• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 64 subscribers
  • Views 2166 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.

TPS63020 Incorrect Switching Frequency w/ Power Save Mode Disabled

Brian Berner
Mastermind 30280 points
Other Parts Discussed in Thread: TPS63020

Team,

I have been testing a prototype board and the TPS63020 seems to always be operating improperly even though it appears to achieve the correct output voltages it was designed for. The 

amount of output ripple.

The layout guidelines from the datasheet appears to not have been followed and we are worried that the thing traces could be causing some issues with high current, but I am not sure how that would affect the switching frequency.

Any troubleshooting advice would be greatly appreciated. Schematic, layout, and scope shots were taken and attached.

Thanks,

Brian

  • 0
    TI__Guru*** 148920 points

    You are correct that the layout is poorly done.  The D/S has a recommended layout.  The traces for L1/L2 and thin and the grounding of all caps is poorly done.  Poor layout can lead to a device that does not perform as expected.  They should repeat their test on the EVM.

  • 0 in reply to Chris Glaser
    TI__Mastermind 30280 points

    Chris,

    Obviously the poor layout can lead to poor performance (e.g. inefficiency), but is it possible that poor layout would cause the switching frequency to be 15kHz as opposed to the desired switching frequency of 2.4MHz?

    As I understood it, except in Power Save Mode, the device will switch at 2.4MHz. Power Save Mode is forced OFF in this circuit, and I am looking for a logical explanation of the unanticipated switching frequency.

    Is there an App Note that could help in this explanation? Or another e2e post that has seen this phenomenon?

    Thanks,

    Brian

  • 0 in reply to Brian Berner
    TI__Guru*** 148920 points

    I believe those waveforms measured the output ripple.  For regular buck or boost converters, the frequency of the output ripple would typically be at the switching frequency. But this may not be the case for a buck-boost converter if Vin is near Vout.  In this case, it has to buck and then boost to keep the output regulated.  So, then, the output ripple frequency would vary as it switches from buck to boost.  The device is not switching at 15 kHz.  You need to look at the L1 and L2 pins to see what frequency it switches at.

    I would try reproducing the issue on the EVM with their Vin, Vout, and Iout and seeing if the behavior can be reproduced.  

  • 0 in reply to Chris Glaser
    TI__Mastermind 30280 points

    Chris,

    Just to add a little more info on this topic for the sake of closure. After zooming in and looking closely at pin L2, it was very obvious that the TPS63020 was indeed switching at 2.5MHz. After seeing this and observing the large 150uF capacitor at the input of the TPS63020 EVM, we decided to add a 220uF cap at the input of the boost on our prototype board. We thought this may help with the very large input ripple seen at VIN, which was being probed right at the power supply. Multiple power supplies exhibited the same behavior and indicated a large amount of current being consumed by the prototype vs. the EVM. The bulk input capacitor did not help, and it appears the only resolution may be a complete layout revision.

    The layout guidelines in the datasheet are great... if only they were followed! 

    Prototype board showing 400ns switching frequency

    Prototype board showing envelope frequency of ripple

    Prototype board showing large amount of input ripple

    EVM exhibiting pulse skipping with VIN~VOUT (with power save mode turned OFF)

    EVM showing 400ns output switching with VIN <<VOUT (with power save turned OFF)

    Thanks,

    Brian

  • 0 in reply to Brian Berner
    TI__Guru*** 148920 points

    Great.  Thanks for confirming that it switched at 2.5 MHz as it should.

    Yes, sufficient input capacitance to overcome the impedance of the input source is needed for any power supply.  This can lead to what looks like instability but their layout needs to be repaired as well.