FW13 AMD (7040) — eDP-only flicker after exhausting software/driver fixes: cable or panel?

Framework Laptop 13 Linux
1

Hi all, I’ve been chasing a persistent flicker on the internal eDP panel of my FW13 AMD 7040. I’ve exhausted what I can do in software and from a kernel-log standpoint the driver looks healthy, so I’d like input on whether the next step should be a replacement eDP cable or a panel replacement.

System

  • Framework Laptop 13 AMD, Ryzen 7040 (Phoenix1)
  • Fedora 44, kernel 7.0.4-200.fc44.x86_64
  • BIOS 03.18 (latest)
  • Internal eDP panel at 60 Hz (native)

Symptoms

Screen flickering when connected to the AC power.

External displays attached to the same machine via expansion bay USB-C never flicker — only the internal eDP.

Kernel parameters already in use

amdgpu.abmlevel=0
amdgpu.sg_display=0
amdgpu.dcdebugmask=0x410
pcie_aspm=off
i915.enable_psr=0

What I’ve ruled out

  • Refresh rate: 48 Hz and 60 Hz both flicker (60 Hz is native, not 120 Hz as I’d initially assumed).
  • GNOME automatic brightness (ambient light sensor): was causing a slow brightness drift I’d confused with flicker. Disabled via gsettings set org.gnome.settings-daemon.plugins.power ambient-enabled false. Separate, resolved.
  • ABM and PSR: disabled via the kernel params above.
  • GPU performance level: forced high via power_dpm_force_performance_level. No change.
  • eDP cable reseat: opened the laptop and reseated the cable on the motherboard. No change.
  • amdgpu driver bugs: journalctl -k -b and the last 7 days show zero amdgpu page faults, ring timeouts, or GPU resets. Driver path is clean. dmesg shows kernel applies a panel backlight quirk (min_brightness: 0) and a custom brightness curve for this panel automatically.
  • USB-C PD / UCSI: I noticed flicker sometimes coincides with slow-charge episodes, and journalctl shows recurring ucsi_acpi USBC000:00: GET_CABLE_PROPERTY failed (-5) errors. I ran a live correlation test (journalctl -k -f -o short-precise with ucsi|typec|drm|amdgpu filter) and observed flickers occur with zero corresponding kernel events, while PD source stays at 20V/3A. So PD is not the flicker trigger — the UCSI errors are a real but separate issue likely tied to the slow-charge episodes only.

Where I’m stuck
The only data left says it’s hardware: internal-eDP-only, driver-clean, BIOS-current, cable-reseat-doesn’t-help.

2

Please can you post a video of the flicker, and add a link to it.
Different types of flicker are due to many different things, so seeing what it looks like helps.

3

Hey thanks for the reply. This is the video. It’s more like a jitter. If a better video is required then I can make another. Hope it helps thanks!

4

Based on your description.
The fact it works OK on external displays tends to rule out GPU software bugs.
The video shows an intermittent corruption of the image. It becomes offset in parts momentarily but much of the time the display displays accurately.
This would tend to mean the display panel itself is OK, leaving the problem being intermittent connection problems.
So that leaves just two things:

  1. The cable between the mainboard and the display. This has a connector both ends of it, so it would be good to try removing and replacing the cable to ensure it is a good fit. It is a push fit, so can sometimes become a bit loose.
  2. A intermittent hardware fault on the PCB, causing intermittent end-to-end connection problems. Either the small PCB on the display panel, or the mainboard PCB.

Of those two, I would suspect the cable as the most likely cause.
Remove and re-seat the cable both ends, and also look at the cable for any obvious damage.

The FW support/repair guides should help you identify how to get to the cable ends.

If you determine it is the display cable damage, you can get a replacement here:

Note: The above is only my, another user, view of the most likely cause.
It is also, in theory, possible for the GPU software to cause these sorts of effects, or even bad RAM, but if that was the case, I think we would see far more people complaining. So, I mention the most likely cause being the cable itself.

5

Had a similar issue. External monitors worked fine, memtest passed, cables where reseated. It still occured. But only when all of these conditions are met:

  • using internal monitor
  • while charging
  • under high CPU load

Turned out the panel was broken. Support send me a new one for free.

6

Thanks for your replies. I ended up changin the whole display. Now it does not flicker anymore. So it does not seem like a software problem and were more an hardware one.

1 Like