TMS320F28377D: The difference between DB_ACTV_HIC and DB_ACTV_HI when initial ePWM

Tobby Guo

Intellectual 770 points

Part Number: TMS320F28377D
Other Parts Discussed in Thread: C2000WARE

Hi:

when USE DB_ACTV_HIC to initial ePWM , a unexpected plus occur on EPWM1B when execute initial, the display as below:

CH1: Trigger signal CH2:EPWM1B, EPWM1A is always low( do not display on scope)

the initial code:

EPwm1Regs.AQCTLA.bit.CAU = AQ_CLEAR;
EPwm1Regs.AQCTLA.bit.CAD = AQ_SET;
EPwm1Regs.AQCTLB.bit.CBU = AQ_CLEAR;
EPwm1Regs.AQCTLB.bit.CBD = AQ_SET;

EPwm1Regs.AQSFRC.bit.RLDCSF = 3; //The active register load immediately
EPwm1Regs.AQCSFRC.bit.CSFA = FORCE_LOW; //FORCE the EPWM2A output LOW
EPwm1Regs.AQCSFRC.bit.CSFB = FORCE_HIGH; //FORCE the EPWM2B output HIGH

EPwm1Regs.DBCTL.bit.IN_MODE = DBA_RED_DBB_FED; //EPWMxA In is the source for rising-edge delayed signal.
EPwm1Regs.DBCTL.bit.DEDB_MODE = 0;
EPwm1Regs.DBCTL.bit.POLSEL = DB_ACTV_HIC; //Active high complementary (AHC). EPWMxB is inverted.
EPwm1Regs.DBCTL.bit.OUT_MODE = DB_FULL_ENABLE; //Enable dead band time configuration
EPwm1Regs.DBCTL.bit.OUTSWAP = 0;
EPwm1Regs.DBRED.all = EPWM1_DBRED;
EPwm1Regs.DBFED.all = EPWM1_DBFED;

then change to DB_ACTV_HI, the display as below:

CH1: Trigger signal CH2:EPWM1B, EPWM1A is always low( do not display on scope)

the initial code:

EPwm1Regs.AQCTLA.bit.CAU = AQ_CLEAR;
EPwm1Regs.AQCTLA.bit.CAD = AQ_SET;
EPwm1Regs.AQCTLB.bit.CBU =AQ_SET ;
EPwm1Regs.AQCTLB.bit.CBD = AQ_CLEAR;

EPwm1Regs.DBCTL.bit.IN_MODE = DBA_RED_DBB_FED; //EPWMxA In is the source for rising-edge delayed signal.
EPwm1Regs.DBCTL.bit.DEDB_MODE = 0;
EPwm1Regs.DBCTL.bit.POLSEL = DB_ACTV_HI; //Active high complementary (AHC). Neither EPWMxA nor EPWMxB is inverted (default).
EPwm1Regs.DBCTL.bit.OUT_MODE = DB_FULL_ENABLE; //Enable dead band time configuration
EPwm1Regs.DBCTL.bit.OUTSWAP = 0;
EPwm1Regs.DBRED.all = EPWM1_DBRED;
EPwm1Regs.DBFED.all = EPWM1_DBFED;

Between the two plans ,the first give EPWM1B force high then inverter in DBCTL[POLSEL], the second give EPWM1B force low then Neither EPWMxA nor EPWMxB isinverted in DBCTL[POLSEL].

The initial code is execute in sequence, the two plans should have same output on EPWM1B, why the fisrt have a 20us high pluse.

Addition information:

fpwm:20kHz

EPwm1Regs.CMPCTL.bit.SHDWAMODE = CC_SHADOW;
EPwm1Regs.CMPCTL.bit.SHDWBMODE = CC_SHADOW;
EPwm1Regs.CMPCTL.bit.LOADAMODE = CC_CTR_ZERO_PRD;
EPwm1Regs.CMPCTL.bit.LOADBMODE = CC_CTR_ZERO_PRD;

Thanks

over 6 years ago

0 Tobby Guo over 6 years ago

Intellectual 770 points

Supplement：

In the first plan , EPWM1A and 1B have same value and action, use DBCTL[POLSEL] to inverte EPWM1B, and use DBRED and DBFED to set DEADBAND time.

In the second plan , EPWM1A and 1B have same value and Complementary action, use +/- Dtime value to set DEADBAND time.

Prefer the first plan ,but can not accept the unexpected high pluse when initial.

Thanks

0 Subrahmanya over 6 years ago in reply to Tobby Guo

TI__Mastermind 24095 points

Hi,

Can you refer this post and let me know if it helps?
https://e2e.ti.com/support/microcontrollers/c2000/f/171/t/838289?TMS320F280049-Active-high-complementary-PWMs This seems to refer to the same issue you've explained here, if i understand correctly.

0 Tobby Guo over 6 years ago in reply to Subrahmanya

Intellectual 770 points

HI：

Thanks for you reply.

As my post above , i use DB_ACTV_HI to dismiss the unexpected plus.But about DB_ACTV_HIC, before EPwm1Regs.DBCTL.bit.POLSEL = DB_ACTV_HI, i set the EPwm1Regs.AQCSFRC.bit.CSFA = FORCE_LOW; EPwm1Regs.AQCSFRC.bit.CSFB = FORCE_HIGH;.

The initial code is execute in sequence, so i think the high plus shouid be the execute time of the two code between . The execute time should less than 20us , why the high plus have 20us

the post you suggest is less than 20us

0 Subrahmanya over 6 years ago in reply to Tobby Guo

TI__Mastermind 24095 points

Hi,

The deadband configuration, be it active high or active high complimentary (edge based configuration/delay) would come into effect upon PWM edge occurrence.
EPwm1Regs.AQCSFRC.bit.CSFA/B forces would take effect immediately. It's not clear to me why you had to do continuous software force in the initial configuration.
Also, in case you use CSFA/B but NOT configure PWM to be EPwm1Regs.DBCTL.bit.POLSEL = DB_ACTV_HI and only use DB_ACTV_HIC, do you still see the same results?
Can you also cross check configuration w.r.t. example below?
C:\ti\c2000\C2000Ware_2_01_00_00\device_support\f2837xd\examples\cpu1\epwm_deadband\cpu01
EPWM3 uses DB_ACTV_HIC.

0 Tobby Guo over 6 years ago in reply to Subrahmanya

Intellectual 770 points

Thanks for your reply.

C2000™︎ microcontrollers

C2000 microcontrollers forum

TMS320F28377D: The difference between DB_ACTV_HIC and DB_ACTV_HI when initial ePWM