• State Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 1 answer
  • Subscribers 63 subscribers
  • Views 207 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.

TPSM33625: TPSM33625 is damaged when power is applied after connecting the load.

nam byungchang
Prodigy 10 points
Part Number: TPSM33625

Tool/software:

The TPSM33625 part receives a single +30VDC from our product and outputs four output voltages, of which the load currents of +4V and +3.3V are 0.6A and 0.4A, respectively. First, connect each load and apply the input voltage of +30V. After application, the power supply enters the C.C state immediately and the TPSM33625 part is damaged. We confirmed that the resistance of the 4th output pin of the part is 0ohm when measured with a DMM. We are inquiring about the cause of the damage and how to solve it after reviewing our circuit diagram, manufactured PCB photos, etc.

  • 0
    TI__Mastermind 23815 points

    Hi Nam,

    Does this consistently happen? Or did it only happen on one deisgn. schematic looks good. Can you post layout?

    Thanks,

    Andrew

  • 0 in reply to Andrew Kutzler
    Prodigy 10 points

    Hello, this part is continuously broken in the PCB of our product. The PCB LAYOUT of the actual product will be organized and uploaded soon. It works normally on the evaluation board (pictured below, FR-4 2LAYER) that we made separately.

  • 0 in reply to nam byungchang
    TI__Mastermind 23815 points

    Hi Nam,

    I'm noticing that you don't have a low ESR capacitor at input of VIN and a high frequency 0.1uF capacitor. Does adding these parts help?

    This is what we have on the EVM.

    Is it possible to have a startup waveform of when the chip blows up? 

    Thanks,

    Andrew

  • 0 in reply to Andrew Kutzler
    Prodigy 10 points

    Hello
    I modified the circuit as below and the component damage has disappeared, but the vp-p performance is not good. I would like to ask how to improve it.

  • 0 in reply to nam byungchang
    TI__Mastermind 23815 points

    Hi Nam,

    Did you add a 100nF high frequency bypass capacitor and a 4.7uF capacitor to the input?. It may also help to have low ESR ceramic capacitor at the output as well.

    Thanks,
    Andrew

  • 0 in reply to Andrew Kutzler
    Prodigy 10 points

    Hello, I am a company representative who is testing the part to apply it to an actual product.

    The TPSM33625 part was continuously damaged, so we purchased and verified your TPSM33625EVM and confirmed that it was operating normally.

    When we made the peripheral circuit (especially UVLO) the same on the verification board we produced ourselves, the part damage stopped and it operated normally.

    Please refer to the comparison results below.

    Finally, I have two questions.

    1. First, the part damage stopped after the UVLO circuit was processed, so do I have to apply that circuit?
    2. Second, the ripple VP-P characteristic of the output voltage is better on the TPSM33625EVM, so I would like to ask if this performance is due to the input filter effect.

     

    Thank you.

  • 0 in reply to nam byungchang
    TI__Mastermind 23815 points

    Hi Nam,

    1. What UVLO circuit did you implement. I'm not sure what you mean. Can you share the circuit?

    2. You can try shorting out the filter and removing the filter capacitor and VOUT ripple should still be good. EVM has optimized layout and input/output capacitor.

    Thanks,

    Andrew

  • 0 in reply to Andrew Kutzler
    Prodigy 10 points

    The UVLO circuit is the circuit described on pages 16-17 of the TPSM33625 datasheet, and it is implemented identically in the EVM, which means we applied it as well. and Even if the input power filter is removed, there is no significant change in vout ripple.

  • 0 in reply to nam byungchang
    TI__Mastermind 23815 points

    Hi Nam,

    The UVLO circuit delays when the device turns on, so maybe the inrush current in your application with your layout causes a spike in VIN. If you copy the layout of the EVM, I don't think you'll have this issue. Because it fixes your problem, I recommend using the UVLO circuit. Worst case you could always Populate the top resistor as 0Ohm and leave bottom resistor as DNP.

    Thanks,

    Andrew