I ordered the tools to flash the bios into the chip. Do you have any instructions on how to extract the bios from the capsule file?
I downloaded the bios file and quickly check the content, but I have never flashed a BIOS from a chip before. The instructions above are fairly clear and I will make an attempt at reviving my mystery box mainboard.
Thanks.
For anyone trying to fix a BIOS, the cable that @moparisthebest links to on Amazon is available on AliExpress for much less. I plan on using it with my XGecu T48. I don’t think that I need anything else.
Wow! My lack of patience (and research) just cost me 60 CAD. Aw well…
I think FW could modify their design to make the mainboard unbrickable. I.e. recover from things like this without needing special tools.
For example, I have a desktop mainboard that can flash the bios even with the cpu and ram missing.
Yes. Ideally they make it easier to flash their boards for instance. It’s something they could easily do. As well as provide the raw bios files.
Trying to replicate what moparisthebest did. Bought the same tools, set the programmer the same way as the picture.
Mainboard is not in a case. It is not connected to any power, but I kept the coin cell battery in. Got red lights for battery and audio module. I have a screen connector. Post codes are BGGGGGGG. My board is the i7-1360p.
I even compile a recent version flashrom in Ubuntu 25.04. I just can’t read the BIOS. In my 50+ attempts, I got something other that EEPROM/BIOS not found about three times, but it seemed like error message. Tried with and without the 1.8v adapter. Even tried to get some reading in Windows with ASProgrammer.
Looking for other suggestion for troubleshooting… or if it’s completely possible that the BIOS can’t be read from the chip for some reasons.
I don’t know if it’s sad or hilarious, but
a. the lack of input from anyone in FW here
b. the fact that something that has even a 0.01% chance of bricking someone’s board is allowed out of the door
is frankly disgusting.
Having to buy a new board for simply applying a BIOS update issued by the manufacturer in a manufacturer instructed manner?
Wow.
Way to stand behind your own morals, @nrp .
In an ideal world, dumb business decisions like this would be front and center on the purchase page. It seems to be that only after purchase people are realising how much they’ve been cucked by this mickey mouse operation masquerading as a tech company.
“omg repair so easy!”
Give me a break.
Wow, that’s an impressive recovery! Soldering directly to the SPI flash and using a Raspberry Pi is definitely advanced, but it’s great to hear it saved you from buying a new motherboard. Thanks for sharing the detailed steps — it could really help others in the community facing similar BIOS update issues.
I sent my board to someone to investigate it, and it seems like the CPU is shorted. CPU was getting something like 0.003VDC instead of 0.860VDC.
I almost hate to do this, but I am going to chime in on this thread. If you are going to use the cheap CH341 programmers, you should first connect a DMM or scope and verify that with the voltage divider board, it is actually programming at 1.8V (or what ever voltage it should be). These had a design flaw and the main board would not program at 3.3V (1.8V with the divider), so you were not only programming at an out of spec voltage, but also back feeding other components with a higher than rated voltage. There is a fix for it, and some of them had the fix, but most do not (search for CH341 voltage fix).
For quick and dirty programming in-situ, I use a buspirate that has a programmable voltage regulator and current limit (I think someone in the thread mentioned a Glasgow which is also good). If I really want to make sure I protect the board, I will unsolder and then flash and resolder, if I can (as was mentioned in another post). I understand this is not something that everyone can or is willing to do.
Why would you hate to do this? You have provided very good information to us amateurs. Thank you.
My input is that the AMD boards (at least the 7040 series) suffer from bugs that’s resulting in a lot of bricked or partially bricked boards. When my fw13 was under warranty, a bad firmware update through the windows updater never completed and left my board bricked, which led framework to just send me a replacement. That worked out fine for me at the time, but i still had to replace the board, return the old one, etc.
Fast forward to after my warranty expired, and after updating to more recent firmware, my system became unstable and showed the random crash/freezing/restart issues that others have had in other threads.
From reading this thread, I purchased this cable WSON8/QFN8/DFN8(8x6) Pitch 8.5 and CH341A Programmer adapter + SOP8 clip+ SOIC8 adapter and manually programmed my SPI flash chip successfully. However unfortunately, it did not solve my issue specific issue. Something is still causing random freezing with under both linux and windows and saving bios settings changes to NVRAM also is not working. I’m giving up on recovering this board and purchasing a core ultra board so I can use this machine again.
I’ll echo that it’s frustrating that fw itself is not acknowledging that their amd systems have these issues that cause bad firmware updates in the first place, and there is no recourse to those like me who have 18 month old mainboards that are borderline useless now because of an apparently unknown problem that’s due to a design flaw in either hardware or firmware.
The final thing is that before I found this thread, I was in contact with the email support that went through many steps to try to confirm what the problem was, and whether or not it was the fault of the RAM or SSD. Basically we confirmed that it’s the mainboard and there is absolutely no kind of discount that can get me a replacement for less than the retail price :(.
I just created an account to thank you @cesfahani and other contributors to this post.
Thanks to all the details mentioned here I was able to recover a failed BIOS upgrade on a standalone setup.
Lights pattern was 00000000000010101000, which I transalted to hex A8, and therefor
#define SMM_ACPI_DISABLE_START 0xA8 // OS call ACPI disable function
I used a T48 programmer with a WSON8 (spacing 8.5) adapter to first write the 3.0.7 bios shared by @cesfahani (thanks again!) and then upgraded following the official docs to 3.1.8 using the EFI method (first left PD and then right PD)
When using the T48, it only succeeded after erasing manually and checking “Write blanks”
I’ve been wanting a board to tinker with so I ordered one of the mainboard mystery boxes for $49. Based on the fairly limited research I’ve done, I think I’m facing this same issue so I’m going to buy the tools and try to reflash the chip myself. I’m experienced in IT and minor hardware repair but this is something I’ve never tried before and I’m excited to do. Anyone got any tips or advice for a newbie?
(Before anyone tells me to cut my losses or whatever, I’m doing this purely because it sounds like a fun way to learn more about this. I’ve already learned enough new things that even if I can’t get the board working, it’ll still have been worth it.)
Welcome to the community @jacobias.
It will be a fun task to try and manually reflash a board or do some board level investigation as to what it might take to repair something from a mystery box.
The best piece of advice I suggest is being patient and careful when tapping into the little pins for the flash chip. The wrong move might just cook a chip by accident.
The miniaturization of components makes it increasingly harder to be connected where desired vs. close and maybe touching an extra pin or two!
Keep in mind that some boards genuinely have a hardware fault while others may just need a clean BIOS flash to all the subcomponents to get them talking correctly again.
The Embedded Controller console has quite a bit of feedback that can be useful when diagnosing too.
I was able to get a CH134a and dump the bios off a Framework 13 AMD HX 370, but I am having trouble parsing the .CAP file/merging it with the BIOS dump. Would anyone be able to help or provide me with a working BIOS for a HX 370.
I actually own the same model of mainboard and am going through a similar process as I write this. I’ve had zero issues with write protection. My mainboard was bricked when updating the bios with an eGPU connected and I am attempting the repair. My current setup is using the Raspberry Pi 4 going straight into the WSON 8 probe. (the BIOS chip’s voltage aligns with the 3.3v output of the Raspberry Pi) I have had a little issue with the chip being recognized by flashrom as it identified two possible chips when trying to interact with it. However, it was easily mitigated by specifying the correct model after a little research. The only other issue I’ve easily resolved was it not recognizing anything. This was easily fixed by looking closely at the probe as usually it was a pogo-pin that was stuck on the edge of the chip’s package. If anyone is looking for the flashrom command I used I have it listed here:
flashrom -p linux_spi:dev=/dev/spidev0.0,spispeed=512 -c W25Q256JV_Q -r oldBadBIOS --progress
The reason I don’t list the write command is because I haven’t been able to write to it quite yet due to some issues. I’ve been struggling with extracting the bios from the .cap file. I’m really just struggling on finding the offsets using UEFIToolfor the head and tail commands. If there is anyone that could help, it would be greatly appreciated as the Intel bios is different in a few ways from the AMD counterpart which has been making it difficult.
I have been using this fork of InsydeH2O-extractor to extract flashable images from the capsule files. It’s worked on prior models of Framework Laptop.
thank you so much for this tool, I was able to get my Mystery Box Mainboard up and running after I was able extract and write the bios over 
Will this method work with an Intel 12th that will not boot up all the way? I am currently getting a blank screen with the ON button light staying on. The POST code is all green except for one blue light right after the orange LED lights up. Any help will be good. I will probably make a new post if this doesn’t get noticed. The motherboard is from the mystery box.
