I think several of us have noticed it happens when the charging rate drops off. As others have mentioned some Microsoft Surface devices had this issue, a couple examples. Likely will be a firmware issue / fix.
3 Likes
Since this seems to be the thread for CPU throttling, here is my report:
After putting the laptop to sleep (unplugged) for a few hours and picking it back up, I immediately noticed that it didn’t feel as “snappy” as it did before. All 8 cores appeared to be switching back and forth between 0.2 and 2.8 GHz (but nothing in between).
FWIW, the battery was at about 72% before sleeping, and 65% when I woke it up about 3-4 hours later.
I tried rebooting and cold-booting, but neither helped. I also noticed that the time between pressing the power button and actually seeing Linux boot was unusually long, which suggests an issue with the firmware rather than the OS power management stack.
In the end, I turned the laptop off and let it cool down for a half hour. Now it is back to its snappy self 
All this to say, my experience has been consistent with the hypothesis that either (a) BD PROCHOT is just a bit too sensitive, or (b) there is some physical defect within the device that is causing targeted overheating (even though the overall device is only mildly warm to the touch).
1 Like
If it is a firmware issue, wouldn’t everyone be affected?
Are most of the issues with users coming out of sleep mode while charging or having had it in sleep mode?
I think this thread has coalesced around a specific issue when charging rather than sleep, and only in certain charge ranges (starting somewhere in the 70-80% range through 80-90% or so, whenever the charging rate lowers towards the end of charging, clocks get stuck at 400MHz, though clockspeeds never fully recover while charging through the end of the charge cycle). Several threads were merged into one so it’s a bit of a mess to read through, honestly if someone is experiencing a different issue probably should start a new thread for visibility.
@Bobby_Reynolds IIRC clocks bumping down to 200MHz might be related more EMI stickers on the USB ports or another issue entirely, have you read up on that?
RE: not everyone being affected, it’s possible sensor tolerance is in the expected range for the vast majority of users, but the range is too narrow so some users are getting false readings.
1 Like
Update from my end… Once again my laptop had been sitting off, cold for close to a week. It didn’t respond to power button presses until I plugged in the AC adapter. Battery read at 76%. The RTC had been frozen across the time it was powered off (Windows thinks it is the afternoon of 10/23). CPU was stuck at 0.39GHz again.
Installed the 3.06 BIOS update. By the time it completed, battery was around 85%, and the CPU is no longer pinned at 0.39GHz. So I guess that’s another data point for the ~80% threshold being a factor.
Took a video of the side LED while waiting for it to boot after the BIOS update. White, 12 green blinks, then orange, blue, green, green, and back to orange (steady, charging).
Tomorrow I’ll check/reseat the CMOS battery and try adjusting the EMI stickers (hopefully without ripping any of them).
@nrp , do you need anything else from us? not asking for daily updates, but a weekly one would be appreciated.
2 Likes
I just got my Batch 4 framework laptop and I am having this same issue. I was using the laptop while charging, once the battery gets to around 80% the CPU drops to 0.39 GHz. I am charging in the top right port relative to looking at the laptop as it is in use.
I emailed support today about it because I have heard of folks going through RMA processes. It’s not a huge deal for me, I usually just unplug my machine when it gets up around 80% and the CPU returns to normal. But still, I would like to see a resolution. It sounds like it’s only somewhat reproducible in Framework’s hands, which makes fixes tricky.
The question is, would an RMA fix this? I already went through an RMA before and this is a replacement unit that I received.
Oh rats…I wonder, it seems that some people aren’t seeing this issue, but others are. That’s at least why I thought an RMA might help (at least to give them a machine - mine - that I know is having this problem reliably) but your evidence is saying it might not help.
I guess we should just be patient, I know Framework is working on it!
2 Likes
Finally got around to reseating the EMI stickers. (For anyone who might not want to try - it’s actually really easy, as long as you go slow and use a set of wide tweezers.) I also reseated the CMOS battery while I was at it - seems to be sitting more snugly in the holder than it was before.
Laptop refused to boot initially without power (probably not surprising, assuming the CMOS battery retains the laptop’s memory of having been powered on at least once?). Came up, no obvious CPU speed issues, though it was at 87% charge. And once again the clock was out of date… We’ll see if that’s resolved by the battery reseat.
If reseating the EMI stickers turns out not to have fixed it, I’ll just open a support ticket, since I don’t have another high-power USB-C charger to try it with.
1 Like
Welp, turned the computer on this evening (same refusal to boot without wall power) and it still thought it was 4pm 4 days ago, so clearly whatever’s happening with the RTC on my machine is still busted. Battery was ~97% so no throttling.
Time for a support ticket…
1 Like
I wonder if it could be as simple as a bad RTC battery?
2 Likes
Pulled it out and checked it with my multimeter - 2.01V. So that’s a definite possibility. That would explain the RTC being bad, at least. Still TBD whether the 0.39GHz behavior was fixed with the EMI stickers.
That didn’t help me although YMMV. Still waiting on some official response on the situation.
Have you opened a support ticket? Based on some other threads on the forum, I think this site is not treated by Framework as an official support channel. I got a reply to my ticket (from Nirav!) in about a day.
@comfydragon Your RTC battery is low, it needs to be at a minimum of 2V to function correctly. I would suggest leaving your laptop plugged in for 24 hours which will give it time to charge up again. After 24 hours it should be at around 2.8-3V.
If you are still having issues after trying to charge it for 24 hours, open a support ticket and we can get you a new one.
2 Likes
Thanks, somehow I glanced over the fact the battery is rechargeable. I’ll leave the laptop plugged in as you said and see what happens.
1 Like
I can confirm that this behavior is related to chargers.
My CPU frequency is limited to 0.39GHz in task manager when charging with SlimQ F100. But charging with Lenovo ADLX65YLC3A or Momax UM20 does not have the issue.
It seems not related to the wattage because SlimQ one is 100W while Lenovo and Momax are 65W.
Maybe it is related to the PD controller of the charger?
2 Likes