• State TI Thinks Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 4 answers
  • Subscribers 63 subscribers
  • Views 220 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

LM5146-Q1: Buck convertor schematic review

Reed Clardy
Prodigy 10 points
Part Number: LM5146-Q1
Other Parts Discussed in Thread: LM5146, LM5145

Tool/software:

I am designing a 48V 12V buck convertor based off the WEBENCH Design Report using the LM5146QRGYRQ1 switcher chip. There were two instances were it said MOSFET quantity 2 and I want to make sure thta those were connected correctly.

Thank you,

Reed Clardy

48V_12V_sch.SchDoc

  • 0
    TI__Guru 73752 points

    Hi Reed, 

    Please send the schematic PDF and a completed quickstart calculator https://www.ti.com/tool/LM5145DESIGN-CALC (the quickstart is useful to understand the design specs, check component selection, power losses, etc.) 

    Regards,

    Tim

  • 0 in reply to Timothy Hegarty
    Prodigy 10 points

    Below I have attached the schematic as a pdf as well as the benchmark file I was following. Is it possible for you to send me the EVM file? I had a problem with the quick start calculator it said that my values for Cin and Cout did not fit with my voltage ripple for input or output but I was informed that our voltage ripple was 1%, I have attached the quickstart calculator filled out to the best of my ability.

    Thank you for your ongoing assistance,

    Reed

     48V_12V_with_EMI_filter_WBDesign12.pdf4452.LM5145_quickstart_calculator_A5.xlsm48V_12Vsch.pdf

  • 0 in reply to Reed Clardy
    TI__Guru 73752 points

    Hi Reed,

    The EVM layout files are available in the LM5145 and LM5146 product folders.

    No need for 100V high-side FET here, as suggested by WEBENCH, as Vin-max = 55V. A Vds rating of 80V should be fine, especially as it provides better efficiency. Take a look at the 80V FETs used in the LM5145 EVM.

    Further comments:

    1. The buck inductance is somewhat low, as the ripple current is 52% (prior to inductance derating with current).
    2. The current limit should be higher than the full-load current. If Iout-max = 17A, then set it at 22A for some margin.
    3. The output caps should be 22uF/25V/1210/X7R for a 12V output (see the datasheet or the LM5146 EVM for part numbers). 3 x 4.7uF is too low for 17A. Make sure to derate the ceramic value for voltage, as this affects compensation.

    Regards,

    Tim

  • 0 in reply to Timothy Hegarty
    Prodigy 10 points

    Howdy Tim,

    Thank you for your continued assistance. I swapped the buck inductance for the 6.8 mH inductor in the LM5146 EVM. I swapped the high side FET for the 80V in the LM5145 EVM. I swapped the 3 x 4.7uF capacitors with 3 x 22 uF capacitors, is this the correct implementation?

    Thank you,

    Reed

  • 0 in reply to Reed Clardy
    TI__Guru 73752 points

    Hi Reed,

    These look like good updates. If needed, send on the updated schematic and quickstart calculator for review.

    Regards,

    Tim

  • 0 in reply to Timothy Hegarty
    Prodigy 10 points

    The values for the Cin are still too small, in the quickstart calculator it uses 5 capacitors but in the WEBENCH it only specifies 3. Below is the current schematic and the quickstart calculator updated.

    Thank you, 

    Reed

    1856.LM5145_quickstart_calculator_A5.xlsm48V_12V_new_cout_newFets.pdf

  • 0 in reply to Reed Clardy
    TI__Guru 73752 points

    Hi Reed,

    I doubt that TDK inductor is sufficient. The DCR is too high, giving high power loss at 22A, especially as it's only rated for 10A in terms of temp rise. Take a look at Cyntec VCUD128T.

    The FETs should be asymmetric - higher Rdosn for the high-side FET given the 25% duty cycle (and lower capacitance for better switching losses), then lower Rdson for the low-side FET as it conducts for 75%.

    If full load is 22A, the current limit should be set at >27A for margin.

    Input caps should be 4.7uF/100V/1210.

    Add more Cout as well given the high Iout, and make sure to derate the value for voltage (a 22uF/25V cap at 12Vout is effectively ~8uF). 66uF in the quickstart is incorrect and will result in incorrect compensation values.

    The power losses shown in the quickstart file are way excessive, so it's critical to select components to achieve 2-3W power loss (each) at full load - otherwise the temp rise will be too high.

    Regards,

    Tim