TIDM-HV-1PH-DCAC: About the inverter duty cycle generation and normalized values

Damith Pavithra

Part Number: TIDM-HV-1PH-DCAC
Other Parts Discussed in Thread: SFRA

HI

1. In this design, i observed there is a mismatch in the design guide and the actual code.

In the design guide at build level 2: invDuty PU = (invIoRefInst – invIoInst) × Gi/invVbusInst
In the code, at line 715 : invDutyPU= (gi_lead_lag_out + invVoInst)/ (invVbusInst*(VDCBUS_MAX_SENSE/VDC_NOMINAL));

.
(invIoRefInst – invIoInst) × Gi is equal to gi_lead_lag_out right? So, what is the reason of the use of invVoInst in the code while design guide doesnt;

Also when extracting DC bus voltage, why use invVbusInst*(VDCBUS_MAX_SENSE/VDC_NOMINAL Instead of invVbusInst*VDCBUS_MAX_SENSE, which is the correct way to extract the DC bus voltage from the PU values.
Also what is the difference of VDCBUS_MAX_SENSE and VDC_NOMINAL.

2. In the Line 762 and 763, seems like the multipliers are used wrongly.

guiVbus = invVbusInst*VAC_MAX_SENSE;
guiVo = invVoInst*VDCBUS_MAX_SENSE;

As i understand, multipliers of above two should be interchanged, right?

Best Regards
Damith

over 4 years ago

0 John_K over 4 years ago

TI__Expert 4200 points

Hi,

I have checked the code and build 2 has the same equation as shown below

#if (INCR_BUILD == 2)

#if (DC_CHECK==1)
invIiRefInst = invIiRef;
#else
invIiRefInst = invIiRef*invSine;
#endif

if(closeILoopInv==1)
{
// gi_out= DCL_runDF22_C1(&gi, (SFRA_F32_inject(invIiRefInst) - invIiInst));
gi_out= DCL_runPI_C4(&pie, SFRA_F32_inject(invIiRefInst), invIiInst);

invDutyPU= (gi_out+invVoInst)/invVbusInst;
invDutyPU= (invDutyPU>(float32_t)(0.98))?(float32_t)(0.98):invDutyPU;
invDutyPU= (invDutyPU<(float32_t)(-0.98))?(float32_t)(-0.98):invDutyPU;
}
else
{
invDutyPU = 0;
}

#endif

You are right. VAC_MAX_SENSE and VDCBUS_MAX_SENSE have to be changed. However, they have the same value and does not make impact the operation.

Best,

John

0 Damith Pavithra over 4 years ago in reply to John_K

Expert 1125 points

Thank you for the help.
Sorry, i forgot to mention, this is for the grid connected mode inverter. Your one is for standalone mode.
Extracting of DC bus voltage from PU values is now clear.

Only question remaining is the use of the invVoInst in the duty equation. Design guide doesn't use it. But the code does.

Thank you
Damith

+1 John_K over 4 years ago in reply to Damith Pavithra

TI__Expert 4200 points

Hi,

It seems like the documentation has to be updated.

The duty is consist of feedback term(gi_lead_lag) and feedforward term(invVoInst). Please check the link below

https://e2e.ti.com/support/microcontrollers/c2000-microcontrollers-group/c2000/f/c2000-microcontrollers-forum/597846/ccs-hv_1_ph_dcac-control-problem-in-grid-connected?tisearch=e2e-sitesearch&keymatch=invDutyPU#

Best,

John

0 Damith Pavithra over 4 years ago in reply to John_K

Expert 1125 points

Thank you for the directive. That helps..

Regards

Damith

C2000™︎ microcontrollers

C2000 microcontrollers forum

TIDM-HV-1PH-DCAC: About the inverter duty cycle generation and normalized values