After leaving my Pocket connected to USB power and not using it much for a few weeks, now the battery meter only shows 40-42% on a full charge.
When I disconnect power and leave the device running I get the battery life I would expect from a full charge but the meter reaches 0% long before the device shuts down.
After such a discharge cycle, when I charge again, it gets to 100% but the problem returns after a while of regular usage with shorter charge/discharge cycles.
Same issue as yourself, if you let it run down to 0 it recharges fully, however once it starts discharging again, I can’t recharge beyond it’s current point, until it gets to 0 again.
We (me and zeha) started looking into the details of the fuel gauge/protector chip and we’ll soon start testing some code to give it a detailled configuration. Currently it’s running mostly on defaults and doesn’t know the “design capacity” of the cells, for example. There’s also a command sequence to reset the gauge which should be useful to implement as a user accessible function. But yeah, at the moment the only way to calibrate it is running to zero and charging up again.
BTW charger board v2 is almost ready and I’m also looking into bumping the battery size a bit.
I’ve been observing this a bit longer and it’s really inconsistent. It seems to matter when a charger is connected, what charger I use, what the current battery level is, but I have not been able to see any pattern to it.
Also it’s a fairly recent issue, only appearing in recent months, but I’m not sure since when exactly.
Today I ran into the “refusing to charge” issue again, I let it discharge all the way until automatic shutdown and then plugged it in but it stopped charging at 11% at first and after unplugging and plugging in a gain it went to 13% and then stopped. I disconnected the batteries from the board for a moment and then tried again. The battery indicator read out 3% and started charging again.
In general it’s happening a lot and I really think this issue was introduced with a firmware update because until recently I had used my Pocket for over a year without ever encountering this problem.
Right now I pretty much can’t use my Pocket unless I bring a power bank even for just short periods in case it didn’t charge over night again.
OK, that’s weird, thanks for the detailled write-up. There’s a few things we can do:
We can send you a charger board v2, perhaps your charger board v1 is developing issues.
If it’s just a firmware issue, we should be able to see more if you could monitor the output of tio /dev/ttyACM0 for a while, especially when you plug in the charger and it doesn’t want to charge. You can just paste/attach the logs here. This is the virtual serial port of the RP2040 System Controller, and it’s quite chatty about charging/discharging/charger PD communication. CCing @zeha here who should be able to interpret the logs best.
@selfawaresoup I think I have the same issue and I’ve got new batteries and the v2 charger board. I’d be interested to see if you have the same problem that I have - when it’s unplugged, the battery voltages sit somewhere around where you’d expect them to be (e.g. both at 3.4V) but then when plugged in one of the cells will go straight to e.g. 4.1V, eventually ticking up to 4.2 which causes the PD negotiation to drop down to sustenance current whilst the other one sits around e.g. 3.7/3.8.
I did some firmware updates over the weekend and left my pocket charging today (it could only get to 7%) and when I checked later it was charging and got all the way up to full, so hopefully somewhere in the combo of things I did was a proper solution (I think it’s probably the better cell balancing on the v2 board that’s resolved the imbalance over time).
