I have been using a FW16 (7840HS, iGPU) for about two weeks. While I like it very much in general, the battery charging is horrible. I tried multiple chargers, even bought a 140 W one at the request of Framework support. For all of them, charging is erratic. Despite the lack of any substantial CPU load the battery is discharging only to start tickle-charging at a later point in time taking ages to reach the configured charge limit.
I have exchanged an annoying amount of emails with Framework support without any clear outcome.
Has anybody seen such erratic charging behavior with your machine?
I am using Linux, but if I understand correctly charging is handled by the EC, so the operating system shouldn’t play much of a role here.
Agreed. I’ve used my FW 16 with a number of non-framework chargers, including the dinky low-wattage ones that come with my Pixel phones. But, others have demonstrated that some devices do not cooperate with the EC very well, i.e. in the case of the Anker 737 external battery’s controller. This may have been addressed in the beta 3.04 BIOS. It sounds like you’ve either got a lot of chargers that have something in common with the 737’s controller, or a laptop that needs at least one part RMA’d. Either way, something’s clearly not working as designed if it behaves as you’ve described. Additionally, the OS does not matter. The device will (or in your case should) charge even without an SSD installed.
What’s interesting is the discrepancy between the battery’s charge state and the state that is supposed to be reported. The operating system only gets to know the first number, not the display number.
There are people reporting Nexode 160W being incompatible here, and Anker 737 here. It’s unclear to me though if one of them means a powerbank while another just a wall charger, or not. But definitely you can see both brands not being 100% compatible, and cables used also playing a role
Wait: the Nexode 160W works very well in standard use and in charging.
Problems appear when you heavy load the dGPU (the 7700S) using the Nexode. Charging wise works wonders. If you don’t have the dGPU works as expected.
I know some of my other brand laptops have a BIOS setting to restrict battery charging to 80% or some other <100% setting to extend battery life overall. Does the FW? If so, the OP could have this set in BIOS and when the charge hits 80% it slows down to trickle charging. Perhaps the O/S misreads the set battery charge limitation and dynamically recalculates battery 80% as O/S 100%?
Framework support has sent me through hours of troubleshooting - with no solution. Either the hardware design is flawed or my unit is defective. Support doesn’t seem to care, though.
This is incredibly frustrating and I don’t want to picture what happens if the laptop is really damaged and not just charging is affected.
I’ve successfully charged my FW16 off the two official chargers, an old 100W Hyper Juice charger, a less old 100W Hyphen-X charger, an 87W MBP charger, and a 100W battery pack.
It won’t charge from a Steam Deck charger, that I’ve seen. Guess 45W doesn’t cut it.
I’d thought that feature was an Apple original – to placate users, the hardware does what it wants and rounds up a few percent on the charge level. I’ve got a MacBook with two percentages in the tray, the actual percentage from a third-party app, and the official, higher, percentage from Apple’s own battery widget. I’ve never liked that kind of hidden-complexity / dishonest UI myself. iPhones do it too.
… poking around the forum a bit, it looks like there’s a fair number of known issues with various chargers, especially from these two threads:
this post from that thread sounds a lot like what you’re describing, has to do with USB cable interactions. What a mess.
I don’t think these are related. I just verified with a Fnirsi FNB-58 USB meter that the power supply is delivering plenty of current. When the battery discharges, it does so with < 0.05 W, so nothing substantial that could be explained by insufficient power from the power supply.
In my Win10 Pro, Ubuntu and Kubuntu, the indicator of battery capacity displays as the limit I set in BIOS when “fully” charged to the set limit of 90%, and when battery is at the set limit, there is no notification of charging happening.
I’m not sure I follow your point. Like the measured current on the cable fluctuates with variations in power usage laptop-side without fluctuation in the battery discharge rate?
Most of the posts I found interesting were talking about failures of protocol negotiation between the Laptop, cable, and power sources. Even down to people putting Anker power banks in 100W modes (or using cables supporting 100W or less) and having success where the 140W mode would be glitchy.
With Framework being, at least historically, slow on UEFI updates, and the entire USB industry being… how it is, higher power USB-PD is probably going to suck for a while. A standard is only a standard to the extent that all the individual pieces of hardware comply with the standard… if it’s new, everybody probably still has bugs or ambiguous incompatibilities in everything that they haven’t identified through testing yet.
Maybe you could get support to mail you a 180W charger for testing purposes. It really sounds like known compatibility problems, and if that’s true, that’d be a recipe for a bad helpdesk experience – like you’re saying, they’d be obliged to essentially try everything even though 80% likely it’s a lost cause. If you bought a charger, maybe they could let you borrow one. It’d be a good data point for them.