F2803x High-resolution PWM period in complementary mode

Channel B is not upgraded to high resolution nor channel A is down graded to normal resolution - they stay as they are originally designed. You will simply not get high resolution with channel B.

Stian

On the 2802x and 2803x devices, the only way the B-channel can have High-Resolution is using the following bits in HRCNFG register:

1. SWAPAB - swaps the output of A with the output of B. In this case, A and  B outputs are swapped so  B will have high-resolution output based on the HR configuration options for A, and channel A will be non-high-res (based on configurations for B)

2. SELOUTB - ePWMB appears as an inverted ePWMA signal. So if ePWMA is configured for high-res, ePWMB will also have high-res (inverted version of A)

Thanks for the clarifications ....forgot about this feature

Thanks. 

Is it possible to use dead band when using HighResolution mode?

JHi,

As mentioned in the posts above the B channel of each PWM is not inherently high-resolution(HR).  This means that if you want to use a PWM modules A and B channel to drive some FETs with deadband you will not be able to use HR on both the A and B channels.

Alternatively, you could synchronize the PWM1 and PWM2 module timers and use the channel A HR outputs of both modules to drive your upper and lower FETs.  The deadband could then be added/subtracted via software each time you change duty cycles in the ISR.


Thank you,
Brett

Thanks for the answer. I did this and noticed that you can use FED and RED for the A channels to have deadtime and you don't have to do additional calculations for it.

JHi

Brett,

         If I set ePWM1A as HRPWM while ePWM1B as ePWM, and give them a HR period, what will happen ? jitter? 

         Will only the ePWM1B freq. jitters or both A and B jitter ? 

thank you!

Jim,

Only the PWM's B channel should jitter.


Thank you,
Brett

Thank you!

Will ePWMxB have jitter in the case where SWAPAB or SELOUTB is enabled?

mark