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.

TPS63710: Clarification with Recommended Inductor value

Part Number: TPS63710

Hi Team,

Our customer would like to know any reason not to use a larger inductance to reduce ripple and to avoid discontinuous conduction mode over a wider load range for TPS63710.

As datasheet, section 6.3 “Recommended Operating Conditions” lists the maximum effective inductance at 6.2uH. 

Their application will use TPS63710 to convert 12V to between -3.5V to -5V, target voltage to be determined.
Output current will range between -80mA to -400mA.

Minimizing ripple and EMI are design priorities.

Please let me know if you have any questions.

Thanks,

Jonathan

  • Jonathan,

    Under these conditions, tt is possible to use a slightly larger inductor value, but just as with a standard Buck regulator, this will slow down the overall loop response of the regulator.   Likewise, the output capacitance may also be increased with the same effect.

    Best practice is to test the load transient response to make sure there aren't any obvious signs of loop instability with the larger values.  You may also want to use the Spice for TI tool and model available on Ti.com to test out different component values under different load transient and start up conditions.

    Sincerely,

    -Matt

  • Hi Matt, 

    Thank you for supporting this thread.

    Upon sharing your answer to our customer, here's the additional questions and comments:

    The TPS63710DRRR converters 12V to between -3.5V and -5V, exact level to be determined.  The load will draw between about 0.12amps and 0.4 amps. 

    Webench simulation with a 5.6mH inductor and 10uF effective output capacitance showed no instability, and recovery from load transients between ~9% full load and ~99% full load in less than 0.08mS.

    Nevertheless, Webench recommends a 4.7uH inductor for all sets of requirements that they tried.  Since the data sheet does not provide dynamic specifications beyond a few transient response observations, dynamic design must rely on the model incorporated into Webench.  Do the results described above reliably demonstrate load step response and stability?

    Also, the customer has some question with Webench design below (I don't know if I need to create separate thread on this):

    First. The problem:  Downloaded simulation CSV files do not include sample time information.  Please see the attached example.
    WBDesignInputTransient5p0V.csv

    Questions:

    Second. When changing the output voltage on Dc to Dc converter design, webench resets all components to their default selections instead of leaving the individually selected components that did not have to change.  They understand that the voltage reference divider resistors need to change.  But, it would have been useful to be able to leave the other components that I had selected.

    Third.  The capacitor selection tables do not show the capacitor size e.g .0805, 0603.

    Please let me know your comments.

    Regards,

    Jonathan

     

  • Jonathan, I would recommend submitting a separate thread on the WebBench specific questions/recommednations as these can then be directly assigned to someone on that support team. I've not spent much time using the WebBench simulator, but the load step behavior the simulator is showing looks reasonable to me. Alternatively, if the customer wants to evaluate different component values within a given design along with various startup and load step responses in greater detail, i would recommend using our Spice for TI tool. Sincerely, -Matt
  • Hi Matt,

    I will create a new thread for Webench and mark this as Resolved.

    Thank you for your help.

    Regards,

    Jonathan