Hello Framework. What is your experience with battery life?
I bought my 16 from Batch 8, so it’s been my daily driver for nearly 2 years. Since then, the battery is due for a replacement as it is now at 41% health. I checked the battery controller, and it reports it only went through 27 cycles. My laptop usually lives in my room on my charger, but that’s still far less than what I expected (I was guessing somewhere around 500 cycles), but I also didn’t expect to go past 1,000 cycles either (The battery store page reports that it should retain 80% capacity after 1,000 cycles). I run Linux – could this be a Linux thing? How accurate is the cycle count on the battery controller, anyway?
How are everyone else’s batteries?
Here are my statistics for the past week (reported on KDE on Manjaro Linux):
It is a known issue (reported to FW in Jan) that the Linux sysfs > class > power_supply > cycle_count metric (read by programs like upower and used in Plasma’s battery viewer) is incorrect for Framework batteries after 255 cycles due to an ACPI implementation bug. The counter rolls over, effectively reporting correct_count % 256. You can read the correct value from the EC with either # ectool battery or # framework_tool --power -vv. The BMS counter should be accurate.
The measured health percentage can drift over time, so it’s recommended to perform a battery calibration periodically (especially if you use charge limits) to ensure you have an accurate reading and so the BMS functions reliably (e.g. so you can use the full capacity available). The steps I linked are somewhat imprecise for laptops though: I personally make sure it has reached full discharge by letting it “die” to avoid the the slack from their approximate method. Typically I do this by using it until 1% and then rebooting to the UEFI to let it “die” there where the OS is unaffected. This method does not require the long wait step either, although you may choose to let it rest for 5-10 or so minutes before charging again anyway.
With a health reading as low as 41%, a calibration probably won’t dramatically change the measurement (unless it’s really gotten out-of-whack), but hopefully these steps give you a more accurate picture of the true situation before you replace it.
Do you limit max charge while it’s on the charger for extended periods? Time spent at high voltage (high charge %), and at the low end, both induce stress in li-ions. People choose different max charge cut-offs for their needs, I personally do 80% for my laptop & phone when they are on charger for an extended period.
I use the firmware’s default settings, which is to limit to 80% charge
after it’s on AC power for 3 days, and then move back up to 100% after
30 minutes of battery.
