Framework 13 AMD Hard Crashing Issue

I’ve raised a support ticket with Framework, but thought it prudent to document the issues with my Sept 2024 Batch Framework 13 AMD (DIY) laptop in case others have the same problem.

October 15th
Here’s the contents of the initial support ticket I created:

I received my Framework 13 AMD about three weeks ago. It functioned well initially, but around October 9th, I began experiencing random hard crashes (screen goes black, and then the system quickly reboots).

System information:

  1. NixOS (stable), but tested with Fedora 40 live USB as per support guidelines - issue persists
  2. MediaTek WiFi card occasionally fails to load, I haven’t resolved this but the issue is preset with the card working properly, not working or disabled.
  3. No crash-related entries in journalctl
  4. BIOS v3.05 (latest); fwupd reports no available updates
  5. Work setup: USB-C monitor with 65W charging
  6. Home setup: USB-C dock driving external monitor via DP Alt Mode

Attempted solutions (unsuccessful):

  1. Reset BIOS Battery Charge Limit from 80% to 100%
    1a. Performed “Load Optimal Defaults” in BIOS
  2. Issue occurs primarily when connected to a charger
    2a. Definitely happens at full charge; uncertain about during charging
  3. No consistent trigger identified; occurs during various activities or idle periods
  4. Tested different USB-C chargers and docks, including a regular 100W charger with no other peripherals connected
  5. Problem affects both USB-C ports (expansion cards in rear-most bays on each side)
  6. RAM: 2x Team Elite 32GB 5200MHz
    6a. Tested with single stick due to potential 64GB incompatibility
    6b. Concerns about 5200MHz compatibility vs 5600MHz requirement (seeking confirmation)
  7. Tested on Linux kernels v6.6.54 and v6.11.2 - issue persists

The problem only seems absent when running on battery power. I’ve conducted further tests using a portable USB-C dongle with pass-through charging and HDMI (no DisplayPort). When using this dongle to connect peripherals without power, the laptop operates normally. However, if a USB charger is connected to the dongle’s pass-through port, the crashing issue eventually returns. This isolates the problem specifically to when USB-C power is connected, regardless of the adapter or dock being used.

October 16th

I’ve now confirmed the issue is also present under these conditions too:

  1. Using a single stick of Crucial 32GB 5600MHz RAM
  2. Whilst on running on battery power
  3. Whilst power connect but the battery is charging

It seems this hard crash issue is far less likely to happen while under battery power; but otherwise it seems unrelated to USB chargers, USB docks/peripherals and battery.

One outstanding idea is to boot into BIOS and see if happens from there too. If it does, I will then retry without an SSD installed too.

November 8th
I’ve confirmed my delivery details with the RMA team, a replacement mainboard is being dispatched :slight_smile:


Related threads by other users:

Power supply problem (note: I don’t have issues with fans ramping up like this user does)
Framework 13 AMD crashing on different OSs - #5 by McAllaster

O.T: Of course this issue started after doing a nixos-rebuild switch --upgrade … leading me to spend a whole bunch of time thinking my Linux setup was the problem. Nope, not related at all and just bad timing haha!

1 Like

Just an update on where I’m at:

Framework support seems to be quite responsive, only taking a few hours to get back to me yesterday :slight_smile:

Among a few other things, support has asked me to try and replicate a hard crash on battery power with all the expansion cards removed. This is a slow process, as it seems I’m much less likely to have a hard crash issue on battery.

Nonetheless, I’ve reverted to working from my old laptop and have my FW 13 chilling on battery with no cards installed to see if the issue still happens. So far I’ve had no crashing occur, once the battery is depleted I’ll retry the same test with the card re-installed.

1 Like

My latest response to Framework support (answering their most recent questions)

Testing without expansion cards

I’ve run the laptop on battery with the expansion cards removed for approximately 7 hours and experienced no crashing. If I connect a charger it directly into the expansion bay port it does crash again (has happened with two different chargers).

I’ll conduct another on-battery test with expansion cards fitted and see if the above behavior (7 hours run time) is common or not.

Memory

The original memory I purchased was 2x Team Elite 32GB DDR5 5200MHz (non-kit) | Model: TED532G5200C42-SBK

I then purchased 1x Crucial 32GB DDR5 5600MHz | Model: CT32G56C46S5

SSD

Team MP44L 2TB M.2 2280 NVMe PCIe Gen4 SSD | Model: TM8FPK002T0C101

Chargers

  • Alogic USB-C Dual Display (DP Alt. Mode) Dock DUPRMX2-100

  • Dell USB-C monitor (this is at work, fairly confident the model is P3424WE)

  • Heymix 100W GaN X100-2C1A-P49S-AU

  • SlimQ 100W USB C Charger F110

All of these chargers/docks were used by my previous laptop without issue.

I’ve had the crashing happen using 4 different USB-C cables that have been working fine for the past year, I’m super confident that 4 different USB-C cables did not fail on the same day :slight_smile:

Not sure if it’s related, but I’m using power-profiles-daemon with default settings in my NixOS setup. Unsure what Fedora 40 does out-of-the-box, but as I said the crashing happen even just running the Live USB.

1 Like

I have dealt with a very similar issue. With 2x32gb ram it happens frequently with one stick less frequently.

The support team ran me through so many tests and eventually sent me 1 replacement.

First time testing it out and within 10 minutes I get a crash again.

Reached out again to see what else they can do. :frowning:

1 Like

I’m having a similar issue and just posted about it in this other thread: Framework 13 AMD crashing on different OSs - #8 by Paul_Brown

Try removing the wifi card and see if the problem goes away.
Not just disable. Physically remove it from the slot.
The FW will power off if it does not like a PCIe device.
It sounds to me that you wifi card has an intermittent fault.

1 Like

Sorry for the delay in an update, but here’s where I’m up to:

Support got me to try a whole bunch of things including:

  • Physically removing WiFi card
  • Removing all expansion cards
  • Sending a dump of logs + configuration
  • Doing multiple mainboard resets (slowly pressing a SW3 10 times)
  • Asking for photos/videos of the issue :face_exhaling:

I did several days of testing with/without the expansion cards fitted. It was a really slow & tedious process to try and get the issue to happen when running on battery. I was never able to get a crash to happen on battery with the cards removed. It could be pure luck though, as even with them fitted the laptop could run for many hours before crash.

Connecting any charger directly to the expansion card USB-C port still caused frequently crashing, so I don’t think there’s an issue with any of my expansion cards.

My ticket has been escalated, so I guess I’m just waiting for a technician to review my logs. Presumably, they won’t find anything either and hopefully green light a mainboard replacement.

I’ve had another request from the support team asking to try using the official charger. I don’t have one, which is clear in my order :confused:

I appreciate that support is able to respond fairly quickly, but just feels like they’re giving me the run-around now.

Having invested a significant amount of my time attempting to troubleshoot the issue, I’ve explained that we’re at the point where they just need to provide replacement parts or a refund.

Very similar problem for me. 2x 16gb crucial ram. Tried removing 1 stick and i haven’t got a crash since. Tried again with 2 and it crashed within 30mins.

Basically only when gaming btw. Really annoying because of dataloss and with 1 stick the performance is meh

Thanks for the suggestion. I removed the WiFi card and used a USB WiFi adapter. Just had another random reboot after a few more hours of usage.

Update from end of last week: Support was concerned that all four of my chargers/docks may have issues/incompatibilities and suggested trying the Framework charger.

I’ve firmly reminded them that the crashing occurs on battery power too (just takes a lot longer) so it’s not related to chargers or USB devices.

Hopefully they’ll finally agree to a replacement mainboard, I’m pretty sure I’ve exhausted all practical options to isolate the problem on my end.

I have seen this symptom once on a FW16 amd.
I have not been able to reproduce it a second time.
Like you. The syslog has no entries at the time of the crash.
No MCE errors after the reboot.
I was using the ethernet port card. Wifi was disabled to the time.

My support ticket has been reviewed and they’ve agreed it seems most probable that my mainboard has a fault. It’s now with the RMA team, hopefully this process doesn’t take too long so I can use my new laptop again :slight_smile:

For my case in the United States it took about 1 and a half weeks from approval for a new mainboard replacement to me receiving it.

That’s decent! I’ve confirmed delivery details with the RMA team over the weekend, should see my replacement mainboard soon too! :slight_smile:

I’m very much looking forward to using my FW13 again.

This might be an EC firmware bug that only happens after the EC has been on for a while.
One way to reset the EC is to power off the laptop and leave it overnight.
That is powered off, not standby,sleep or hibernated.
It should then take longer before the issue returns, if it is an EC firmware problem.
EDIT: The laptop also needs to be unplugged from the PSU. If the laptop is plugged into a PSU, it will not reset the EC.

Cool, I’ll give this a try. Just confirming I don’t need to disconnect the battery?

You do not need to disconnect the battery for the EC test.
EDIT: The PSU needs to be disconnected.

Well my laptop has been off (shutdown) for several days, I just booted it up and got a crash with a few mins.

Replacement mainboard is due to arrive tomorrow, fingers crossed it works!

The EC can cause these symptoms.
To find out if the EC crashed, you can use the command “ectool panicinfo”.
If it has an entry in there, the EC crashed, causing the reboot / unexpected power off.
To reset the EC.
Remove the PSU and power off the laptop.
After the Power button LED goes off, the EC will reset 40 seconds later. So, leave it off for 1-2 mins, then power on.
After the EC reset, “ectool panicinfo” should be empty, until the next crash.

So, I would be interested to see if there is any ectool panicinfo output.
An example of the output is:
ectool panicinfo
Saved panic data:
=== PROCESS EXCEPTION: 00 ====== xPSR: 61000000 ===
r0 :00000000 r1 :200c738c r2 :00000000 r3 :00000000
r4 :100a6328 r5 :200c738c r6 :00000000 r7 :00000001
r8 :00000001 r9 :100afb49 r10:100a7180 r11:00000000
r12:100ad781 sp :200c7338 lr :1008b551 pc :1009f2da

cfsr=00000000, shcsr=00000000, hfsr=00000000, dfsr=00000000, ipsr=00000000

The panic info is stored in the RAM of the EC. So it survives a reboot of the CPU, but does not survive a EC reset.