• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 64 subscribers
  • Views 661 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.

LM25574 inverting buck converter - output voltage is too high

Dale Tardiff
Intellectual 555 points
Other Parts Discussed in Thread: LM25574

I used webbench to design a converter to provide -5 VDC from +12V source.

Here is the design: 5556.Power Supply -5.pdf

Instead of -5V, the output is -18V. 

Probing the circuit reveals that the converter is running at a too high duty cycle (approx. 60%) even though the voltage at the FB pin is 4V referenced to -ve supply rail. Although it is hard to tell, I think the duty cycle is only being limited by the ramp or current inputs.

Note that for ease of prototyping, my layout is mostly through hole, except of course for the LM25574. 

I used a poured copper plane for the -5V output since everything on the LM25574 is referenced to this level.

Are there any particular things I can look for in the circuit the may indicate the source of the problem?

Also, are there any components or traces that are particularly layout sensitive that I can investigate as causing the problem?

  • 0
    TI__Guru 65905 points

    Although Webench picked the values, the input capacitors seem small to me.  I would try increasing Cin to 10uF or so.

    And paralleing Cio with a 10uF or so.  Be sure that Cio is connected directly between the VIN and AGND of the device.

    The critical layout paths associated with the FB, COMP and SW.  Components Rfbb, Rfbt, and the comp components should be close to the FB and COMP pins.  The inductor and diode should be near the SW pin; keep these paths short.

    The data sheet has more hints about layout and an example.  You could also order an EVM and hook it up for inverting.

    FD

  • 0 in reply to Frank De Stasi
    Intellectual 555 points

    Hi Frank

    I changed the 1 uF capacitor to a 10 uF.

    Still no change on the output.

    The original 1 uF was connected to the AGND of the input source. The 10 uF is also connected to this point. Would you expect a better result if this was connected to the -5V plane.

    I did not have access to a scope last week, but I will scope the waveforms this week and see if there is anything else that might explain this.

    Dale

  • 0 in reply to Dale Tardiff
    TI__Guru 65905 points

    Yes, try a 1uF to 10uF from the Vin of the device to the Agnd on the device.

    FD

     

  • 0 in reply to Frank De Stasi
    Intellectual 555 points

    Hi Frank

    I have been able to scope the power supply.

    Here are some of the waveforms: 

    This is with 10 uF on the Vin input to the ground of the 12V source, and 1 uF from Vin to AGND on the LM25574.

    An obvious problem is the amount of noise on the RAMP signal. 

    I could not see this on the previous scope I was using.

    I suspect that this a result of my layout. I will try some filtering, but I suspect I will need a new layout or evaluation board. 

    One other note, the output of the device to the inductor is at 593 kHz and 70% duty cycle.

    I currently have an 11k resistor from the RT pin of the device to AGND. I had expected < 500 kHz.

    The running frequency was almost 800 kHz with the original 7.5k resistor from the webbench design.

    Dale

  • 0 in reply to Dale Tardiff
    TI__Guru 65905 points

    I can send you a eval board; that might be the easiest way to help troubleshoot.

    Please email your shipping adress and phone to:  simpleswitcherapps@ti.com  and I will send to you.

    FD

     

  • 0 in reply to Frank De Stasi
    Intellectual 555 points

    Hi Frank

    That sounds like a good idea.

    I have sent my shipping info to the email address you provided.

    Thank you

    Dale