• State Resolved
  • Locked Locked
  • Replies 2 replies
  • Subscribers 62 subscribers
  • Views 261 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.

TPS54325: TPS54325 applied to 12V to 1.0V BODE PLOT phase margin less than 45 degrees

Kao Mick
Intellectual 330 points
Part Number: TPS54325

Hi Sir

TPS54325 is applied to the power supply of MAC 1.0V 1.6A (when the power supply is tested, the MAC chip is not installed), the electrical conditions are as follows~~

Input : 12V

Output : 1.0V  / 1.6A

Cff: 22pF  change to 560pF

R123 : 7.15k change to 30.9k

R137 : 22.1k change to 100k

C1033 : POSCAP 470uF change to MLCC 1206 22uF 6.3V

C131   : POSCAP 470uF change to MLCC 1206 22uF 6.3V

L7 : 1.5uH change to 1.0uH

We tried to increase the cff capacitor and reduce the output capacitor value to open the phase margin to more than 45 degrees.
But the increase of cff is very large, and the bandwidth has exceeded 1/5 of the switching frequency.
Is there any doubt that the control system is unstable?

We have used a lot of TPS54X25 in more than a dozen related circuits, but have never seen Cff increase so much that the phase margin can be greater than 45 degrees.

The load we added is at the near end of the POL, but the output power supply path has been changed three times via the pcb layer to the Net1/Net2/Net3 area ten centimeters away. Could this be the main reason?

Differences in bode plot before and after parameter change of circuit parts

Before

After

  • +1
    TI__Mastermind 23400 points

    Hi Kao Mick,

    The feedforward cap should not be making so much difference because the effective zero and pole for such a small VOUT (1V) should cancel each other out. I suspect the "after" bode plot is showing larger bandwidth due to the smaller COUT compared to before (2x470uF before down to 2x22uF).  The after plot looks good to me. The bandwidth on these parts with the 2x22uF caps should be high as you are observing in the "after" bode plot. You can try removing the 560pF Cff and just keep the 2x22uF Cout caps and measure the bode plot to see if you still observe about the same plot. I am not sure why the prior plot has such a low phase margin and I would expect some ringing or instability issues when there are large transients on the "before" design. 

    Thanks,

    Amod

  • 0 in reply to Amod Vaze
    Intellectual 330 points

    Hi Amod Vaze

    We try to adjust the output capacitor configuration as LCL filter, as follows
    22uF 2pcs and 100uF 1pcs + bead 120ohm + 22uF 1pcs and some 1uF
    PM : 48degree
    GM : about 12dB
    Bandwidth: 122kHz

    We passed the bode plot specification after correcting the Mac factory recommended capacitor values.

    But reducing the capacitance value, we will still pay attention to whether its electrical can meet its transient needs.
    Thank you for your suggestion.