Xiao nRF52840 Running Hot, High Power Consumption and Green LED Always On

Products & Technology XIAO
,
1

Hi everyone,

I have a partially fried Xiao nRF52840 + Wio SX1262 Kit running MeshCore Repeater firmware.

The device was set up as a solar node powered by a 3.7 V Li-Po battery, which was initially (and incorrectly) connected to the GND and 3V3 pins instead of the BAT+ and BAT- pads. The node was left unattended like that for a week.

After that, I noticed that the green LED is always on and the device is running significantly hot, with a current draw of 0.08 mA (measured with a USB tester). This happens when powered either via USB or from the battery connected to the Bat+ and Bat- pads, even with the bottom LoRa board disconnected.

It’s also possible that at some point there was a battery short due to a janky assembly, as the CN3791 MPPT board connected to the Li-Po had some components desoldered and fallen off, most likely due to overheating.

Not sure if the issue was caused by the wrong battery connection or the possible battery short.

The problem seems to be isolated to the charging/power circuit, as there’re no performance issues with MeshCore firmware or LoRa operation. It’s just the high power consumption.

I suspect the BQ25101 chip may be damaged.

Would greatly appreciate any thoughts.

2

Hi there,

YES, you can count on it.. :100:
So TELL me :hear_no_evil_monkey: , The Battery Charger doesn’t work anymore…, Don’t ask how I know.. :grin: :crossed_fingers:

I have some familiarity… What GPIO you got going on ?
You have the MPPT charge controller in Parallel with the XIAO’s BATTEY PADS, with NO DIODE ? :persevering_face: tell me that is NOT so.. :see_no_evil_monkey:
The GREEN LED is the give away. AO. :light_bulb:

GL :slight_smile: PJ :v:

Ah’

Either of those is NOT a good thing. So fix that!

You have a schematic or even a picture of said Apparatus :folded_hands:

LLM–

The Multi-Charger “Fighting” Loop

When two independent battery management ICs (the CN3791 on the solar board and the BQ25101 on the XIAO) are tied to the exact same battery node without blocking diodes, they completely confuse each other’s regulation stages:

  • The CV Stage Conflict: As the battery nears full charge, both chips try to read the battery’s voltage to drop into Constant Voltage (CV) mode. If the MPPT board pushes the rail slightly higher (e.g., $4.21\text{V}$) than what the Xiao expects, the Xiao’s chip perceives this as an over-voltage state or an infinite charging loop.
  • Termination Failure: Neither charger can cleanly read the true internal resistance or termination current ($I_{TERM}$) of the 18650 cell because they are constantly sensing the current being pushed or pulled by the other charger. The chargers stay locked in high-current phases indefinitely, resulting in extreme thermal dissipation. :fire: :firefighter: and a GRN LED… :speak_no_evil_monkey: