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.

LMZ23603: Failure, not regulating and not shuthing down

Part Number: LMZ23603

We are using LMZ23603 to convert from 24V to 5V7 with 1A max output. We have found a number of boards where the regulator starts at the right output voltage set by the resistors (5V7 - 5V8) but then rises quickly (one minute) to 7V and above. The FB voltage goes up accordingly and the regulator doesn't shut off when reaching 0V86 (I have seen 0V95) as per the datasheet. This happens in boards after some time in the field.

They are clearly damaged but still work in an uncontrolled manner. One thing I noticed from the datasheet is the maximum recommended junction temp is 125C and the thermal shut down triggers at 165C. Our heatsinking is not particularly generous and I'm thinking that maybe it has been running for some time at temperatures in between and damaged. But we have other similar regulators in that board showing no problem. I would appreciate any suggestions on what else to check and how to improve the design.

  • Hello Juan,
    Would it be possible to share your schematic and layout here?
    We can do a quick review and see if we see any issues there.
  • Schematic attached, nothing special. The only load to LMZ23603 is a secondary linear regulator that supplies different parts.

    Layout is more difficult as this PCB was done by a contractor who used a different tool... I only have individual gerbers and the attached assembly drawing below. LMZ23603 components in green (some in the bottom side, not highlighted) and second regulator in orange. There isn't much heatsinking, LMZ23603 has 6x10 vias to internal ground planes (2) but they are not connected to copper on top or bottom. Nevertheless, there are other regulators on the same board with similar heatsinking scheme and they don't seem to fail.

  • I think it's fixed by adding a 5K resistor to draw 1 mA from the output. When the LDO afterwards is disabled, there was no load.
  • Hi Juan,

    I see that the datasheet specifies using feedback resistors between 1k and 10k.
    See page 19:

    It looks like the device may need a small load and the 5k resistor you added is probably enough. If you use lower value feedback resistors you may not need the additional 5k you added.