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.

LM5117EVAL: I would like to reconfigure the eval board to output 5V.

Part Number: LM5117EVAL
Other Parts Discussed in Thread: LM5117


I have received the evaluation board p/n: LM5117EVAL and would like to set it up to output 5V. I have calcualted the feedback resistors should be 357 ohms for  R_FB1 (R16) & R_FB2 should be 1870 Ohms (R17).

I wanted to ask if there is anything else I am missing or if I should not use this board for that voltage output? I did the web bench designed and it does not list chip given my input parameters of 8V < Vin < 56V with nominal voltage been 12V. We expect a current peak at 10A for about 30 seconds and nominal current would be about 3A.

Thank for the help,

  • Quick update,
    I changed my parameter in webench to 8 <Vin<26 with 5V out @ 10A and now I do get the LM5117 as an option. Based on the design output, it shows Rfb2 5.04K & Rfb1 @1.15kOhms. I am reviewing the simulation and would just like confirmation that this chip could be used as noted above.

    Thank you,
  • I think that this device should be OK.
    Using the EVM is the best way to test your application over all of your variables and environment.
  • Another update.

    I have taken the schematic and BOM found on Document AN-2103 LM5117 and updated webench design to match. I did use the recommend values from Webench for Rfb2 & Rfb1. I got 6.04kOhms for Rfb2 & 1.15kohms for Rfb1. With that, I got a transient 10a spike to sag the voltage as low as 4.87V and reducing the load to 1 amp give us a voltage spike of 5.18V. Well within my intended application. In steady state it looks like my voltage is around  4.985V to 4.996V. So about a 0.011 volt ripple.

    Now I am wondering how webench determines the default circuit values, such as operating frequency. Using the default values I would get 4.996V to 5.004V ripple. I can also see the switching frequency has change and therefore assume that is the reason the system performed better. When I go to design my own layout for my application, I plan to use the webench values. However, the webench doesn't have as many parts as the evaluation board. Although some of these parts seem to be used for test points and optional components, other parts  look to be crucial such as R10, 2.49 ohm resistor between the LM5117 & High Side Mosfet. The Low Side mosfet does not have this resistor on the evaluation board or webench. Should I add this resistor to my design? I would assume it is used to removing some ringing on the gate pin but there is no ground. Current limit the gate pin? Any input is appreciated :)

  • Webench will try to optimize efficiency or size depending on the position of the knob.
    The calculations it makes are somewhat ideal so they may not match exactly with the bench.
    If the EVM is giving good results, then I would stick with that design and layout. Of course there may
    be some extra stuff on the EVM. I think it is OK to keep a place for R10; you can use a zero Ohm as a place
  • Hi Frank,

    I'll keep a note to put some zero ohm resistors on the gate pins. I suppose it wouldn't hurt if later down the line I have to use a different mosfet and just need to update the BOM. Thank you for your feedback.

    I have seen some designs use the resistor on the gate pin and also a resistor (1Mohm or so) from gate pin to ground. I read that the gate pin to ground resistor is used to help shut down the mosfet faster. I believe it does this by give the current built up on the capacitor a path to ground. The resistor between driver and mosfet might be for current limiting.
    I am considering the equivalent model for the mosfet gate to source as a capacitor and would assume the gate resistor from the driver creates a low pass filter, therefore that would remove some ringing oscillations based on the mosfet input capacitance and resistor RC constant? I intend to replace the components on the EVM I purchased and see how closely the simulation matches the performance.

    Thank again,
  • You are correct about the gate R. The series gate R will slow the gate drive due to the RC time constant. Also, this resistor will help to damp
    any high frequency "source follower" type of oscillation. The shunt gate R will help to turn off the FET and keep it off.