Build DIY Framework 13 Ryzen 5 7640U
Memory: DDR5 -5600 - 32 GB (2 x 16 GB)
Storage: WD Black 1 TB
OS: Windows 11 Pro 23H2
BIOS: JFP30.03.05
Windows AMD driver bundle: installed
I received this unit a little over 2 weeks ago, assembled it with my own SSD, loaded Windows and it was great… until it died. It all started with a hard freeze while I as in Windows - no error or BSOD. I performed a hard reset and managed to get back into Windows but only briefly until I actually got a BSOD to the tune of hardware error or failure. It disappeared before I could take a pic but now it no longer boots. No flashing lights, nothing. Sometimes it will get really hot and the fans will spin up to 100% interrupted / stopped only by a hard shutdown. This is all without even making it to Windows.
I have tried removing all modules and booting. No change.
I have reseated the SSD and RAM. No change.
I have tested RAM sticks one at a time in either slot. No change.
I have removed the SSD and tested its health and it is 100% OK.
The BIOS allows access but if I attempt to make a change and save it, hard freeze.
I have disconnected the battery and performed a hard reset (I don’t see where the CMOS battery even is on this unit). No difference - the BIOS remains as is.
I managed to disable secure boot but all this has done is force me to recover bitlocker with my online encryption key. I have managed to get into Windows a time or two for a couple minutes before a hard freeze occurs. Not enough time to examine logs or anything else.
This smells like hardware failure and I have initiated contact to support a day ago and as of yet have not received a response. Evidently, everything’s delayed due to the recent release of the new Framework 16. I find it frustrating that I can’t get a prompt response because everyone is busy supporting that while I am dead in the water with a brand new product. I’ll try to be patient but I’m afraid that if I don’t get a response within the next couple of days, I will request a return while I’m still within 30 days of ship date. Not too jazzed right now.
Any suggestions or useful info out there that I can try?
Sorry to hear that you are having issues Frank, our support queues are a bit backlogged and replies are delayed. Hope we can help you troubleshoot that issue here and find a quick solution.
Welcome to the community! That is odd about your FW13 with the Ryzen 5 7640U. You have gone through what most everyone would have thrown at it. If you leave it in the BIOS screen just idling does it still lock up/shutdown? If so then something is panicking at the hardware level and the system can not recover nor does it have a means to report on it.
I don’t know if there is a way to force a reflash of the same BIOS as that is suppose to reset some NVRAM registers too. If those did not get reset properly when new BIOS is updated.
If you can get into the BIOS, try setting it to the Defaults (F9 or F10) then rebooting the machine. There was a post that mentioned in rare instances the FW16 BIOS update did not reset a register and defaulting the settings in the BIOS then rebooting corrected it. Just a thought that came to mind as they are both AMD hardware and similar BIOS structures.
If you are feeling ambitious and support has not gotten back to you; you could try disassembling the board from the laptop shell. Take it out, and blow gently on both sides and inside the laptop shell, in the really rare event something conductive managed to get in there and bridge a few connections where it should not have.
Another remote possibility is the thermal paste is either nonexistant or there is a gap so it is not conducting heat like at all. The change in temp for this processor in open air (i.e. no solid heat sink connection) could explain the sudden freezes. If it heats up so fast it can not throttle itself down or it doesn’t throttle itself down fast enough a system will exhibit this behavior. This comes from experience in the 90’s when processors would cook themselves because some of this protection circuitry just was not implemented on consumer chips.
Just a few ideas I thought of after reading your post. Another way to narrow it down to a thermal issue is to put it in the freezer for a bit then take it out and boot it up. If it lasts longer before lockup it is probably thermally related. Though this could still be a bad board, bad traces, bad components that work ok until they warm up and separate just enough to cause issues.
Hope you hear from the support team sooner than later.
Those are some good thoughts. I’ve still not heard from support.
Honestly, I have not left it sitting in the BIOS long enough to notice what it does after idling for some time. It does seem more responsive on cold startup though vs after it has warmed up but hot or cold, it will freeze as soon as I try to save a setting up to and including the one that sets everything to default. I did try the freezer trick though and I was at least able to get in and actually set factory default successfully. That’s it though. I was able to get into Windows briefly before it froze again but the laptop wasn’t hot at all.
I’ll see if I can disassemble, blow things out and check the “state of paste” - maybe I’ll get lucky!
How long was it working for before it died?
Was it like seconds, or did it work for a few days and then die?
I am trying to understand whether the FW13 worked for a few days OK and then failed, or whether its never really worked at all.
Did it work long enough to go through the install windows process, or was windows already on the nvme before hand.
I was fully using it normally with Windows 11 for a little over 2 weeks. Working perfectly no issues at all. This is a failure scenario not a “never got off the ground” scenario.
You could look in the logs for “MCE” messages or “Machine Check Exception”.
I don’t know how to do that on Windows, they just appear in syslog on Linux, on the reset/boot after the failure.
MCE, if you see errors, generally points to a failed CPU.
This could be a RAM problem, but to prove that you would need to run “memtest86”. You can download a usb bootable version from the memtest86.com web site. It comes pre-installed on Linux systems.
The problem is that I’m not able to get into the OS at all and if I do, it’s only for about a minute before it freezes up again so accessing any logs while in the OS is out of the question. I could slave the drive to my other Windows machine and pull them that way if I can find the event logs. I will get a USB memtest going to see what it does and record my results.
In other news, support did respond this morning and they asked me the battery of questions around what kind of SSD I am using, whether or not Windows was installed with latest patches on board, whether or not the AMD bundle and BIOS were updated, etc… They also provided instruction on how to manually reset the BIOS on the mainboard I had to depress both of these at the same time 10x for 2 seconds at a time while waiting for the MB LEDs to blink each time in between.
It did something but it’s actually worse now. Now, boot goes no further than the Framework logo or just sits at a black screen. I’m also not consistently able to get into the BIOS again now either. I made it in once and as soon as I selected the main settings section, the screen went black.
There is something very obviously wrong with this mainboard. At any rate, support is now involved (thankfully) and I’ve submitted all this. Will see where this takes me.
It might be worth removing the nvme storage. Then see if you can get into BIOS and save settings.
It might be worth also removing the wifi card and again see if bios setting save works.
Essentially removing everything you possibly can to narrow down the fault.
I’ve removed the nvme and the wifi card - along with every expansion card and still getting bad behavior. Maybe able to get in and boot into BIOS once or twice, even got the “no boot device detected” splash but now it’s back to the bad. While in there, I also removed the RAM too.
I’ve also double-checked to make sure I wasn’t doing anything like plugging a USB-A into into either rear slot (1, 3) - no matter all of them unplugged make no diff anyway.
Growing concerned now. I had a couple of different support personnel contact me with things to try which was now 2 business days ago. Nothing has worked and I have sent in videos of everything they’ve asked for - silence now. I’ve sent a reply today asking for next steps in getting a proper RMA for the mainboard. - Leaning more towards a return now and chalking this up as a poor experience. This is bad.
That is rough @Frank4, sorry you are getting the battery of suggestions from support without any resolution. Based on what you are describing something is going on at the hardware level with the board, clearly resetting the firmware or other defaulting is probably not going to fix this.
I know support has to do their job. At this point though, sending you a replacement board to swap out or replacing the whole machine would be a more prudent fix vs. the bad experience and your wanting to just return the machine. This is just my personal opinion though from years in the customer service industry.
Maybe @Destroya can check on something to expedite this. In either scenario, they are either going to get a questionable board back or loose a client and still end up with the questionable board back along with the rest of the laptop.
The Ryzen machines are still really new and amazing at what they can do. If you do decide to return the whole unit, we do not see the same number of issues with the later Intel boards on the forums. This is if you are still willing to try Framework out on the Intel side of course.
I bought a factory 2nd Framework 13 to replace an old iMac at work and its amazingly fast for the price. This is the 11th gen stuff too. The 12th and 13th gen are even better. Now the guys in the shop can take it with them when they work on equipment and if they bang it up, drop it, spill oil on it, etc.
Thanks for the ping pkunk, this is not something we can expedite actually.
I see that the case is escalated internally, Frank should receive a reply soon. Please do not send another email meanwhile since it might cause extra delays.
We had a backlog and we are mostly recovered, you should receive responses faster now, thanks for your patience.
@pkunk@Destroya thank you both for chiming in with the additional insight and moral support. I did send something yesterday asking for status / next steps - I hope this doesn’t set me back any more than I already am. Will refrain from doing so any further.
Seems unlikely that it might come to this now but if I don’t get a suitable path to resolution say by 7/9 - I’m going to request a return to cover myself. That is 2 days prior to the 30 days from ship date and I don’t want to be saddled with a broken machine and then have to spend even more time trying to get it to a working state.
One thing it could also be is a battery problem. If you can get into the bios, there is a setting that disables the battery for the next boot. So you could test it without actually removing the battery.
The battery connector has pins that are easy to bend, so if you do remove the battery, be very careful.
My experience with FW support is that they first collect as much information as they can, logs, pictures etc. Then pass that to the next team, be it the windows team or linux team etc. If they conclude it is not OS specific, it might get passed onto an engineering team. It might take a while to get to the answer, but I found that they did get to the right answer in a reasonable amount of time (days). To be fair to them, I was always slow. Maybe a day delay in me answering each question they asked.
This morning I powered it up for the first time in several days and I actually got into Windows but it froze about 3 minutes into my excursion and it was back to the failure behavior I’ve grown accustomed to. Unfortunately, I can’t consistently really get into the BIOS anymore. I have weird power on behavior now where the power button will flash a couple of times and then do nothing. I’ll hard reset and I’ll tap tap tap F2 and maybe the Framework logo will appear maybe it won’t. The one time I got in the last several attempts, it completely froze before I could look around for this setting your describing. After that, I hard reset again and got nothing but black screen. Then the fan spun up to 100% with nothing on screen whatsoever and I had to hard shut it down.
Everything I’ve described here is literally all I get from this thing now. I’m not going to waste anymore of my time and energy on it.
Just sitting here waiting on support to get it together. Tomorrow will be a week since I last heard from them and the clock is ticking.
Sorry that you are having issues. Please follow up with the outcome, whether a root cause or return/exchange. If I had to guess I’d say it’s either RAM or main board.
I don’t think there is any harm in sending them a “status?” email as you did. Your request might have got lost in their system. It has been 8 days now since you first raised the problem.
The community here is pretty good, all of us helping each other out and maybe saving some of the load on the FW support teams.
I think I have helped a few people here, but I think what you are describing is definitely a hardware issue, and not something the community can fix.
My guess is either Mainboard or Battery replace, or both.
