TL;DR : Device couldn’t be charged to full. I replaced the voltage divider responsible for battery sensing. However more investigation, and everyday use is needed to fully evaluate this fix.

Symptoms

  1. Near the end of the charging cycle, the power button flashes in red/green erratically. Screen shows 95% and 100% interchangeably.
  2. After disconnecting the wall wart and switching to voltage measurement it shows only ~75%, even after 24h charging.
  3. Single charge gives only about 10h of operation.
  4. After a ~month on the shelf it does not turn back on (it somehow discharges even though turned off).

The lowest voltage my multimeter was able to work on was 7.55V which in my opinion is way to high. Below that it shuts down. I assumed that the critically low value should be around 3.3V * 2 = 6.6V (or even lower): Discharged even though batt is 7.6V

At the other hand, after full charge the battery shows 8V. I’m not sure what to think about it. If I were designing this, I would charge to 8.4V, but I haven’t seen the battery’s specs, so maybe it’s desirable. Then I discharged the battery using an active load, and it seems to be in great condition. If not the charger nor the battery, then the voltage measurement done by the MCU has to be wrong. After poking the PCB I’ve found the divider:

The voltage divider

At 8V the original divider gave (measured) 2.16V (while it should output 2.253). So I replaced the resistors as so:

  R63 R64 out @ 8V in (theoretical) out @ 8V in (measured)
original 392k 1M 2.253 2.16
replacement 35k7 82k5 2.362 ?

Values were chosen mostly because I had those on hand (1%). It fixed the problem, but more investigation, and everyday use is needed to fully evaluate this fix.

Is id fixed?

Thoughts

  • Voltage sensing during charing and during normal operation is realized differently.

Links