Framework 13 AMD Ryzen 7040 BIOS 3.03b

Framework Laptop 13
63

I installed the update using the usb boot method. It appeared to work correctly, I saw a flashing screen and “update successful”.

However, when I run dmidecode to check my firmware version, the output looks like this:

        Vendor: INSYDE Corp.
        Version: 03.03
        Release Date: 10/17/2023
        Address: 0xE0000
       product: AMD Ryzen 5 7640U w/ Radeon 760M Graphics
       vendor: Advanced Micro Devices [AMD]
       physical id: 4
       bus info: cpu@0
       version: 25.116.1

The date and version number appear to be older than expected. ‘3.03’ instead of ‘3.03b’ and a date from last year.

It this expected, or should I try reflashing?

64

As far as I understand the initial post, this is expected, as the actual BIOS version stayed the same. The only update included in 3.03b is the EC version. And I don’t know if this version number is visible somewhere (it was visible on the update screen).

5 Likes
65

Updated here, no change - but I didn’t have an issue with my third party charger ( " UGREEN 65W USB C Charger Plug 2-Port GaN ") so consider this a good negative test :slight_smile:

1 Like
66

Is the source code for the embedded controller in the Ryzen Framework 13 available anywhere?

67
4 Likes
68

@Kieran_Levin Any updates you have to share regarding the full 3.04 release, such as the boot-related issue that originally prevented 3.04 Beta from being released a few weeks ago?

Any information and help is appreciated. I’m eager to see the improvement can be brought by the full 3.04, and perhaps finally being able to promote the AMD13 as daily. Thank you.

2 Likes
69

Looks like its doing a lot better, not perfect though. I dont have an issue with attaching of power sources like a dock to a dead battery Framework and it not working. The PD IC seems to power on and do the handshake now (I think this was the bug but at least for me this behavior is now now solved). So far what I have tested with is handshaking correctly. Anker GAN chargers, Kensington Docks, Dell Docks, HPI docks, Logi Dock, etc all seems to work. Will keep testing but I am running out of devices to plug it in and try to find issues in my corp test suite at work.

70

I did this same thing, but I still get the message on Fedora 39 that it can’t install because that OnlyTrusted line hasn’t been added…guess I should reboot eh?

71

I’m also curious when it will be officially released (approximately)

72

Potentially some time, track record hasn’t been impressive. Although communication within the past 5 days is pretty great.

1 Like
75

I have tested the new update with various chargers and haven’t noticed any improvements. I have yet to find single USB-A port that will charge my Framework. Be it a dedicated single port charger or some multiport chargers, USB-A is a no go. I have yet to test my tiny 35W USB-C charger, but I am not hopeful…

My previous Honor-laptop took every charger you threw at it, be it ancient 5V/1A to modern 20V/3A, coming from C to C or A to C. It didn’t give a shit and just sucked all the electrons.

My framework on the other hand is nearly as selective about power supplies as the Amish…

76

On my side of things, the beta made my 30W charger working properly.

77

Same, I am especially hoping a newer firmware can cleanup the i2c errors that fly around dmesg on occasion.

[ 2659.309293] ucsi_acpi USBC000:00: ucsi_handle_connector_change: ACK failed (-110)
[ 2685.068778] i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
[ 2848.129662] i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
[ 2951.717997] i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
[ 3792.769549] Lockdown: systemd-logind: hibernation is restricted; see man kernel_lockdown.7
[ 3829.730585] i2c_designware AMDI0010:00: i2c_dw_handle_tx_abort: lost arbitration
[ 3834.769293] i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
[ 3887.272118] i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
[ 3894.547399] i2c_designware AMDI0010:00: i2c_dw_handle_tx_abort: lost arbitration
[ 3946.550057] i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
[ 4355.942117] i2c_designware AMDI0010:00: i2c_dw_handle_tx_abort: lost arbitration
[ 4508.471982] i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
78

The 3.03b bios fix is unrelated to charging from dumb USB-A chargers. The fix only affects charging from (underpowered) USB-PD capable chargers (which always use USB-C).

1 Like
79

USB Power Delivery does not require a USB-C port. USB-PD 1.0 applies to basically all USB connectors. Going past 100W does require USB-C as well, which is defined in PD2.0 and beyond and does apply to USB-C only afaik.

On another note I can confirm that my previously incapable 30W USB-C charger finally works with my framework. I guess it is at least some improvement…

On another note: for the framework to charge, does it always require a PD charger with 20V support? Or could it also charge with 12V or 15V? I only ever saw a 20V handshake when I tested my chargers…

P.S. This was the “35W” charger I previously mentioned. Unfortunately it is just a 30W charger :sweat_smile:

1 Like
80

The framework seems to be able to Charge with all PD voltages, even as low as 5V.
The only charges I have not been able to use are chargers without PD.

Surprisingly, even the 67W Xiaomi charger is working fine (using the correct USB-A to C cable with the additional pin), and it even handshakes at 20V. I didn’t even know that the charger supports that voltage.

81

Proper USB PD requires communication over the CC pins, and these are only present in USB C. What you mean is probably “USB Battery Charging” / USB BC which is defined in the USB PD standard, but just supplies standard 5V 3A. This is still unrelated to the fix here, which affects the voltage ramp-up and communication over the CC pins.

82

Both of you are somewhat correct.

PD Version 1.0 was actually defined for USB-A. It existed prior to USB-C and was just never implemented. It does not exist in the real world.

The current versions of PD that are actually implemented are not backwards compatible to that first attempt and have replaced most of the stricter concepts of PD 1.0 with better and more flexible alternatives and require USB-C with its CC wires. So any PD today is only valid in combination with USB-C.

USB BC is not defined in the PD standard. It is its own standard designed for USB-A and similar. USB-C is just backwards compatible to it so that legacy adapters work for example with USB-A chargers. And BC is limited to 1.5A for a total of 7.5W. The 5V3A mode of USB-C is defined in the USB-C Standard and called “Type-C Current”, same as the 5V 1.5A mode that USB-C supports independent of USB BC.

And EPR for going above 100W was only added from PD 3.1 onwards.

At least the 12th gen Intel version works with 9V, 12V whatever you want. There were some problems with 11th gen, where 15V I believe was excluded post launch with BIOS updates because it had some hardware problems with that. But since the 12th gen fixed those hardware issues I would presume all the voltages are supported since then. And Framework never had those strict limits to only 20V while operating like many other notebooks have for historical reasons. Though it may not be smart to use that while the FW is running, as chargers that do not supply 20V do not have the power to reliably supply the notebook with more power than it requires itself (i.e. charging the battery needs to be stopped often to supplement with power from the battery).

3 Likes
83

after this update the FW13 AMD supports most if not all PD ratings except 5V3A, which either won’t charge, or supply 5V3A charging the power bank… Some other computers with PD 20V5A input 5V3A output have the same problem

84

When can we expect a patch for the LogoFail vulnerability? I’m mostly concerned about that.

Also, if anyone is looking for a great charger, I got this one Nomad 130w. It is a total beast, and has worked flawlessly for me. It’s pricey but on sale right now, 20% off. Can do a 100/30 split and even 70/30/30. It’s small and has great build quality. Paired it with Belkin USB-C Cable. Amazing combo

2 Likes