• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 62 subscribers
  • Views 571 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.

TPS40170: TPS40170 - Audbile noise at middle range load.

tuv0k
Prodigy 90 points
Part Number: TPS40170

Hi,

I have a potential problem with my TPS40170 design. Everything works ok at light and high purely resistive load, however when I connect CPU on my PCB Buck starts producing barely audible high-frequency noise. I try to remove ceramics completely(left only Aluminium Poly) and it is still present. My second clue is the shielded inductor - I could replace it in the next design with unshielded one. I don't think this is stability issue, however I am a bit worried since oscillating frequency is much above audible(300kHz) range.

Do you guys have any other idea? My schematic is given below:

  • 0
    TI__Mastermind 25590 points

    Tuv0k,

    Thanks for sharing the schematic and the info.

    A) Which component is making the noise? You can probably use the mic on your cell phone and a spectrum analyzer app.

    B) Is the audible noise:

    • A solid single tone?
    • A wobbly single tone?
    • A couple tones?
    • Noise (like white noise)?

    C) Check the current going into the CPU. Is it pulling current periodically at the audible frequency? Or at a different frequency?

    D) Do you see any subharmonic oscillation? Check the SW node and check the min/max/avg duty cycle.

    E) Check the output voltage. AC couple and zoom way in. Is it flat (as compared to the SW ripple)?

    -Sam

  • 0 in reply to Samuel Jaffe
    Prodigy 90 points

    HI Samuel

    Thank you for your questions. 

    A) The intensity of audible noise is so low that both S7 and iPhone7 were not able to catch it on spectrum analyzer app

    B) It is definitely not white noise(I had that problem on some old switcher - output ceramics were "singing")! I would say it is something between wobbly single tone and couple of tones.

    D) and E)

    SW node

    VOUT

    All measurements were taken using "short" mass probe.

    Removing all output ceramics and leaving only 150u AL-POLY - noise was still here. My second thought was inductor - adding glue around it didn't help. However, when I pushed C127(ceramics in the feedback loop) noise diminishes - everything is quiet. I don't think the reason behind that was capacity change.

    I think it is definitely my CPU board(ARM-A9) which draws current in audible noise frequency band - however I don't have equipment to verify that.

    How to solve feedback ceramics whining without glue/epoxy etc.?

  • 0 in reply to tuv0k
    TI__Mastermind 25590 points

    Tuv0k,

    Here's a great blog about reducing audible noise in ceramic capacitors. The highlights:

    • Use acoustically quieter capacitors
    • Optimize layout
    • Reduce voltage variation

    You could also try using higher voltage-rated capacitors.

    That said, there may still be circuit optimization that we can do to help this.

    1) Check the SW node with the CPU load. Measure the min and max duty cycles as it runs. Is it wobbly?

    2) Check VOUT with the CPU load. Zoom out a bit to a timescale where you can see the audible noise frequency (e.g. if it sounds like a 1kHz tone, change timescale to ~500us/div).

    -Sam