I am using the TPS62142 to regulate from 12V input to 3.3V output. The output also feeds a supercap circuit but is protected by a diode so there is no back current. The supercap generate 3.3V for an MPU. We have noted brownout issues with the MPU that we had not been able to correlate with any failure mechanisms but now we have been able to find a repeatable issue. With the TPS62142 at 12V input there are not issues. If we increase the input voltage up to 16V from 12V there are no issues. However, if we power up the system with the input voltage >14.76V (but <16V) the TPS62142 (which is rated 17V input max) starts to lose output regulation, leading to oscillation and reduced current that disables supercap charging. Per the data sheet the only issue I would see at the higher input voltages would be the switching frequency being reduced by approximately 500kHz (but still over 1MHz). It almost seems like the higher input voltage causes the regulator to operate like a lightly loaded regulator with oscillation. The circuit is directly from the datasheet and the layout looks very much like the layout in the datasheet (Rev. E of the datasheet). If we power up at 15V so that we have the issue but reduce the input to 14V then ramp back up there is no issue. It only occurs when the input voltage is >14.76V at power up. We were thinking that the supercap inrush at power up might be an issue but think that this would be more of an issue at 12V than 15V due to the higher headroom. Can you give me some direction on what we should look at next? Although the issue is intermittent is happens often enough that just saying we have bad parts doesn't seem valid. The part does not get very hot so we do not think we have a thermal issue. Thank you.
