• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 62 subscribers
  • Views 435 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

TPS55165-Q1: Power supply working only as buck, in 12V output mode

Nicolas Giussani
Prodigy 10 points
Part Number: TPS55165-Q1


Hi,

I have an issue I'm not understanding. The SMPS should work as buck-boost in the range of ~5.3V/36V (8V/36V is a good range for our application as stabilized constant voltage LED supply)

Over 15.2V I get a stable 12V output. Under that voltage, I wasn't able to make it work. I hope adding some extra info would help:

SCH Design

Design (I tried adding extra input capacitors and both 0.1uF and 100pF output, for high frequencies)

Working_IMG

Captured waveforms (Yellow: OUT - Cyan: IN): Working fine

Fail_IMG

Captured waveforms (Yellow: OUT - Cyan: IN): Not working as expected with lower Vin

Hope I get help. Thanks in advanced.

  • 0
    TI__Mastermind 46905 points

    Hi Nicolas,

    what is output current in your application? please try to increase the output capacitor.

  • 0 in reply to Jasper Li
    Prodigy 10 points

    I already tried that. I also tried adding 0.1uF and 100pF caps, as mentioned above. I'm also trying different values on input, an adding 100pF to VinL.
    It seems this particular IC is extremely sensitive to small layout changes. I don't think it's a good solution to rely on 4 layers, for what should be a fairly simple converter for automotive applications.

    Thanks anyway, Jasper.

  • 0 in reply to Nicolas Giussani
    TI__Mastermind 46905 points

    Hi Nicolas,

    I was thinking about adding another 22uF capacitor, not just 0.1uF. because it seem the device is unstable. could you also try 22uF capacitor?

    The layout is always importance to a switching power supply, especially a fulling integrated device such as TPS55165. 

    but how can you confirm the issue is caused by the two layer PCB?

  • 0 in reply to Jasper Li
    Prodigy 10 points

    That's what I meant when I said "I already tried that" I added another 22uF capacitor. And ALSO 0.1uF and 100pF (incrementally, to test if that help with stabilty).

    I'm aware about routing. I've been working with SMPS for almost 15 years. But never encountered (having designed for many different and complex applications) a chip that is so suceptible to little changes and that has a recommended 4 layer layout (https://www.ti.com/lit/an/slvaej2/slvaej2.pdf and various post answers by Helen Chen, the author of this document). Not saying it's wrong (how could I?), but it's a little harsh for a "simple" converter for automotive market.

    I've been using tons of TI SMPS controllers and regulators, and never faced one this frustrating. Of course, I may have been a little overconfident with my skills and experience, but I never thought It would be in this situation, regardless of the mentioned recommendations. Very often layout guides tends to be a little to conservative. Sorry this doesn't seem to be the case.

    Once again, thanks for your time and advice, Jasper. I appreciate it. I'll keep working on it.

  • 0 in reply to Nicolas Giussani
    TI__Mastermind 46905 points

    Hi Nicolas,

    thanks a lot for sharing detail. I think I understand the situation. the internal control of the device is powered by the input/output voltage. it replied on cleaning voltage in VIN and VOUT pins. otherwise noise coupling into the internal circuit would impact its operation.  so the input loop and output loop is very critical. with four layer PCB, the input loop and output loop are easy to optimize, and the EMI and thermal benefit from the whole ground panel. 

    from pins out of the device, two layer PCB is not impossible if the component can be placed in both top and bottom side.  unfortunately i can't find some sources to verify this in short period.  and in a real system board, not sure if there are other signal traces that will impact the power layout.

    let me know if there is anything I can do to help.