Various posts in this forum recommend “mem_sleep_default=deep” for increased battery life on the 11th Gen Framework. I also used this setting for years now without problems.
But after updating my BIOS from 3.17 to 3.20, I thought I had bricked my laptop. After putting it to sleep, it would only turn on the fan on high speed, the display would stay black (backlight didn’t turn on), and after 15s or so, would turn off again, only to turn back on after 5s and repeating this forever. Turning it off manually by holding the power button would not help. While it would stay off, the next time I tried to power it on, the same effect happened without me ever being able to boot.
The only way to fix it is resetting the mainboard state by removing the battery and the CMOS battery. The laptop will then show the same effect once (15s high fan speed without backlight), but after turning itself off it will then boot up normally. Phew!
Through trail-and-error I managed to find that this would always happen whenever “mem_sleep_default=deep” is in the kernel command line, and never without. “sleep=deep” alone does not cause the issue.
Thank you so much for pinning down the issue on the sleep mode!
I have exactly the same problem as you described. Now with s2idle sleep it works fine.
Which distro are you using?
Did you already create a bug report for this issue?
I changed the CMOS battery, but had the same issues.
Debian Testing, but I don’t think it matters, the problem seems to lie on the firmware level, not the kernel
Did you already create a bug report for this issue?
No, I’m not actually sure how to do that. I was assuming Framework monitors this thread (they moved my initial report to this one), and are aware of this issue. I might be wrong though, where can I file a bug report? Just sending a message to support?
Regarding “Please Note: After updating to 3.20, you will not be able to downgrade to an earlier version.”
Why this restriction? It seems to go against Framework’s philosophy of giving choice & control to the user, and it takes away a useful troubleshooting/recovery option for those who have issues.
I wouldn’t see it as a restriction rather more a warning.
Frameworks’s philosophy is probably not as simple as you would imagine, they have many things to consider.
Business is a game of ethics so then philosophy can only help guide, not demand or control how a business evolves.
Philosophy is more a person’s way of seeing the world, ethics is more how to act in the best interest of those things a person, or in this case a business, holds important for survival.
Still there must be a reason and yes maybe your topic will provoke an answer. Let’s hope the answer makes sense and therein an acceptable reason for both Framework and the users.
By the way are you concerned about doing an update and something going wrong and not being able to undo - that does make sense?
At first, shows updating correctly - hooray!
I reboot again after the update, and then go in to the BIOS to re-enable Secure Boot
I notice that the BIOS shows version is still/only 3.17 (now due to this situation I’m not sure what version it was previously…)
VERY STRANGE…
So I go back to the webpage, download the file again to be 200% sure I’m getting the right file, repeat the process all over again, except this time it will not update, saying “Cannot update to the same version as already installed”
But after I boot back into Ubuntu and check the BIOS version again, it is still showing 3.17 !!!
Anyone else had this issue?
Is it possible that a mistake has been made with packaging the wrong firmware file into the EFI ZIP package??
Evening again,
Cracked it! Reading the posts you linked @Second_Coming, especially about updating the bare motherboard configs, made me think “I wonder if I need to have the power connected?” Apologies if I’ve just plain missed this in the main instructions, but I don’t remember reading it. Anyhow, yesterday I was trying without AC power connected. Today I tried with AC power connected, and also on the LH side of the laptop (usually I have it on the right) - this was from a comment made on the bare motherboard thread which stated the AC needs to be connected on the LH side…
When I did this, I tried to execute the update from the USB again - it told me once again that it couldn’t update to the same version, HOWEVER, when it re-booted it on the initial splash screen it showed the green progress bar which I didn’t get previously, and took 5-10 mins to complete the update. Then when finally booted up again, now shows BIOS version 3.20.
Apologies for changing 2 variables in one try, but it was either having the AC power connected, or having it connected AND on the LH side of the laptop.
@amoun I’m not using the Windows update so I didn’t read those instructions. I’m using Ubuntu so only read the section on the EFI updater. I re-read those this morning, and there isn’t a mention of any need to have AC power connected - perhaps that should be added @Destroya ?
It starts to run the setup script but very quickly crashes with this message: “Security issue on line 1” …(or 11 - can’t recall).
I type “exit” to get out of the terminal and reboot my machine and go to the internet for help.
Anyone else experiencing this issue?
Another possibly relevant information: I recently updated from Ubuntu 22->24 and found that I was having issues with my WiFi no longer working, which is what prompted me to get around to installing framework updates because I was thinking there may be an issue with my driver. I’m doing this all on ethernet.
It looks like you have Secure Boot enabled; it must be disabled for the EFI updater to work. (This was mentioned in instructions for previous updates, for example 3.17, not sure why it was not mentioned in the latest instructions.)
Had been having many problems since new (11th Gen Intel) with battery not charging, CMOS battery dying, and not powering on unless the main battery is disconnected, even needed to replace the main battery on support’s recommendation (but problems remained).
I updated BIOS from 3.06 to 3.17, then 3.20 today and all seemed normal until I check battery status. Concerned that I may be repeating previous sagas with dead battery, not charging and not powering on with connected battery (making this a desktop).
So you have an ‘older’ 11th Gen but is this a recent purchase as I note you only just updated from 3.06, which I did two years ago.
I see you have a new main battery but what about the RTC/CMOS battery, the ML1220 you don’t mention any change to that.
The ML1220 may well be dying, it requires at least 6hours charging a week or 1 hour a day. If left for a few weeks it will degrade rapidly.
Further you don’t say which OS you are using and there has been a similar on/off charging icon when using Ubuntu 24.04 on my dual boot with Windows but not when using windows.
That Ubuntu issue has been mentioned earlier in this topic, you may like to search and find that.
Yes, an older one and have had it from new, but it became unusable for a long time after power on and battery not charging problems. Running Mint but was on MX initially.
Kept on resetting the board at first, changed the ML1220 at least twice and eventually resetting did not work any more. Main battery was then ordered (support diagnosed that) but as I had relocated, I had to wait for months before I could return to fetch the new battery…
So, only managed to persuade the device to charge and run on the new battery recently by some “black magic” trial and error of procedures. And finally, BIOS got and update.
However, I am getting a few hints/reminders of symptoms that make me worry. For example, when I plug in the power, the bolt deos not appear and it still runs on battery until I reboot. And there is that charge-pending status.
Read that Windows does not have these issues, but I am not eager to run that and far prefer Linux especially vs 11.
