Reported charge level suddenly drops to 0 for a sec, then back to normal

Framework Laptop 13 Linux
1

Tldr: I was sitting in a cafe when suddenly my system reported charge level to be critically low (0%) and initiated emergency hibernate. Whereas in reality the battery was charged to ~85% since half an hour ago I was at home and my notebook was connected to external power.

The thing is, this low charge state was reported for, like, 200 ms, and then back to normal. Apparently OS still registers the event and does its thing.

This happened a couple times, ~15 minutes apart. First time I cancelled the emergency shutdown, second time I did nothing and the system eventually hibernated as it should.

My setup is:

AMD Ryzen AI 300 AI9
Ubuntu 24.04.4 LTS
6.19.3-061903-generic
BIOS 03.05
FRANMGCP 61 Wh battery 2023/11/3, serial 00A7, reported health 90%

[RESPONDED] Battery percentage gets misreported as 0% occasionally
"Battery is empty" message/popup in Gnome on Fedora
2

This happens on my FW16 as well, running Arch in my case.

It’s just a kwirk of the framework firmware I think.

1 Like
3

Yup, my old 11th gen FW13 does it too occasionally. I thought it might be my button battery wearing out and needing replacement so it’s curious (to me) that relatively new model seems to have the same issue.

1 Like
4

IIRC similar thing happened with my gen11 mainboard, and maybe with gen12.

Found similar post with more info and anecdotes.

Also,

cros_ec_lpcs: packet too long: Common on Chromebooks or laptops with embedded controllers (like Framework), leading to 0% reporting.

I recall seeting messages like these and specifically about battery being broken. I discarded them back then because apparently the battery was fine.

Have this in my dmesg:

[   24.649905] cros_ec_lpcs FRMWC004:00: packet too long (4 bytes, expected 0)
[   24.650461] cros_ec_lpcs FRMWC004:00: Chrome EC device registered
5

I guess this is a classic case of bad errror handling. I guess a better fix would be setting this value to 255 instead of 0. Then at least the software could discount it then because 255% charge is unlikely.

6

I just got the laptop so good to know this isn’t just a flaw with mine

7

That’s really interesting. Not the button battery after all, perhaps. The odd thing for me is that it certainly didn’t do it for the first few years.

Thinking back it may have only started once I started using the fan control software; maybe that is what “breaks” the module that also monitors the battery. I’ll have to look into it.

8

I am pretty sure it has nothing to do with the button battery, as the whole reporting logic is tied to the main one. Not to mention that the mainboard can function without CMOS battery just fine.

9

yea I literally just bought the computer and I have the issue, clearly not the CMOS battery

10

Does the system maybe also randomly report very high charge or dischange spikes?

I have this happen for the last couple weeks, and so far chalked up to the battery flipping problem or me updating to the latest bios firmware

Opened a github issue for this behavior