LAUNCHXL-F28069M: Deadband for ePWM in Simulink and EPWMxB enable

Gergosz2

Hello Guys!

I think you will answer my question very fast, but I am l littlebit uncertain about the deadband unit end the EPWMxB.

So, I am using F28069M based launchpad and I am using MATLAB Simulink for Code generation. I have read the Simulink reference of the EPWM

modules and I am also using TMS320x2806x Piccolo Technical Reference Manual (spruh18g).

I understand how the deadband unit works, but I don not know whether I have to enable EPWMxB ? As I have read, the deadband unit could use

only the EPWMxA signal to generate the appropriate output signals. But in this case do I have to enable the EPWMxB on the third tab?

And this is my config for the deadband unit. It will be used for an inverter using 1.5us deadtime.

over 7 years ago

0 Cody Watkins over 7 years ago

TI__Mastermind 42195 points

Gergosz2,

I have never used the MATLAB software to setup the PWM, however I suspect that you will at least need to check the "Enable ePWMxB" check box. I assume that this gets translated into some GPIO setup enabling ePWMxB to reach the IO buffer.

If you wish to simply use ePWMxA's output to generate a complementary signal on ePWMxB then I think you should check the "Inverted version of ePWMxA" checkbox.

Regards,
Cody

0 Gergosz2 over 7 years ago in reply to Cody Watkins

Prodigy 55 points

Hi Cody!

Thanks for your answer!

As I have read in the references, the "Inverted version of ePWMxA" option in this case is unneccesary, because the Deadband unit using EPWmxA as source, and using the Active High Complementary option does the invertion.

But you are right, my main issue is that, whitout Enabling EPWMxB, will it reach the the IO buffer? I think I have to disassemble my setup, and measure it directly.

0 venkatesh chilapur over 7 years ago in reply to Gergosz2

Mastermind 6460 points

Hi,

In this case enabling ePWMxB is not required as the source for both RED and FED is ePWMxA.

With your setting we get IN_MODE = 00 and OUT_MODE = 11. Which means ePWMxA will be considered as the source for both RED and FED delay and deadband is enabled on both ePWMxA and ePWMxB output.

0 Cody Watkins over 7 years ago in reply to venkatesh chilapur

TI__Mastermind 42195 points

Has this been resolved?

Thanks,
Cody

0 Gergosz2 over 7 years ago in reply to Cody Watkins

Prodigy 55 points

Yes!

Everything is working now as I expected.

C2000™︎ microcontrollers

C2000 microcontrollers forum

LAUNCHXL-F28069M: Deadband for ePWM in Simulink and EPWMxB enable