TMS320F28P550SJ: Ask help for TMS320F28P550 Implement full bridge LLC + SR PWM configure solution

Hello,

I will recommend you the reference design based on what mode of control you want.

If you need voltage mode control, you can refer TIDA-010062 

If you need current mode control you can refer hybrid hysteretic control HHC in PMP41081 and TIDA-010062 

Both of these design are based on half bridge LLC and two switch based SR. For full bridge, second leg is complimentary for the original half bridge leg. SR side remains same unless you are using full bridge type of rectification.

Here are app notes to explain this scheme as well: 

1. https://www.ti.com/lit/an/spradf4/spradf4.pdf

2. https://www.ti.com/lit/an/spradj1a/spradj1a.pdf 

Regards,

Sumit

Now I am study the LAUNCHXL F28P55X HRPWM function, I am try to create the LLC full bridge HRPWM drive, but I meet a issue that E1PWM OUTA TON not equal to OUTB ton, but E2PWM OUTA TON is equal to OUTB ton.

Below is the details info, could you please help to provide some suggestions.

LLC frequency = 1MHZ.

TBCLK = 150MHZ

Use up-down counter mode.

So the period = 75.

HRPWM with SFO().

AC configure like below, so only compare A impact the duty:

EXPWMA and EXPWMB output is complementary:

Both Period and compare A are implemented high resolution, I can use CMPAHR for duty control.

As above, when the period is 0x4B, COMPA is 0x00257F,  CMPA is 50% of the period, as the FED delay and RED delay are the same, so the ton time of OUTA and OUT B should be the same.

But unfortunately, OUTA ton time is greater than OUTB ton time, could you please provide me some suggestions.

However i also configure EPWM2 same as EPWM1, from EPWM2, the OUTA ton time is equal to OUTB ton time, it is very strange.

Below is the waveform of my test results:

Purple: EPWM1 OUTA, Blue:EPWM1 OUTB,

OUT A ton = 436ns, OUTB TON= 430ns, this is not measurement error, I have double check many times.

Yellow: EPWM2 OUTA, Green:EPWM2 OUTB,

OUT A ton = 436ns, OUTB TON= 436ns, this is the expect result,

Did you refer to HRPWM example in C2000ware at following location to see if there is any configuration that you may have missed. Especially how duty and period is calculated:
C:\ti\c2000\C2000Ware_6_00_00_00\driverlib\f28p55x\examples\hrpwm

Choose appropriate example based in your requirement depend on counter mode and dutyhr / periodhr control.

Regards,

Sumit

Hello, i have use the example for test, it have the same issue.

Simon, 

I have received your config file in email via our FAE. I will try recreating this in our lab and will get back to you in next week. We appreciate your patience!

Regards,

Sumit

Simon, 

I am still looking into this and re-creating issue. Meanwhile, I appreciate your patience. I should have reply for you by end of this week.

Regards,

Sumit

Thanks Sumit, can you repeat the same issue on your side.

I used your sysconfig to recreate the issue. When you load TBPRD= 75 in sysconfig, did you change the following highlighted as well considering this used in duty and HRPWM factor calculations. Value is 76 because of the up counter mode formula? When I changed that, it fixed the issue as shown in scope shot below. Could you please try this and let me know if this resolved your issue?

Regards,

Sumit

when you change the period to 76, the CMPAHR will change to 0, so the issue will disapear, if you change the period to 75 or 77 then the CMPAHR will non-zero, then the issue will repeat.

Could you please elaborate your question? Like under what condition you are getting asymmetric duty? Also when you change TBPRD are you also changing corresponding duty HR factor?

Regards,

Sumit