This review was linked in another thread on the forums, and summarizes the issue with the RTC battery well, and it gave me an idea.
Other than @anon81945988’s letter to Framework, to get them to actually give this issue the attention, and response it deserves, anyone who is having issues like this should vocalize on other platforms as well.
If they don’t want to give this the transparency we want (as they are still not answering many questions on this thread), then we need to make this a bigger issue for them to consider.
I agree with @Richard_Lees and @Senhara here. Having read through this entire thread, it seems like a largely isolated problem that not many people are facing.
This thread seems to largely be filled with overreactions, and input from people who aren’t facing the problem.
On the other hand, second opinions by media or individuals are also important. It’s an important role of journalism. Hope that tech media write not only a typical product review, but also write deeper insight about these kinds of Framework’s major challenges.
When you left your laptop unused, was it at full battery before? Did you have to connect AC power to get it to boot? Did you find that even though you had to connect AC power, the main internal battery still had plenty of charge? Are you using windows? Did you have to manually set the system clock?
There are two separate issues:
RTC high drain and low capacity design that causes the RTC battery to drain much more quickly than other laptops. Design requires AC to be connected or RTC battery to be charged for the laptop to boot even if there is enough juice in the main battery. Also, system clock will not keep requiring it to be manually set if you use windows. This presumably affects everyone when leaving laptop unplugged for 2-3 weeks. Some people are having this happen even more often (5 days).
CPU state / dead RTC - after #1 occurs, laptop will not boot at all even if AC is connected requiring the full reset procedure (remove RTC battery, plug in 2nd charger, etc). This is more rare.
Framework has said #1 is by design and they have used same design in 12 Gen boards but have lowered idle drain. No fix for 11th gen users.
Framework has said #2 is caused by Intel. Obviously #2 is worse because of the design decision made by framework in #1.
For those saying to contact support, they have stopped replying to several tickets and nrp has not answered questions to his post regarding design decisions.
This is outright false. Like a sane OS, Windows uses NTP. This isn’t 1998, this is 2022. Everyone uses NTP. It only doesn’t use it if you manually disable automatic timekeeping.
I can echo what @Richard_Lees is saying. I have 2 Framework laptops that I have never had any problems with. I have never had to reset the mainboard or remove the RTC battery.
It is unfortunate that your situation is not the same, but please refrain from calling other people here liars.
I appreciate you commenting in this thread even though you use your framework often and don’t have the issues others are experiencing. I am happy the laptop matches your use case. Unfortunately you likely have less insight on the frustration because of this. I’d invite you to install fresh Windows 11, charge up your framework to 100%, then shut it down and set it away for a few weeks like many of us do who don’t use framework as our primary daily machine.
You will discover that the framework will not boot without connecting AC power even though the main battery still has plenty of juice. When you do connect AC power, it may boot after a few minutes, or you may have to perform the reset procedure on the RTC battery. If it does boot, you’ll not the internal main battery will still have plenty of life. You’ll note that the system time is off and isn’t automatically syncing. You’ll note that even if you go to Date & Time that in Windows 11 “Set time automatically” is enabled by default. If you click Sync now, you’ll see it still does not sync.
You’ll notice errors in the event viewer that explains the issue.
Yes Windows uses NTP but Sync automatically only works if the clock isn’t too far off. You can try to launch Edge but you won’t be able to load any websites because of certificate validation errors (due to system time being off). You have to manually set the time to be closer to the correct time, then Sync time works. Try it yourself.
This is a very wise statement! I agree, others may have different experiences. However, with regards to Windows NTP, there are actually several articles which state the tolerance of windows “Set time automatically” and “Sync now”. Many threads out there do conclude the issue boils down to BIOS, CMOS, or RTC battery failure causing the time to drift outside of the Windows tolerance.
I’ve talked with people outside the forum, turns out yes this actually correct. I’m used to Chrony’s behavior when dealing with NTP, so I apologize for my inaccuracy on knowing how NTP services work.
However, I have set my Framework into similar conditions described above that should trigger the issue, and I haven’t experienced them, additionally there’s so much sheer misinformation, overreaction, and whining I can’t help but respond.
No problem at all, super cool for you to reply with this! I really appreciate it. I was just as shocked as you when I experienced it because I have never seen this before. I too thought NTP would save me as it has on every other machine I’ve ever used. Framework was the first time I even learned there was such a NTP tolerance in Windows because the sync time not working was driving me crazy! Turns out, those CMOS batteries lasting for years and years were keeping system time all along. When they fail in other platforms, swapping the battery is truly no big deal because it buys so many more years of usage. Reminds me of the first time I had to replace battery in a Casio watch (after 12 years of service)!
You weren’t able to trigger the issue where it won’t boot without connecting to AC? Or you weren’t able to get it to go into the dead-CPU state where it requires the reboot? My understanding is as my above post - it should be easy for anyone to get into the #1 state where the main battery is juiced but the framework won’t boot because RTC is dead. This state might be fixed quickly (plug in AC and wait a few minutes for RTC to charge) or it might trigger #2 (harder to recover).
I very much understand this. I think this is likely a passionate community. I think that’s what got framework so much attention out the gate. For me personally, I am loud and obnoxious but I’m also fair about it. I cheer just as loudly when I think Framework does something great - I tell different groups of friends and coworkers about the good just as well as the bad. I expect several visiting this thread or forum may be similar ¯_(ツ)_/¯
I wasn’t able to trigger the issue at all, perhaps I got lucky and my RTC battery was completely topped off before the system was shut down, or the CPU never got into a bad state and Chrony recovered the time just fine before I noticed, not sure. After all, this problem is caused by a mistake on Intel’s side and an oversight Framework made due to not knowing about the Intel issue. 12th gen boards allow the battery to actively supply the RTC chip with power and has an automatic reset function. Combined with other changes, it should outright circumvent any issue experienced from RTC dying other than time resetting.
I hope that’s true. It’s not completely clear from @nrp post:
It does confirm the RTC power consumption is lowered that should extend the idle time - which is great. I’m not sure what “designed in a path” means with regards to the charging. I’d love to know more about this. It seems they have maybe planned for it but haven’t figured out how to implement it? Definitely seems like an improvement for sure.
I’m optimistic about the 12th Gen revisions. My big concern as someone who ordered an 11th Gen in July and received it in October and only has a half dozen hours of total use is - where does that leave me? There is no fix for 11th Gen users or any trade in process. I need the idle time to be as long as possible for my use case as I am using this computer specifically for field. I have tried to engage support on this and offered to pay for a trade up also but they just stopped replying to my ticket mid last week.
This is the most concerning piece of it all that can really infuriate us super loyal and passionate framework fans - nrp points to support, support just ignores
This most likely means “We’ve designed an electrical path so the main battery can directly charge the RTC battery”, a physical electrical fix.
Admittedly I’m not sure if it’ll be charged during poweroff or just when the system is alive. Considering the hardware and firmware poweroff drain fixes coming in the future, I hope it’s the former.
Yep maybe you needed longer idle but hey, I’m super happy for you honestly. Possibly this issue just won’t effect you at all and that’s great. Be happy and don’t go looking for things to be upset about.
For me, I have a work issued MacBook and a gaming desktop so Framework is a 3rd machine that I only get to use for certain hobbies and vacations and so my time between uses is spread.
That makes sense. Maybe an electrical path that needs some logic / firmware on when to “wake up” the main battery to charge the RTC battery (when it gets too low). I was thinking of path more like “a series of steps or directions we need to take but we haven’t traveled down it yet”
I’m not lol, it’s mainly other people trying to find things to be upset about that upsets me.
Sometimes electrical circuits are described as “paths”, like how you would walk a path to say, a convenience store, electricity would go through a path to where it needs to go.
To clarify my post to @Tai_Kahn and to reply to @Shiroudan’s comment about it, I am not looking to take advantage of anyone’s situation. In the case of @Tai_Kahn, if the return period still applies, it would appear to be the path of choice. If not, and the poster is frustrated with trying to achieve a resolution, I am curious as to the price for which the machine would be sold.
There are several reasons behind this:
I have had an excellent experience with Framework’s Support team and the company in general when I have needed to engage them. In fact, they had to prod me politely to return a part for which they had shipped a replacement. This was because I had been busy and not had a chance to swap the parts. But that was my fault, not theirs, and they were extremely polite about it. They wanted the old part back so they could analyze what had happened.
This would not be my primary machine, which would make any failure to boot, etc, significantly less impactful.
Again since it would not be my primary machine, I would have the luxury of time to work through things with Framework’s Support team.
I am genuinely curious to try to see what is going on. I am under no illusion that I could resolve anything where Framework themselves have not, but I like to tinker and experiment, and this would afford me that opportunity. The ability more easily to take things apart and rebuild them is what intrigued me about Framework in the first place. I was a bit disappointed that I didn’t get to build the whole computer piece by piece. Maybe with this one I would strip it all the way down and put it back together. I will admit that I was a bit chicken to do so with my primary machine, and didn’t go there with the second family machine, either.
So, @Tai_Kahn, my questions stand, they were not intended to be snarky, I have no hidden agenda, and I am sorry that you are so frustrated, I wouldn’t wish that on anyone. If you are interested in selling, please reach out either here or via a direct message.
