shifting the low-to-high timing.

Nick

Nick,

I think I understand most of it, but the "one-time sync" is throwing me off a bit. Is the goal to have the waveform only run for one period, or it runs many times and will terminate after the trigger?

Could your "trigger" be to just change your CAU to be greater than PRD, therefore it is never reached and the waveform is never set low? Assuming you are using shadow mode, this would result in the current PWM period completing as normal and the following one staying constant high because there is not an AQ event to change it.

Just a note, the EPWM guide you linked to is for an older device. Use the following document for future reference. It won't make any difference in this scenario, but may help you in other areas. Your EPWM has more features than the F2833x.

http://www.ti.com/lit/ug/spruhe8d/spruhe8d.pdf

Regards,
Kris

Hi Kris,

As far as the ADC is concerned, I only need to generate a single pulse once at startup, so you are correct in that sense, that is, the waveform only needs to run for one period. But since the occurrence of this pulse needs to be synchronous with other peripherals, having several such periods (e.g., 1 or 2 seconds worth) provides a good opportunity to verify that this synchronous condition is met prior to starting data acquisition.

I have not tried the change the CAU. I will try that. I have not understood how shadow mode works. Do I do anything special or is it enabled by default? The documentation isn't clear. I will check out the newer EPWM guide. Thanks for pointing that out.

Thank you for your suggestions!

Nick