• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 2 answers
  • Subscribers 66 subscribers
  • Views 551 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

Original question:

LM5175: lower efficiency than expected

LM5175: schematic and layout review support

Bruce li
Genius 5580 points
Part Number: LM5175
Other Parts Discussed in Thread: LM5176, LM5122,

Dear TI Experts,

My customer designed schematic and pcb based on webench with 9~27V--Vin,42V/2A--Vout. After finished prototype, they measured key parameter and below questions:

1) When Vin=9V, Vout=42V, add 2A load, the efficiency is only 84%, is this the best efficiency by calculating the condition theoretically?

2) From the schmatics generated by webench , PIN24--BIAS seems over 40V, yet in the datasheet it suggests that "The BIAS pin voltage must not exceed 40 V", is this a potential risk?

3) When load removed and output opened, the power consumption is about 1.2W, is it normal? Since from the datasheet, the Iq of normal mode is only 1.65~4mA.

About all , do you think LM5175 is the best option for such application? Any better choice? Thanks a lot.

Vin=9~30V

Vout=27~42V

efficiency > 90%

4807.INPUT 9-30V OUTPUT 42V2A.pdf

5773.TTIBBMSGEN2P010.pdf

LM5175_42V_2A by webench.pdf

  • 0
    TI__Mastermind 40080 points

    Hi Bruce,

    The expert of this part is on holiday. They will address your issue after the new year. Please wait in patience.

  • 0 in reply to Helen chen
    Genius 5580 points
    Dear Helen,
    Got it, thanks for your notice
  • 0
    TI__Mastermind 21720 points
    Hi Bruce,

    Thanks for reaching out to us.

    Below please see my responses to you questions.

    1) When Vin=9V, Vout=42V, add 2A load, the efficiency is only 84%, is this the best efficiency by calculating the condition theoretically?

    This is heavily dependent on the MOSFET, and inductor selection of the design. The LM5175 quick-start calculator can be used to estimate the efficiency of the circuit based on the component selection.
    www.ti.com/.../lm5175quickstart-calc


    2) From the schmatics generated by webench , PIN24--BIAS seems over 40V, yet in the datasheet it suggests that "The BIAS pin voltage must not exceed 40 V", is this a potential risk?

    The BIAS pin should be below 40V to ensure that the device is not damaged.

    3) When load removed and output opened, the power consumption is about 1.2W, is it normal? Since from the datasheet, the Iq of normal mode is only 1.65~4mA.

    What is the input voltage? Is the output voltage still in regulation, just not loaded? Since the LM5175 is configured to operate in CCM at light load 1.2W seems reasonable. When operating in CCM the switching losses and conduction losses of the MOSFET as well as the losses in the other components must be considered. 1.2W seem reasonable.

    As for other components, I would recommend the LM5176. This is the next generation of the LM5175 and has some performance improvements. The other option would be to use a boost converter like the LM5122 since this design only calls for a the boost operation of the LM5175.

    Please let me know if you have any questions.

    Thanks,

    Garrett
  • 0 in reply to Garrett Roecker
    Genius 5580 points
    Dear Garrett,
    Got it, thanks a lot for your support and recommendation. Here are another several questions I want to consult with you again,

    1) For LM5176, is it P2P compatiable with LM5175?
    2) Do you think LM5176 will achieve better efficiency than LM5175?
    3) Is there any improvement suggestion from customer's layout attached ?
  • 0 in reply to Bruce li
    TI__Mastermind 21720 points
    Hi Bruce,

    1) For LM5176, is it P2P compatible with LM5175?
    Yes it is P2P compatible. Please be aware that the oscillator resistor will change.

    2) Do you think LM5176 will achieve better efficiency than LM5175?
    The efficiency should be fairly close for both devices.

    3) Is there any improvement suggestion from customer's layout attached ?
    The layout is alright. Here a few small suggestions.
    1. The gate drive traces going to QT23 look like they run under a switch node and are no shielded by a ground layer. I recommend routing these are around the hi dV/dT signals

    Please let me know if you have any questions.

    Thanks,

    Garrett
  • 0 in reply to Garrett Roecker
    Genius 5580 points

    Dear Garrett,

    Thanks a lot for your support. Customer has measured QT23 and found some abnormal spike on the falling and rising edge. They compared the LM5175EVM same test points and found similiar result.

    Do you think these spikes have some potential risks which may influence board's stability and safety? Thanks a lot.

  • 0 in reply to Bruce li
    TI__Mastermind 21720 points
    Hi Bruce,

    Are you probing the signal with a short ground lead? If the ground lead is long noise can be injected onto the measured signal.

    Thanks,

    Garrett
  • 0 in reply to Garrett Roecker
    Genius 5580 points
    Dear Garrett,
    Thanks a lot for your help, the spike reduced a lot after customer increased the input voltage