Wow you were right, my chromebook sits right next to my laptop and it has a really strong magnet. If I move my framework close to the laptop the framework will shut off. Thank god its not an actual issue then! Thanks
Another one to add to the list. In my case, it was just sitting there idling with the lid closed for nearly a day.
AC plugged in. 80% battery. No event log messages aside for “The previous system shutdown was unexpected.” No BSOD. No dumps. Nothing. System died as if someone powered it off.
12th Gen DYI - 3rd Batch, but I purchased their memory.
Whatever this is hasn’t been fixed. Using a 12th gen i7-1260P. I was on battery around 85% with a little bit of WebGL going and it instantly powered off mid-keystroke with no logs. I was nowhere near max TDP and had minimal disk usage but notably using the GPU.
I’m running Arch with the beta 3.06 BIOS, so this isn’t a windows-only thing. My power settings are tuned for savings, with some cores turned off when not in use, turbo disabled on battery, etc.
Putting my engineer hat on for a second, I’d guess it is a power management problem for the CPU/GPU package in response to big spiky demand changes. I’ve seen other machines, notably desktop GPUs have brownouts and their VRM regulation algorithms needed tweaks in firmware.
Which kernel version are you running and which kernel parameters did you set?
uname -r && cat /proc/cmdline
Nothing exotic:
6.1.7-arch1-1
initrd=\intel-ucode.img initrd=\initramfs-linux.img cryptdevice=<BLAB> root=<BLAB> rw rootfstype=xfs mem_sleep_default=deep resume=<BLAB> resume_offset=<BLAB>
I haven’t been able to replicate the power failure despite trying a bunch of stuff. It did happen from a clean boot, so I’m sure there are no S3 sleep or hibernation effects. I also have hid_sensor_hub blacklisted, and here’s the powertop report on battery showing which power management options are available/enabled/disabled
Hi guys,
I´m using a 13 inch 12th generation laptop with an i5 and RAM and SSD from Framework. In the last few weeks I also have the described problem of crashes with the error message kernel power in the event viewer.
I have already tried a few things to fix the problem (checked BIOS, drivers, changed the power options, removed the expansion cards (only used USB C and USB A)).
Now I’m not an electrical engineer and bought the laptop more as a repair fan and am slowly reaching my limits.
Framework support had also not yet given me a solution. Since the last post here some time ago, is there anything new here to solve the error?
1 Like
Hi,
i have the same Problems with my 12th gen mainboard and Windows 11. I also tried to fix it with changes in the power settings. But nothing helped so far. please help.
I wonder why this thread doesn’t have a [TRACKING] tag.
That tag is used by Matt and the Linux support team to show that they are aware of the issue. The regular support team don’t use tags like that, and this thread was started before anyone started using tags like those anyway.
Yes, I’m aware of those points. Matt and the Linux support team have the ability to update / modify thread titles, do they not? Regardless of when the thread was created. There are Linux users in this thread reporting “instant power loss” behaviour. So why don’t we see this thread updated with [TRACKING]?
Matt joined in Nov last year, and this has been an [quietly] active thread since then.
I guess I’m asking: Are these isolated incidents?
@Second_Coming So just tag @Matt_Hartley and @Loell_Framework and be done with it. You are making it look like some big conspiracy that they haven’t responded. Geez.
I don’t believe it’s an OS problem, but a hardware issue, considering that a RAM shuffle fixed it for me at least.
2 Likes
I’d say that this is the exact reason, and why I mentioned that only the linux support uses those tags. If this isn’t a linux-specific issue then I would assume someone who doesn’t tag posts is working on it. Also remember that while Matt and Loell come here to search for issues, it’s still a user forum, so just because Framework hasn’t responded here does not mean they are ignoring it.
Sorry about that, the miscommunication came from the fact that my original statement is that I meant Matt and Loell use the tag to show they are aware and tracking a relevant linux issue, and they don’t really do any work with hardware issues on the forum, my bad!
1 Like
Hi guys will label this as tracking.
4 Likes
Going forward, contact support if you’re experiencing the issue in the title of this thread.
6 Likes