I am using two MP3432 boost converters on a board design. One is set to regulate a 3.7V Lipo input to 5V, one to 12V. The 12V stage behaves as expected, but the 5V for some reason is set at 10.7V.
I did quite some testing but can’t come to a sensible conclusion:
- Vout is 10.7V instead of 5V, and the voltage seems quite stable
- the FB-pin is at 2.11V, which is of course wrong
- the resistors R1/R2 show a voltage drop corresponding to the intended resistance ratio 4:1 for setting 5V output voltage
- AGND and PGND are connected, no resistance measured that would explain a GND potential difference
- it doesn’t seem like FB is shorted to any adjacent pin like COMP
The behavior was observed identically on two different boards, on one of which I completely isolated the boost stage (on the other the 10.7V broke other components and only then I realized that it was not working properly). The board showed the described behavior right after powering it on the first time.
As you can see in the attached screenshot, the layout is quite dense due to space constraints. But as I see it, that should not cause such a behavior. Thermal issues maybe… also, the 12V stage is almost identical and works. So it can’t be a bad batch or wrong handling by the PCB fab as well. What can cause something like this? Any hints would be greatly appreciated, then I’ll try to debug further
