• State Resolved
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 64 subscribers
  • Views 72 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

LM5019: Lm5019 Subharmonic oscillation

Ameer Baidusi
Prodigy 10 points
Part Number: LM5019

Hello everyone,

I’m currently testing the LM5019 as a buck converter. The desired output voltage is 15V, and the input voltage range is 18V to 50V. The maximum output current is 60mW.

I used the Excel calculator provided on the component page under Design Tools & Simulation. According to the calculator, it’s recommended to use a Ron resistor of 410kohm, which
yields a nominal switching frequency of about 350KHz. This value is below the maximum allowed frequency at minimum Vin (which is 1MHz).

However, during my testing, I encountered subharmonic oscillation issues at low input voltages (18-20V), the output voltage was 14.46V. There is no mention of subharmonic
oscillation in the datasheet, and I was able to eliminate the issue by raising the switching frequency to 400KHz by reducing the value of Ron.

What could be causing this subharmonic oscillation, and how can I address it?

Thank you for your insights!

  • 0
    TI__Guru 51902 points

    Hi Ameer,

    The issue may be related to insufficient ripple at low Vin (i.e. as Vin approaches Vout) for this COT architecture. If you need additional support, send on the quickstart file for review.

    Regards,

    Tim

  • 0
    Prodigy 10 points

    Hi Tim,

    Thank you for the quick response.

    The output ripple is larger than the minimum limit; I am observing a voltage ripple of approximately 75mV. I conducted additional tests and found that, although it appears to be subharmonic oscillation, it might actually be resonance between parasitic components, leading to spikes in the FB leg and causing the controller to shut down or turn on within each period.

    I concluded this because I observed spikes in the current waveform at every triangular peak, which should ideally be continuous. Additionally, when I reduced the value of the Cac capacitor to ~3nF, the issue was resolved. This suggests that the problem is related to the Cac capacitor and the frequency, which is associated with the Ron value.

    I will attach some oscilloscope screenshots under different test conditions, along with the design scheme for your review.

    Thank you for your assistance.

    Best regards,
    Ameer

  • 0 in reply to Ameer Baidusi
    Prodigy 10 points

    Hi Tim,

    Thank you for the quick response.

    The output ripple is larger than the minimum limit; I am observing a voltage ripple of approximately 75mV. I conducted additional tests and found that, although it appears to be subharmonic oscillation, it might actually be resonance between parasitic components, leading to spikes in the FB leg and causing the controller to shut down or turn on within each period.

    I concluded this because I observed spikes in the current waveform at every triangular peak, which should ideally be continuous. Additionally, when I reduced the value of the Cac capacitor to ~3nF, the issue was resolved. This suggests that the problem is related to the Cac capacitor and the frequency, which is associated with the Ron value.

    I will attach some oscilloscope screenshots under different test conditions, along with the design scheme for your review.

    Thank you for your assistance.

    Schematic:

    parameters summery:

    Test with Vin > 21V, where everything looks good:

    Yellow is the inductor current, and blue is the output voltage ripple.

    same test as above zoomed in, where the spike in the current waveform can be seen at the peak:


    Test with 18V < Vin < 21V:

    It looks like subharmonic oscillation because the frequency (from the low peak to the higher one) is twice the desired frequency.

    15V < Vin < 18V:

    Best regards,
    Ameer

  • 0 in reply to Ameer Baidusi
    TI__Mastermind 28640 points

    Hello Ameer,

    Suggest reducing the ripple as you have done.  By reducing Cac you have done just that.  If it's now working with a lower Cac value, then I would check that the loop is stable under all operating conditions.  

    David.