This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LM25118: VCCX pin, Webench Design Clarification

Genius 14159 points
Part Number: LM25118
Other Parts Discussed in Thread: CSD17309Q3

Hi Experts,

I am looking at a design with the LM25118 power controller. Cx have a design from Webench , doubting one of the connections in that design. Could you please confirm that connection is correct?
PDF design is shared below from Webench for this specific question.
 
Looking at the VCCX pin, it is tied to the output. Based on the description for this pin "Optional input for an externally supplied bias supply. If the voltage at the VCCX pin is greater than 3.9 V, the internal VCC regulator is disabled and the VCC pin is internally connected to VCCX pin supply. If VCCX is not used, connect to AGND."
 
And looking at the reference design in the datasheet where there is an "N/A" resistor on that line to the output, I am believing this pin should have no connection but would really like to confirm that with an expert familiar with the chip before my customer commit a PCB.

For your confirmation.

Regards,
Archie A.

web bench design.pdf

  • Hello Archie,

    Thank you very much for this question. Webench is showing a wrong connection of VCCX to VOUT. I will take care that this will be updated.

    Please recommend the customer to either leave VCCX open or connect to GND whichever is easier in his layout.

    Best regards,
    Brigitte

  • Hi Brigitte,

    Thanks for your confirmation. Customer needed assistance with an expert:

    In light of this Webench error, I think it would be prudent to ask an engineer/expert to look over the complete design to see if there are other errors coming out of Webench. Could you please arrange that? Again, the parameters for this supply are as follows:  8 to 36vdc input, 12vdc at up to 6 amps output.
    Also, attached are my BOM and corrected Schematic for the TI LM2118 power supply.  As you know, it is really hard to find FETs in inventory at major suppliers. I am wondering if we could use the CSD17309Q3 for both M1 and M2 I want to use this case style for both FETs and the BSC037N08NS5ATMA1 for M1 is hard to find while the other alternatives listed by the Webench are out more than a year for delivery.
    Surely, there must be some other TI FETs we can use that are actually available.
    Any advice? Thanks again.
    Regards,
    Archie A.
  • Hello Archie A.

    There should not be any further issue with Webench, but as a first step checking the design with the quickstart calculator would help. Please download here: https://www.ti.com/lit/zip/snvu065 and then check if your design fulfills the requirements given there. Be careful with 6A current, you might get into trouble with the power dissipation in the diodes.

    Best regards,
    Brigitte

  • Hi Brigite,

    Cx: Thank you for the spreadsheet.  However, that has really thrown a wrench into the works.  The spreadsheet is coming up with wildly different values AND the components recommended by the Webench are not within the current ranges that the spreadsheet shows as minimum... For example, the spreadsheet is telling me that using a 4.7uH inductor, the minimum current spec is 32.7 amps but the inductor spec'd by the Webench is a 24 amp inductor.  Similarly, the sense resistor specified by the spreadsheet is spec'd at a minimum of 2.5W while the one coming out of Webench is a 2W resistor (and the largest made in that series).  Another example is that

    I'm not sure where to go from here.  Again, I think I need one of your engineers to confirm the Webench design which I am not able to confirm myself with the spreadsheet you shared.

    This is the link to my design: https://webench.ti.com/appinfo/webench/scripts/SDP.cgi?ID=79D518BCBDB141C8

    I have attached the spreadsheet where I completed as many of the inputs as I could.  Please note the second tab, BOM, and the current values I mentioned above.

    I had use the substitute M1  BSC037N08NS5ATMA1 as the recommended part is out of the supply chain for at least a year.  I would still like to find a less expensive FET for here, but can live with this.

    Also, I am spec'ing E192 (0.1%) resistors for the feedback resistor divider using 8K760 and 1K000 values which gets me much closer to the target Vout of 12.0V than the 1% values would (Vout 11.88).

    Just for consistency, I also attached a spreadsheet of the BOM I am using.

    I'm stuck here until we can sort out these inconsistencies.  Should one believe Webench or not?

    Thank you VERY MUCH for any help you can provide here.  My PCB design using this power supply (see that section in photo below) is complete and I am very excited to use this power supply design if I can just sort out these inconsistencies before we commit to the proto production scheduled for this week.

    Regards,
    Archie A.

    TI Calculator snvu065a.xlsm

    ...

  • Hello ArTzy,

    I am sorry, but I cannot find the attachments you mentioned, sorry for the inconvenience.

    I checked the Webench design and it looks good just VCCX connection is wrong which will be updated soon.

    For the peak inductor current, Webench is taking care that the design does not get too big. If you review the inductor behavior over current in the inductor datasheet, the inductance just changes slowly, so it can be used for higher peak currents than the saturation current value given in the table in the datasheet. The spreadsheet just gives a calculated value without checking the behavior of the used inductor.

    But please always keep in mind that a tool only can give a starting point for a power supply. Measurements on the real design are essential, as a tool does not include all parasitic components as well as the layout parasitics. Layout can make and break a power supply.

    Best regards,
    Brigitte

  • Hi Brigitte,

    Thanks for your response. I updated the attachment from my previous comment.

    Regards,
    Archie A.

  • Hello ArTzy,

    Thank you for adding the attachment. Please tell me if there are further questions. Otherwise I will close the thread.

    Best regards,
    Brigitte