LM3477: LM3477 Output Current Limitation and Instability Issue

Part Number: LM3477

Tool/software:

Dear TI Support Team,

We are facing an instability issue with our LM3477 buck converter design and would appreciate your guidance.

Application details:

  • Input voltage range: 23 V – 29 V
  • Output voltage: 14.9 V
  • Output current (target): 3.5 A
  • Switching frequency: ~500 kHz
  • Inductor: 10 µH

Observed behavior:

  • With Rsn = 12 mΩ, the output current is limited to ~300 mA, which is unexpectedly low.
  • With Rsn = 4 mΩ, the output current increases to ~2.5 A, but then periodically drops close to zero and recovers, repeating in cycles.
  • The output current also changes with output voltage, and we are not obtaining stable current regulation.
  • We attached oscilloscope captures of the sense resistor voltage, as well as our schematic and PCB layout for your review.

Our analysis:

  • The expected peak inductor current for 3.5 A load is ~4.2–4.3 A. Based on the current sense threshold (~100 mV typ.), we calculated that an Rsn of ~20–25 mΩ would be more appropriate.
  • We suspect that the instability and oscillation behavior might be related to:
    1. Current sense noise / spikes on the ISEN pin.
    2. Insufficient slope compensation at duty cycles above 50%.
    3. Inductor saturation at high current peaks.
    4. Loop compensation tuning of the COMP pin.
    5. PCB layout coupling between SW node and sense traces.
    6. Input supply foldback or droop under load.

Questions:

  1. What could be the most likely cause of the output instability (periodic drop and recovery of the current) we are observing?
  2. Could you confirm the recommended RC filter range for the ISEN pin to avoid false triggering?
  3. What is the typical current sense threshold we should use when sizing Rsn for ~3.5 A load?
  4. Are there any additional layout or compensation guidelines you would recommend for stable operation with LM3477 at these conditions?

Thank you for your support.WG_MainBoard_V2.1_Battery Charge.pdf