• State TI Thinks Resolved
  • Locked Locked
  • Replies 1 reply
  • Answers 1 answer
  • Subscribers 63 subscribers
  • Views 276 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.

TPS54362-HT: I DON'T BELIEVE SPECIFICATION, OR IS SOMETHING WRONG.

VICTOR BEST
Intellectual 350 points
Part Number: TPS54362-HT

This we Buck is really giving me a hard time.  It seems to be oscillating much slower that specified.  Changing the pull down resistor does make device run faster(RT) pin 6.

Let us start with basics, this device is configured as per EVAL board.  I start my testing with the supply set to 15V DC, and actually with 10R load we ge very little ripple voltage and a lovely 3.3V DC output, one might think that is all dandy and good.  However as load changes even to 20R we get lots of LF ripple(2khz) voltage and device is unstable, the output voltage is 3V with over 500mV ripple voltage on top.

Cin =22uF 100V MLCC 

Cout=100uF, but it photo changed to 47uF

RT=220k. To me Fb is totally screwed, but would this explain the oscilator operating at 2kHz, i think perhaps we have a faulty part, but given at high loads it works makes me feel that the device is conditionally stable? What are your thoughts?

Yellow=Vin / Blue= Vout.

  • 0
    TI__Expert 3835 points

    Hi Victor,

    If you believe it is a fault part, try replacing the part and see if the issue is repeatable. Using a different board as well.

    Looks like this is low load and may be in the boundary of DCM in low-power mode. Check the sw node frequency since it may be in PFM. Try removing the snubber RC as well. Is the output caps ceramic as well? Can you show me a scope shot at 10 ohm load (with sw node)? 

    If you think  the design is unstable, then a quick load transient test will tell. You can also do a bode plot with a frequency response analyzer.

    Regards,

    Ethan