I got the exact same problem with my Ai 350 and Anker 737 charger.
Honestly i love this charger so this is very disappointing. My charger has always worked great with all my other devices. This and the bad wifi on linux really makes me consider returning the laptop.
I know, one good thing about framework is that i can just replace the wifi card with another one and i could also buy a charger but i would somehow expected better first impression for the price of 2100 euro.
I seem to be having the same issue with an Anker PowerPort 3 (A2144). HX 370, BIOS 3.03. Fedora 42 is installed, but the issue still occurs while the laptop is fully powered down.
With ports 1 and 2 (100W and 60W), the charging indicator on the laptop turns on for a split second, then turns off. A USB C power meter confirms the port is active and set to 20V, there’s just no current being sent.
With ports 3 and 4 (20W), the laptop does charge, albeit slowly.
I tried the same charger and cables with different devices, and they all work as expected. I also tested with an Anker 735, and it works as expected. The issue seems to be between the laptop and the PowerPort 3.
For anyone interested. I contacted both Anker and Framework support.
Anker basically told me “Sorry dude, you got that charger as a present and thus have no bill, nothing we can do about it” - which is sad, but understandable.
Framework is currently escalating the ticket, so lets see if something comes out of that.
Man we desperately need some more polished pd controller firmware.
i also have an anker 737 and a previous amd gen (7840u). The power adapter works just fine in this combination. It also worked without a battery plugged in, as that needed to be replaced. So the issue seems to be specific to the newer amd models.
Well there goes my hopes of maybe getting hand-me-down pd-controller fixes from the ai 300 models since they seem to use the same one.
According to the “schematics” they use the same pd controller and power mux chips but different redrivers so yeah there goes that.
I do hope those new redrivers are actually better than the ones on the 7x40u platform otherwise they just produced unnecessary product fragmentation.
Anker 547 and Framework 13 AMD Ryzen™ AI 300 running Fedora 42. Seeing the same problem. It will charge slowly if I plug it into one of the lower power phone ports, but the charging light blinks briefly and goes out when attached to the high power computer ports.
dmesg shows:
[398538.206428] ucsi_acpi USBC000:00: unknown error 256
When I plug it in to the computer ports now.
I ran into the same issue with my Anker 737 120W (Model A2148).
I got my hands on a USB-C PD tester and ran PD Analyser to see whats happening, and tried the following chargers:
with both my AI 370 Framework laptop and my Macbook. The Anker charger + the framework was the only pairing that didn’t work.
Here are the PD logs with each of the Anker, Mac and Samsung chargers after they are connected to a framework laptop or macbook:
I’m no expert in the PD protocol, so please correct me if I am wrong. Both the framework and the macbook manage to negotiate 15V or 20V power with all chargers.
The framework attempts to “DR Swap” which means it wants to swap data roles, i.e. making the charger the USB host without changing the direction of power flow. The anker charger accepts, while mac and samsung chargers reject.
After the anker charger accepted becoming the host, the laptop sends a “VCONN SWAP”. My understanding is that VCONN is whats powering the chips inside the type C cable plugs and is different from VBUS which is what charges the laptop. The anker charger responds with “Not Supported” and then a “Soft Reset” message (I don’t know who sends this) appears, resetting the session and preventing any charging…
The same session reset thing seems to happen with the Macbook + anker, but it retries without doing DR and VCONN SWAP after the first failure.
I don’t know which device is more wrong here, but it seems weird to me that the anker charger accepts becoming the USB host. Whats the point of that?
Maybe someone more knowledgeable about USB Power Delivery can share some insights.
That “soft reset” should not be worrying. That seems to be just how communication with the cable’s eMarkers is initiated.
If you check the PDOs (in Source Cap) that is probably limited to 3A in the first one. And after they confirmed that the cable is good for 5A, the charger updates its Source Cap with >3A abilities. That part is all fine.
Its the Framework ever picking 15V, 1A on a charger that supports 20V that is weirder. And the FW not starting to draw power again after the PS RDY was received for its 20V request.
But the Macbook also only requests 20V, 0.75A at first, but later 4.7A. Is the charger advertising it weird? I think this may be a specific combination of this Anker power supply behaving weirdly and the FW not being as robust as it could be in handling the weirdness…
Because no problems with my Anker 140W charger (the one with the display, before they renamed them “prime”).
Issue confirmed on
Framework Laptop 13 AI 7 350 (DIY), with Anker 737 120W (Model A2148).
same dmesg USBC000:00: unknown error 256
Tried two cables.
Same cables with powerbank works.
Different power-adapter with integrated cable also works.
Update -
Anker A2669 67W GaN does charge the laptop.
dmessg provides the same unkown error, but does charge.
same for me, any updates on this issue from Framework?
The new AI Models have several Issues,
