• State Resolved
  • Locked Locked
  • Replies 12 replies
  • Answers 6 answers
  • Subscribers 63 subscribers
  • Views 407 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.

LM5157: Can LM5157 boost converter output be tied together?

Kazuya Nakai59
Guru 21575 points
Part Number: LM5157
Other Parts Discussed in Thread: LM5122

Tool/software:

Hello guys,

One of my customers is considering using LM5157 for their next products.
Their power supply requirements are the follows.

Vin=5V +/-5%,
Vout=+24V
Iout= 2A~2.5A

They selected LM5157 because the device has a internal switching FET.
They know that a single LM5157 cannot supply this current, so they are trying
to tie the outputs of the LM5157 boost circuits (Figure 9-3. Typical Boost Application) together
and drive them in parallel.

Is this usage no problem?

Your reply would be much appreciated.

Best regards,
Kazuya.
 

  • 0
    TI__Genius 15575 points

    Hi Kazuya,

    Thanks for reaching out.

    Yes, it is possible to parallel two LM5157.

    There is also the possibility for external clock synchronisation. Wit this feature you could achieve two interleaved phases.

    Best regards

    Moritz

  • 0 in reply to Moritz
    Guru 21575 points

    Hi Moritz,

    Thank you very much for your reply.

    Could I ask you an additional question?

    Q.
    They want to use 4pcs of LM5157 with multi phase topology（4 boost output will be tied together).
    Also they think to use 90 degree phase shift for lower output ripple?
    Is 90 degree phase shift no problem for LM5157?

    Thank you again and best regards,
    Kazuya.

  • 0 in reply to Kazuya Nakai59
    TI__Genius 15575 points

    Hi Kazuya,

    Yes, indeed that would be possible.

    But keep in mind that if all outputs are just connected together, also the COMP pins need to be connected and only one FB can be connected to the output voltage. Otherwise only one device would take the full power most likely.

    Best regards

    Moritz

  • 0 in reply to Moritz
    Guru 21575 points

    Hi Moritz,

    Thank you very much for your reply.

    I understood that FB must be connected between LM5157s.
    But is it no problem to tie COMP together?
    I worry a little bit about phase margin and a interference between the GM amplifiers
    when COMP pins are tied together.
    Also have you ever tried connecting the outputs of LM5157 in parallel?
    And do you have any schematic for parallel driving of LM5157?

    Thank you again and best regards,
    Kazuya.

  • 0 in reply to Kazuya Nakai59
    TI__Guru 65300 points

    Hello Kazuya,

    Sorry, but if you connect all FB pins together, the phases will not share the current. You need to use the FB pin of only 1 device.

    You need to connect than the COMP pins together to make sure that all devices are operating at roughly the same duty cycle.

    As far as I know this circuit is not tested as other customers are using a synchronous boost for similar requirements.

    Best regards,
    Brigitte

  • 0 in reply to Brigitte
    Guru 21575 points

    Hello Moritz, Brigitte,

    Thank you very much for your strong supports.

    I want to confirm you the following point again.

    Q1.
    Is it possible to tied four LM5157 outputs together for parallel driving to increase the output current?
    Or is it not possible because each output current can not be averaged? 

    Q2.
    If the answer to Q1 is four LM5157 outputs can be tied together,
    could you please tell me how to connect "FB" and COMP" of each LM5157 and how to use SYNC again?

    The FB pin of only 1 device needed to used. -> How to treat FB of other tree LM5157?
    Are these three FBs needed to higher voltage level than VREF?
    Are COMP pins needed to tied together?
    Please tell me how I should input SYNC signal for phase shift of 4 pcs of LM5157?  

    Thank you again and best regards,
    Kazuya.

  • 0 in reply to Kazuya Nakai59
    TI__Genius 15575 points

    Hi Kazuya,

    Q1: It is possible,however, we have no reference design for it and it is not tested.

    Q2: One FB pin is connected to the output voltage with a resistor divider. The other 3 FB pins then have to be connected to GND. The COMP pins all have to be connected together to achieve the same dutycycle for each device.

    Chapter 8.3.6 in the datasheet explains how to use the external clock synchronisation. For a 4phase operation, the phaseshift from phase to phase should be 90 degrees.

    But as Brigitte mentioned, it would be a good idea to use a sync boost for this application, the the requirements can be achieved with one or 2 phases and there are tested reference designs.

    Best regards

    Moritz

  • 0 in reply to Moritz
    Guru 21575 points

    Hi Moritz,

    Thank you very much for your supports.

    Could you tell me URL of the tested reference design documents because I'd like to inform the URL to the customer.

    Thank you again and best regards,
    Kazuya. 

  • 0 in reply to Kazuya Nakai59
    TI__Genius 15575 points

    Hi Kazuya,

    Please find the reference design here: https://www.ti.com/tool/PMP22819

    Best regards

    Moritz

  • 0 in reply to Moritz
    Guru 21575 points

    Hi Moritz,

    Thank you very much for your telling the URL.

    In the document, FB pin of sub LM5122 is connected to VCC.
    In LM5157 case, which must FB pin of sub LM5157 be connected to VCC or GND?
    The customer wants to use converter like LM5157 not controller.

    Do you have any tested reference design of synchronous converter?

    Thank you again and best regards,
    Kazuya. 

     

  • 0 in reply to Kazuya Nakai59
    TI__Genius 15575 points

    Hi Kazuya,

    For LM5157 the FB of subs have to be connected to GND.

    Unfortunately there is no comparable sync boost converter, so also no reference design.

    Best regards

    Moritz

  • 0 in reply to Moritz
    Guru 21575 points

    Moritz, Brigitte,

    Thank you very much for your supports.
    I will tell your reply to the customer.

    Thank you again and best regards,
    Kazuya.