Hello again,
It seems like I’ve encountered a design issue in MPQ8623, but I hope I am either mistaken or there is a workaround.
MPQ8623 supports starting into pre-biased loads. And it works great - in my project it correctly starts into 5V.
It is the shutdown that is problematic. Namely, when EN goes below both input and LDO falling thresholds, the buck activates output discharge mode, as described in the datasheet.
Unfortunately, the second voltage source on Vout is still present, and is still supplying the 5 V (at around 500mA). The buck is sinking around 50mA of that continuously, but naturally cannot make any progress reducing the readout on FB. And since the voltage does not change, it just hangs in an eternal discharge.
Any attempts to restart it also fail, because in this state it does not react to EN going up. I have to turn off the second voltage source momentarily or short FB to GND, and only then the buck shuts down entirely and is ready for another startup.
Is this a known problem? Is there a mitigation? I think can work around it by adding an EN-controlled sinking NPN at FB, but I am worried it could affect loop stability.
Thanks,
Milek