• State Resolved
  • Locked Locked
  • Replies 9 replies
  • Subscribers 62 subscribers
  • Views 2448 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.

LM5148: Design Review for a 12V - 20A buck converter and 5V 8A buck converter (LM5085)

Mo Mizzi
Intellectual 290 points
Part Number: LM5148
Other Parts Discussed in Thread: LM5085, LM5146, LM5143

Tool/software:

Hi,

Can you please help review the attached schematic below and see if you notice any issues.

There are two buck converters, the one based off of LM5148 is to give a 12V 20 A output, while the one based off of LM 5085 is expected to give 5V 8A output.

Please feel free to suggest any changes.

Thank you.

batteryBoard_v2b-Voltage Regulator.pdf

  • 0
    TI__Guru* 78725 points

    Hi Mo,

    Thank you for designing with our products.  The responsible engineers are currently out of office, so please allow us some time to review the design and we will try to get back to you in the next few days.

    Best Regards,

    Youhao

  • 0 in reply to Youhao Xi
    Intellectual 290 points

    Thank you for the heads up. We'll look forward to your response.

  • 0 in reply to Mo Mizzi
    TI__Guru* 78725 points

    Hi Mo,

    I checked the LM5085 converter design, and the schematic looks okay, and it confirms the webench design. However, I do have some suggestions.

    • The rectifier diode would have ~6W or more under 70Vin and full load.  Even under 19Vin, it would have ~5W losses. To use the diode, you need to consider good heatsink of the diode.  Or you may consider to use a synchronous buck like the LM5146. 
    • The PFET has a large gate charge, while the LM5085's gate peak current is ~1.5A. The switching losses may become high, too.  Instead, the LM5146 has stronger driver and the NFET with similar RdsON would have smaller gate charge. 
    • You can still proceed with the LM5085 design, but make sure you have adequate cooling for both the MOSFET and diode.

    I will get back to you on your LM5148 design sooner.

    Thanks,

    Youhao

  • 0 in reply to Youhao Xi
    TI__Guru* 78725 points

    Hi Mo,

    Regarding the LM5148 design, my concern is the same for the max power capability.  It does not mean it would not work at all. It is a question about thermal and a reasonable overall efficiency.  To me, the LM5143 look like a more suitable device to support your 240W in dual phases.

    Anyway, I need to consult other engineers and get back to you early next week. Sorry for the delay.

    Thanks,

    Youhao

  • 0 in reply to Youhao Xi
    Intellectual 290 points

    HI Youhao,

    Thank you for getting back and advising. I appreciate you looking into these designs with such detail.

    Just wanted to add that the current draw for the LM5148 is expected to be only 10A (20A is just a safety buffer, which I don't think will ever be used).

    Similarly the LM5085's max current draw expected is under 5A. Let me know if that changes anything.

  • 0 in reply to Mo Mizzi
    TI__Guru* 78725 points

    Hi Mo,

    If the normal load is 10A for the LM5148, and 5A for the LM5085, the situation is remarkably changed, and please allow me to review this again and get back to you next week.  Please remind me if you don't see my respond by Tue next week.  Also, please do not post on this thread now, for it would trigger the e2e clock and I will have to respond in 24 hours. 

    Thanks,

    Youhao

  • +1 in reply to Youhao Xi
    TI__Guru* 78725 points

    Hi Mo,

    Sorry for the delay.  I did some calculations, for the LM5085 circuit, the diode  would dissipate about 4W, and the PFET about 3W.  I would suggest you to reserve heatsinks to cool them.

    Regarding the LM5148,  I think there is am mistake in the COMP and FB connections. Please see below datasheet pin config for 12V output (not the value but the circuit configuration).  Please refer to our design calculator to find the Rcomp, Ccomp and Chf recommendations.

    Also, I would suggest to reduce the switching. At 2.2MHz, you would get higher switching losses. However, if you reduce the frequency, you may need to raise the power inductor's inductrance.

    Hope this helps,

    Best Regards,

    Youhao

  • 0 in reply to Youhao Xi
    Intellectual 290 points

    Hi Youhao,

    Thank you for getting back and for your attention to detail. Also, thank you for catching my mistake.

    I took your advise to use the LM5146 instead and revised the buck regulators. I'm attaching the revised schematics for a one last review from you.

    The first one is a 5V - 8A buck converter, and the other is a 12V - 10A buck converter.

    2746.batteryBoard_v2b-Voltage Regulator.pdf

  • +1 in reply to Mo Mizzi
    TI__Guru* 78725 points

    Hi Mo,

    It seems your schematics came from the Webench tool, and they look good.  You may consider to reserve more input capacitors positions on your PCB in case you need them after test. Please refer to our EVM layout examples when designing your PCB.

    Let me close this thread here.  For new questions, please start a new thread such that our experts who are responsible to support these parts can see directly. 

    Good luck!

    Regards,

    Youhao