On a more positive note: I’ve been monitoring the EC thermal readouts very closely on the 3.07 beta, and it has so far never frozen (the readings - getting ‘stuck’ until the machine is rebooted) as it did on 3.06. This is why I no longer think there’s a direct link with the tsc clock-source @erudyne, although the sensor drop-out may still be related.
cpu@4c (presumably what @Quin_Chou refers to as “ACPI thermal_zone3”) still sometimes vanishes from the bus, but it does so far less frequently, and, as I said, the readings have - so far - never got stuck. Although more testing is needed, I’m feeling quite optimistic that this critical component is fixed. Never-mind the sensor sometimes dropping out, the important thing is that it comes back, and that the readings continue to be updated.
I’ll report back if I see the problem again but for now it’s looking really good.
Regarding the battery-saver conversation: @Brian_Gregory I read some of your replies and on my end still holds: I had the same problem as you, did what you initially said you did, and it solved it for me. I’ve not seen any odd behaviour since then. I admit though that I’ve not kept close track of when I have/have not connected the charger or the duration at which I’ve done so; only that I’ve not seen my charge stray away from 70% with the charger connected. I’m probably an unreliable source though: the discussion confuses me. As far as I’m concerned right now though, my charging behaviour seems to follow what I expect (be that right or wrong)!
How to trick fw into using chargers it normally would not (also works for old bios versions that don’t work with a lot of pd chargers):
Plug in known working PD charger, plug other charger into other port, unplug pd charger, it should now charge using the other charger. May sometimes need multiple tries.
Have not done it in a while but it worked with everything I tried and I have a lot of chargers and power bank boards and stuff because of reasons XD.
For 5V chargers it also often gets stuck at 1A instead of the allowed 2but still better than nothing.
I do hope they fix that eventually, maybe wile fixing the iphone thing while they are at it.
Peace of mind knowing that no matter where I am or what chargers are available, any of them will work. I lived the opposite life on my XPS 13 9350 which would only charge on 20v which meant many sub-60w chargers did not work. The situations of “uh oh, my battery is low” “I have a charger, will this work” “I literally have no idea” is no fun.
Then framework released the magic 11th gen Intel board that would charge off literally anything.
Then framework released the AMD board which is physically capable of being just as magic (as demonstrated by Adrian’s trickery), but due to software issues it isn’t. Its still a lot better than the old XPS 13 9350 days but I want to be able to say with confidence in any situation “Yes, that charger will work”.
Following up on the charge limit issues: I have not had the laptop go past the charge limit since I disabled the battery extender in the UEFI settings. This has been through a couple of days of unplugging and replugging, different chargers, and reboots, so it seems to be working fine now. Obviously if the issue only occurs under some edge cases (like plugging in a charger with the laptop closed like someone had suggested) then I might just not have run into that by chance so this is by no means definitive, but it’s a good sign.
For everyone still experiencing charge limit issues, I’d suggest going into the UEFI settings, resetting to defaults, and then setting your charge limit and disabling the battery extender at the same time, since that’s what seemingly resolved it for me.
For Framework, I’d ask that you please either clarify whether these features are supposed to be mutually exclusive in the UEFI settings, or resolve whatever interaction between them is causing the charge limit to not be respected.
I have issues with reconnecting an usb c hdmi adapter after restarting from standby on Linux Fedora. The issue started to happen very often after the said Beta Upgrade. Since the Linux 6.12 to 6.13 happened around the same time i am not sure what could be the issue:
The issue is that the usb video adapter part of the multiport adapter doesnt show up at all, no signs of its existence, however the connected networking cable and other usb peripherals work. It could be an Bug in the firmware of the adapter as well.
Has anyone else found better power efficiency with 3.07 beta? I suspect that hibernation modes are better implemented. Or: Manjaro Linux unstable, kernel 6.13.6-1, KDE, X11, FL137040 BIOS 3.07b
Hey all, here to add that I made a mistake here. The firmware i posted here was version 3.06, which honoured the battery extender & charge limit settings. I’ve now updated to 3.07 and I am getting the same issue reported here, which is that the charge limit setting is not being respected.
What I noticed though is that it is respecting the charge limit I set for it from bios 3.05 (98%), so I’m not sure if its just a bug in the EC firmware or this is indeed from my earlier settings. Just wanted to report my observation in case it helps the team.
Similar issues with charging limit not honored. I have tried doing what other people said, but so far nothing has worked long-term. I changed all of the battery settings, one at a time, rebooting after each change. Still broken. I tried resetting back to default settings. Still broken. One person said that disabling battery extender fixed it for them. Others say disabling battery extender definitely will break charge limit. I have tried both ways. Still broken. I tried changing the battery limit to different percentages. Still broken.
Once, and only one time, I saw that when I plugged in the power, it was discharging even though the power was plugged in. I had thought it was fixed, but then it happened again since then.
I updated from 3.06 to 3.07 through Linux fwupdmgr commands. I wonder if doing an inplace flash from 3.07 to 3.07 (same version) using a USB stick might help. I was much happier with the 3.06 BETA. I wish I could downgrade, but that is not possible.
@Quin_Chou@Matt_Hartley@Kieran_Levin To make sure we’re all on the same page – is the broken charge limit in Ryzen 7040 BIOS 3.07 BETA going to be fixed in the final version of 3.07 ?
Or do we need to wait for the fix in BIOS 3.08 ?
Or is 3.07 BETA going to be held, like the previous 3.06 BETA ?
Thanks for the enlightenment!