Does Framework have the resources for it though? The two firmware engineer roles are still waiting to be filled…it’s been vacant for quite some time now.
It’s almost as if the competitions have been keeping the firmware engineers to themselves…and indirectly trying to stop a startup from maturing.
It seems that when Framework hires new people, they announce it on their Linkedin page. But Framework may still be looking for the firmware engineers as I couldn’t find the announcement there.
oh, ok. Well, let’s wait then.
I think the current situation is that the CMOS will not charge from the internal battery only from external power. Thus if you have the battery charged but left it on the shelf for, say, 2 month the CMOS battery will die and you will have a bad time.
Which is very stupid, but, not terrible (as in, you can at least charge it under certain circumstances)
I think somewhere they mentioned a “trickle charge” by plugging a 5V charger (NOT aUSB-PD charger) and have the CMOS charged up.
2.5V under load is definitely bad. 2.77V is still bad.
https://industrial.panasonic.com/ww/products/pt/coin_rechargeable_lithium/models/ML1220
Someone already linked to it. Interesting.
Anyhow the specification said that the standard drain (for the battery) is no more than 30uA. If framework is right on the edge (e.g., 28uA), then a cheap battery will quickly die (or fail) as it’s not up for the job.
I think the 2.8-3.2V mean that if the battery drops below 2.8V (unloaded) you need to get the battery charged.
Can/will Intel patch it? Microcodes?
Perhaps it’s also an answer that requires patience.
This has already been answered. They cannot.
Looks like my RTC battery died after about 18 days (following 24 hours of charge time) with 2 hours of AC power while shutdown at around the 14 day mark. I booted the laptop twice within that 18 day window and had about 15 minutes of use each time.
When framework wouldn’t boot, I connected AC and was able to boot after a long bootup process and remaining main battery was 60%. Windows had the date and time of the last time it was booted, 4 days prior and wouldn’t sync or connect to secure websites.
I’ve installed the NetTime open source SNTP utility from timesynctool.com as a mediation effort for the Windows time issue being such a PITA.
The struggle is real (someone else posted this on Youtube):
Well, the behavior of the laptop in this video doesn’t seem to be what happens when the RTC battery has been drained. If that were the case, it would have either turned on after being plugged in, or not had any charging or debug lights at all.
Actually, what’s not known / documented is how long does the RTC battery needs to maintain / supply voltage for during this startup maneuver (?).
The general instruction from Framework seems to say, leave the 5v / dumb charger plugged in for a few minutes. i.e. It’s not just a matter of the RTC battery needing to supply some minimal voltage while the power button is pressed at 0.000 second. But it seems that the RTC battery actually needs to be able to continue to supply that minimal voltage for some duration during that power up sequence.
i.e. I suspect that the RTC battery has not been charged that few-more-minutes enough to complete the power up sequence such that it can switch RTC power source from coin battery, to continue to be charged by external power.
For more data points - it’s been 23d since I last used my laptop (just to perform the turn-on test I described at the time). It hasn’t been plugged in since then, either.
Laptop turned on fine without plugging it in and currently has 42% main battery level.
Moderators (@2disbetter @Fraoch @Mirage),
I have a suggestion. Shall we change this thread or another thread Laptop won't power on unless i plug in AC power 's first comment to the wiki, changing the thread title a little bit?
While we have nrp’s comment to summarize this issue, Viability of an ML 1220 rechargable battery for RTC | CMOS - #43 by nrp, I think it’s beneficial for users to update the summary day by day. This issue is impactful enough to summarize.
Just chiming back in to say that I have not forgotten this issue. Just today, after a few weeks without using my laptop I tried to turn it on to see if it would still work before I took a trip, and guess what? It didn’t.
Now I have to plug it in until just before I leave for my trip to make sure it will actually work while I am away from home, and won’t necessarily have a place to charge the laptop.
This laptop does so many things so well, yet one of the few things it can’t do well is absolutely nothing, and Framework doesn’t seem to want to do anything to fix it.
Unfortunately, this won’t help in my case, as I will be on my feet all day, not in a car. Had I not checked yesterday, I would have had to take time at the beginning of the event I’m going to, to turn on my laptop at the beginning of the day, and hoped the battery lasted until I was done working with it, which is unlikely.
It’s not that the place I’m going doesn’t have places to charge devices, I’m going to be constantly moving from place to place with no time to plug in my laptop just to turn it on. That would waste enough time to inconvenience the other people I will be with. Now can we stop arguing whether or not I know what I’m talking about?
My argument was never about the main battery, but about the RTC battery having already been dead if you recall. Had I not checked it yesterday, and set it to charge, it still would have been dead, therefore preventing me from turning on my laptop once I was already traveling.
With these quotes, maybe you can see where the confusion came through.
If this makes the whole thing happen less frequently, this sounds like a decent workaround to me. If I leave the thing unused for a month now, it’ll be in the bad state anyway. If the battery is empty, too - what have I lost? I need the laptop on a charger for 24h anyway.
I already manage the charging state of the main battery for the laptop anyway, but I can’t even see the charging state of the RTC battery (as far as I know). If the RTC-battery-life would be somewhat tied to the main battery, I just need to keep managing the main battery and top that of every few weeks - and I can check the state of the main battery, so I have feedback from the system. Currently, I have a laptop that works fine yesterday evening at 90% main battery, and today it’s dead without any warning. That is very annoying.
Alternatively, give us an option for this reset that does not involve fiddling with a fragile battery holder and a fragile main battery connector, that would be fine, too. Press the power button for 45sec with the AC adapter, or something, whatever.
I like this idea…but not sure if it poses any security related risk / issue.
(e.g. In the case of a password protected BIOS…anyone with physical access can reset your board with the power button…and no password needed?) A matter of able to reset with 45seconds physical access, vs 90 seconds (unscrew / pull / put back / screw)…
I’m not exactly clear on what information is stored by the use of RTC battery other than time, vs what’s in the NVRAM.
Didn’t occur to me at all - that might be a problem, I agree.
Just chiming in to say this just happened to me, again. Laptop wouldn’t power on until connected to a charger, then went through a lengthy boot process, and Windows had forgotten the time (it was two weeks in the past). As a result a lot of web services wouldn’t work presumably because of certificate security issues, until I corrected the time manually.
I cannot recommend the Framework laptop to anyone until this issue is resolved in some way. I suppose I should be thankful I haven’t yet had to do the full mainboard reset routine and interfere with the fragile CMOS battery holder.
You’re lucky…for now.
Happened to @Christopher_Yarch today:
Laptop Fails to Turn On - Community Support - Framework Community