• State
  • Locked Locked
  • Replies 5 replies
  • Subscribers 64 subscribers
  • Views 387 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.

LM5118 stability during power up

Walter Kotz
Prodigy 30 points
Other Parts Discussed in Thread: LM5118

LM5118 power up under load not possible - stable operating point is not reached

I've a LM5118 design which works properly if the load is switched on after power up. But if the load is on during power up the output voltage do not reach stable operating Point - the output is ringing.

With lower load during power up it works but not with higher one.

I've checked the design with WEBENCH - the part which I have used match to the suggested part from WEBENCH.

I've tried several changes on the compensation network - unfortunately with no success.
My special question: which parts of the compensation network impact the stability during power up?

Thanks for any hint that can help me to find a way out...

 Walter

  • 0
    TI__Guru 51881 points

    Hi 

    During start-up, output voltage rise can affect stability, but usually it can be resolved by using a bigger Css. 

    Also, during start-up, usually switching mode changes from buck to buck-boost. You have to check both mode to see stable start-up. 

    Regards,

  • 0 in reply to EricLee
    Prodigy 30 points

    Hi Eric, thanks for your reply

    I've tried your recommendations and I saw the influence of Css on the power up behavior – please see the fotos below:

        

                                  Css = 50nF                                                                             Css = 150nF

    Iout = 500mA  

    With 150nF there are three breakdowns. After the third breakdown the output voltage is stable. So I want to ask my next question:

    Why the output voltage breaks down in spite of it is stable in the first 2 ½ seconds after power up?

    I think I’ve still an issue with stability.
    Additionally I want to get more output current – I need 1,3A not only 500mA - (also because of stability problems?). The design should handle it…
    Should I reduce the current sense resistor? - At the Moment I use an 56 mOhm Resistor (SMD)

    Thanks for your reply

    Walter

    p.s.: You can see the WEBENCH design in the appendix.

     

  • 0 in reply to Walter Kotz
    TI__Guru 51881 points

    Hi 

    That's right approach. Also, please check if any big switching noise is there. 

    Regards,

  • 0 in reply to EricLee
    Prodigy 30 points

    Hi Eric!

    I came closer to the answer. See the picture below:

                                 
                         Vout @ 500ms                                       Vout and I_shunt @ 1ms

    The left picture show the power up and 2 ½ seconds later the break-in of the output voltage. The right picture shows the same one with a closer time scale and additionally the voltage drop from the shunt resistor. As you can see, there is a step in the voltage drop about 3 times, because this is the moment, where the load is switched on.

    The design should handle this load current. Can you give me any suggest to solve my problem. Should I try a lower shunt resistor value? Is there any other possibility to get an higher output current.

    Thanks

    Walter

     

  • 0 in reply to Walter Kotz
    TI__Guru 51881 points

    Hi Walter 

    Looks like you are seeing a hiccup mode current limit operation of LM5118. 

    LM5118’s UVLO pin has dual functions. If a cycle-by-cycle current limit fault exists for more than 256 clock cycles, the regulator will enter a “hiccup” mode of current limiting and the UVLO pin will be pulled low by an internal switch.

    If your load turn-on time is less than this 256 cycle or your sense resistor is small enough not to trigger the cycle-by-cycle current limit during load turn-on time, the hiccup operation will not be activated.

    Regards,