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.

WEBENCH® Tools/TPS54340B-Q1: Op-Val & Bode plot cross freq. not matching for TPS54340B-Q1 based design.

Part Number: TPS54340B-Q1

Tool/software: WEBENCH® Design Tools

Hi TI support,

I recently made the design using TPS54340B-Q1 with details as under,

VIN=12-18V, Vout=4.0V, IOUT=2.5A max, Tamb=60'C, Freq=918KHz. I replaced the compensation components as per calculation from datasheet with cross freq of 29Khz. Now Op-Val shows correct cross over freq. after replacement of compensation components, however bode plot simulation doesn't shows those values. 

Can you please check this. If some email ID is available I could share the design.

Regards,

jagdish

  • Hello Jagdish,

    Please can you email robert.blattner@ti.com with your question. I would include what simulation you are using?


    Thanks.

    David.
  • Hi David,

    Please note that I have shared the design with above said email ID with details written to it.

    Regards,
    jagdish
  • Hi Jagdish,

    Can you please share you design with "Share with Public" option? If you can share the design, we can check if you are running into any of the 3 issues mentioned below.

    Some reasons why the opvals and bode plot sim might not match:

    1. If Cout has been changed, is it a custom cap? If so, the “Derated Cap”, “Total Derated Cap” and “Cap” all need to match up and set the number of parallel caps to 1. This is an issue with custom cap form today. 
    2. Bode plot simulation runs at VinNom = (VinMin+VinMax)*0.5. so need to make sure to recalculate the “Opvals” at VinNom to compare with bode plot sim values.
    3. Also, when using “Recomp”, make sure the power stage components has the correct derated value of Cout that is expected.

    Regards,
    Amod 

  • Hi Amod,

    Please my response as below,

    => Design is shared with link details as below,
    webench.ti.com/.../SDP.cgi

    => Yes Cout is changed to ceramic, these are 2 in parallel, I checked total derated cap matches to value in Re-comp tool. However, parallel caps aren't set to 1. I will check this.

    => As of now I am not using Re-comp tool.

    Thanks,
    jagdish
  • Hi Amod,

    Tried using single capacitor of 94uF instead of 2X47uF but results are same. Cross over in bode plot is 16K while in Op-Val is 28K.

    Awaiting your further response.

    Thanks,
    jagdish
  • Thanks for the additional info Jagdish. I am taking a look at your design.

    Regards,
    Amod
  • Hi Jagdish,

    It seems the issue is arising due to discrepancy in the sim and opval model parameters. Someone from our team is looking into it and will get back to you with further updates. Apologize for the inconvenience caused and appreciate your patience on this matter.

    For now, I would recommend using the opvals and recomp values to compensate and confirm the results and stability of your design.

    Regards,
    Amod

  • Hi Amod,

    Thanks for your response. OK, I will await the results from your team.

    However, using recomp components as per datasheet equations results in stable design as per opval, so I will proceed ahead with this as of now. I believe you are talking same " recomp values to compensate and confirm the results and stability of your design".

    Do I understand correct, please advice.

    Regards,
    jagdish
  • Hi Jagdish,

    Seems like the difference between Simulation Bode plot & Opval Bode Plot was due to edited Rload (20ohm).

    Could you please let us know if you had manually changed Rload to 20ohm ?

    Note: Any changes made to Input source & Rload in the schematic will not affect Opvals. Hence we might be observing discrepancy.

    I made following changes in the schematic (Below is the shared design FYR) :

    https://webench.ti.com/appinfo/webench/scripts/SDP.cgi?ID=D9FCFEA65079651C

    1. Cout : Made Cap, Total Derated cap and Derated Cap as 52uF with Quantity = 1

    2. Rload: Changed to 1.6ohm [Iout = 2.5A]

    Thanks & Regards,

    Harish

  • Hi Harish,

    Even without changing Rload in bode plot fco was 16K (in simulated bode plot) rather than 28K. After changing Rload to 20E it gave fco near 5K.

    BTW, I can't open the link, nor I can see shared design in my list. Can you please look into this.

    Also while checking in opval, I was just changing values in top tab of op-val (VIN & IOUT) & not checking there after doing any schematic modification. This is FYI.

    Regards,
    jagdish
  • Hi Jagdish,

    Could you please try below link ?

    https://webench.ti.com/appinfo/webench/scripts/SDP.cgi?ID=D9FCFEA65079651C 

    Yes, we should be updating Vin & Iout = (Vout / Rload) in Opval accordingly & then click on recalculate as well.

    But there isn't much change in Cross-over frequency & phase margin in Opval as contrast to SPICE Simulation. We recommend to use SPICE simulation as reference especially for lower Iout as it considers DCM mode.

    This is because the Opvals doesn't consider DCM mode for this device & hence are not very accurate for lower load.

    FYI - As seen from below image, the Gain of power stage increases for lower Iout but since gain-bandwidth product is almost constant hence, the overall cross-over is not getting affected.

    Thanks & Regards,

    Harish

  • Hi Harish,

    Thanks for sending the link, I could see 28.2K Fco at 1.6E load & this matches datasheet value. I agree for lower load op-val is not correct since datasheet also consider considers full load during fco calculation & that will change with load.

    I will also be adding enable resistors (not shown in schematic) as per datasheet equations.

    Harish: Do you have any other inputs for this design as this will be used in mass production or you see it fine. Please advice.

    In general, please get correct all these errors in Webench particularly with custom components.

    Regards,
    jagdish
  • Hi Jagdish,

    We would recommend you to build some evaluation or test boards to ensure the performance is as expected for your operating range before mass production.

    The analytical solutions we use in the tool cannot substitute actual measurements done on a board.

    Yes, we have a plan to fix errors w.r.t custom components. Thanks for your feedback.

    Thanks & Regards,

    Harish

  • Hi Harish,

    We will certainly do the actual measurements but idea here was to do characterization with everything included after review so I asked if you see anything else to be included or we proceed ahead. Kindly advice so we can move towards actual protos.

    Thanks,
    jagdish
  • Hi Jagdish,

    Thanks for your patience.
    The design seems good to us. Kindly proceed with your proto-types.
    Please feel free to reach us in case of any further queries.

    Thanks & Regards,
    Harish