nRF52840 ProMicro

The nRF52840 ProMicro board is also known as the SuperMini board. It's a clone of the nice!nano board with a mix of v1 and v2 design features. The schematic is available online, albeit with a small mistake, but I was unable to find board design files. The aim of this repository is to reverse engineer the board design and determine which component is which so that I can desolder them from the board and reuse them on my custom PCB.

Schematic

This is the original schematic as posted on https://chat.nologo.tech/d/80. The voltage divider for the battery is marked as being connected to P0.04, which is an analog input pin. On the actual board it's connected to P0.24 instead, which makes it unsuitable for battery voltage measurements. Instead you can measure the battery directly on VDDH pin, but the voltage will rise when USB is plugged in. To prevent this incorrect reading, you can remove the NPQ2, NBD1 and NPR7 components, and instead bridge pins 3 and 2 on the NPQ2 MOSFET. This way the nRF chip will always draw power from the battery and USB will only charge the battery and thus power the board indirectly. The voltage on VDDH will then be the actual voltage of the battery (slightly higher while charging but it won't jump to 100%). This could be problematic if your board consumed more current than the charger provided, and the battery would discharge while plugged in, so keep that in mind.

Sleep current problems

I wanted to use these boards to create some low power BTHome sensors. I have boards from many different sources. Some exhibit great power consumption, only 4 uA in sleep, with EXTVCC disabled. Others had a consumption of 60 uA, which is still pretty good but it's 15x more. After some component swapping, I found the problem to be the W5 marked diode. This is supposed to be a BAT60B Schottky diode, with a voltage drop of 0.24 V. On the high power consumption boards I instead measured a drop of 0.7 V, indicating a regular silicon diode. What's more concerning is that apparently the reverse leakage current of this diode was also much worse and this is what was causing the power consumption.

KiCad project

The project in the KiCad folder is my attempt at replicating the board to identify every component's location. The USB connector is not the correct footprint, but the name in the schematic is correct. KiCad doesn't seem to have a footprint for the antenna, but the model is CA-C03.