UCD3138ACCEVM149: Unable to test UCD3138PFCEVM-026 using UCD3138ACCEVM149

Hello NSW,

I am reviewing your question and I will get back to you soon.

In the meantime, are you sure that your UCD3138ACCEVM149 is working properly? Ex. have you tried testing your control card on a separate EVM, such as the UCD LLC EVM (https://www.ti.com/tool/UCD3138ALLCEVM150). If you do not have another UCD EVM, try running the training labs on the UCD3138ACCEVM149 control card and probe the test points to measure an output PWM signal.

Regards,

Jonathan Wong

Hello Jonathan,

I have tested UCD3138ACCEVM149 control card for PWM using training labs firmware, it is working properly.

Regards,

Hello,

I tried UCD3138ACCEVM149 with UCD LLC EVM, LLC is not working (UCD LLC EVM is working properly with its control card).

Regards,

Hello NSW,

Is the following summary of the problem correct?

1. You have one UCD PFC EVM (UCD3138PFCEVM-026). The UCD PFC EVM comes with the UCD3138 control card (UCD3138CC64EVM-030).
   • This works with the included control card 
2. You have one UCD3138A control card (UCD3138ACCEVM149).
   • This works with the training labs. You can probe a working PWM signal 
   • Deleted 

     MiscAnalogRegs.CLKTRIM.bit.HFO_LN_FILTER_EN = 0;

     per UCD3138A Migration Guide (https://www.ti.com/lit/an/slua741d/slua741d.pdf)
   • This does NOT work with the UCD PFC EVM. 
3. You have one UCD LLC EVM (UCD3138ALLCEVM150). The UCD LLC EVM comes with the UCD3138A control card (UCD3138ACCEVM149).
   • This works with the included control card 
   • This does NOT work with the control card from above #2 

This seems strange because the UCD LLC EVM actually comes with the same UCD3138A control card that you are testing. So, the included LLC control card is the same PCB as the control card that does not work on the PFC controller. If the control card that was included on the UCD LLC EVM works, then this is also a UCD3138ALLCEVM150 control card that you can try flashing the PFC firmware onto and testing on your UCD PFC EVM.

Have you also set the event update window to 132ns as suggested by the UCD3138A migration guide? 

Regards,

Jonathan Wong

Hello 

Yes, the above summary is correct. 

I have not made changes in PFC firmware for event update window, 

as it is already present in UCD3138FW-PFC.

#if ((UCD3138A==1)||(UCD3138064A==1)||(UCD3138A64A==1)||(UCD3138128A==1)||(UCD3138A64==1)||(UCD3138128==1))
	Dpwm1Regs.DPWMEV1.all = 132*4; //aviod 132ns events update window
	Dpwm1Regs.DPWMEV3.all = 132*4; //aviod 132ns events update window
#else //non A version
	Dpwm1Regs.DPWMEV1.all = 290;
	Dpwm1Regs.DPWMEV3.all = 290;//aviod 72ns events update window

Regards,

Hello NSW,

Have you unpopulated jumper J2 and populated J6? If the control card works with the PMBus, that means that power is being supplied from the PMBus cable via J2 and not the UCD PFC EVM.  By populating J6, you will be getting power from the UCD PFC EVM.

If that does not work, then fault 3 may also need to be pulled down to GND

Regards,

Jonathan Wong

Hello, 

I have tested UCD3138ALLCEVM150 again, LLC is operating on both UCD3138A control cards.(one Included with LLC EVM and other standalone control card). Here J6 was populated.

I tried testing PFC EVM (UCD3138PFCEVM-026) with control card (UCD3138ACCEVM149) again after verifying working with LLC EVM, PFC did not worked.

Am I missing something in firmware side while testing UCD3138PFCEVM-026 with UCD3138ACCEVM149 ?

Hello Nikhil,

Apologies for the delay in response. TI USA had a holiday and we are back in office.

Glad to hear that both the UCD3138A control cards work! That shows that the control card and the hardware is working properly and it is something on the firmware side that is not set up properly. So, we are one step closer to solving this .

One thing to note is that the UCD LLC EVM is by default set for the UCD3138A (notice how it is the UCD3138ALLCEVM150). However, the UCD PFC EVM is by default set for the UCD3138. When you are setting up your project in CCS and Fusion Studio, make sure to change the configuration to UCD3138A. In PFCFirmware-1.2 --> PFC --> Firmware --> Device and in PFCFirmware1-2 --> PFC --> Firmware --> PFC --> UCD3138_PFCEVM-026_CCS6 --> Device you will see header files for each UCD device. 

Regards,

Jonathan Wong

Hello,

The build configuration was already changed to UCD3138A  and dumped .xo file from UCD3138A device folder ( PFCFirmware1-2 --> PFC --> Firmware --> PFC --> UCD3138_PFCEVM-026_CCS6 --> Device--> UCD3138A ) in failed test of PFC using UCD3138ACCEVM149.

Regards,

Hello 

In last test I did same steps.

Cleared firmware on CC#1 then dumped new .xo file built for UCD3138 configuration using PFCFirmware1-2, UCD PFC EVM was operating.

Then as per "Step-2" : Used the same .x0 file  built for UCD3138 configuration and dumped onto CC#2, UCD PFC EVM failed.

Regards,

Hello Nikhil,

Can you try doing step 2 with CC#3? Also, when you say that the PFC EVM is failing, can you describe what is failing exactly? For example, are you still getting DPWM signals out of the CC#2 but the power stage is not working or are there no signals coming out of CC#2 at all? I will also look into other steps that you can take and I will get back to you.

Regards,

Jonathan Wong

Hello, 

Thanks for concern,

In step 2 (test PFC EVM with CC#2), UCD3138A was in program mode as desired but DPWM1B signal was not present on control card so PFC was not operating. 

I will update test results of PFC EVM using CC#3.

Regards,

Hello,

I have tested PFC EVM using CC#3 (tried step 2 with CC#3) PFC was operating.

Still, control card CC#2 not working with PFC EVM.

Regards,

Hello Nikhil,

Glad that CC#3 is working. Are any other output signals working with CC#2? Or, can you also set another GPIO or PWM signal to trigger when DPWM1B is supposed to trigger and measure that?

Regards,

Jonathan Wong

Hello, 

I have probed signals AC_L and AC_N, which were coming normal at AD07 and AD08 pin respectively on CC#2, hence Fault 1 pin should go high but it remains low (Fault 1 pin indicates AC_good ) and DPWM1B does not enable until Fault1 goes high.

As I am using classic boost topology on PFC EVM, DPWM1B drives the switch Q3 so can not set another PWM.

Regards,

Hello,

I have checked ADC resistors they were good as per schematic. 

I have again tested PFC EVM with CC#2, and in the process I found fault 1 pin got high for some seconds, which turned out to be due to loose connections (in some pins) of CC#2 when connected to PFC-EVM (unlike LLC-EVM, where it connects properly)

Thanks for the help.

Regards,

Hello Nikhil,

Glad your control cards are working. Thanks for the update and thanks for choosing TI's UCD3138 products. If you have any further questions do not hesitate to create a new thread on E2E. 

Regards,

Jonathan Wong