Hi,
I have seen multiple threads on this forum regarding MP9486AGN lock-up / start issues, but no real resolution or detailed explanation. We had some similar issues at the beginning, approx. 2-3 years ago while also having issues with badly regulated output voltage when Vin>60V. But later after we removed additional output capacitor and replaced electrolyte one with tantalum type, start-up issues were not detected. Schematic follows evaluation board/reference design.
We made 1000+ devices, with no major issues or at least we did not detect them. But now, in our last production batch, we have multiple devices with this issue.
We have two models of electronics - one without LTE modem. This one works just fine - charts as shown in datasheet. You can see attached ramp up of power supply (Yellow) and output voltage (5V-purple). First at VIN = 10V, second VIN = 32V. No issues here.
Second product - with LTE modem, has compared to above one additional switching buck converter. Typical working board performs like this. I can see a few dips t the beginning (due to inrush current) while caps are charging, but switcher starts and works.
Some boards behave like this. Similar to above, but after two or three dips, switcher locks somewhere in the middle and does not recover. What is even more strange, some boards show this issue at 10V, and work when connected to 32V, while others work at 10V, but not at 32V. Manually you can sometimes recover from that state if you do the same as described here - but this is not a solution for our use case (MP2456GJ-Z Startup problem).
In case of lock-up state, feedback pin is like this - yellow - obviously SCP is triggered.
I will try to measure peak output current but this will not be an easy task, since PCB board is not made in this way due to use of polygons.
Any suggestions, what to do to improve start-up stability?
