The UEFI ofe the Surface Pro 5 broke due to an automatic firmware update. Thank you MS!
So I did some research and tried everything. In the meantime I desoldered the winbond 25Q128JVPQ UEFI chip and since I don’t have an CH341A adapter I read out the chip with the raspberry pi and flashrom.
I also downloaded the latest UEFI file from microsoft and now i have concerns because of the huge difference in the file size.
Here is the UEFI file from the chip - 16.384KB
CrXrZLM) is the latest UEFI file from microsoft - 4.355KB
Bios has a high level of redundancy, everything stativ there you’ll find twice, in your dump all redundant parts identical (which makes an update failure already very unlikely) and in addition all these volumes are identical to the stock volumes you find in the corresponding stock update file.
Identical volumes/ paddings marked in same colour (both arrows for the paddings were supposed to be purple)
The only thing that happens with the surface is that you see the Windows logo every few seconds and it keeps “restarting”. Looks like it is in a boot loop. And that happend right after the firmware update.
I’ve tried an soft reset, hard reset, USB recovery, I can’t enter the BIOS. When it was opend i disconnectet the battery and tried everything again.
So my last thought was that something went wrong with the update.
No luck. I did erase bios chip reprogrammed with test file you kindly shared and exactly same behavior as before. I press power, windows logo appears for 3 seconds and it shuts down. What does that mean, hardware problem? Anything else you can advise? Thanks
Yes it could be an HW problem…why not…such a surprise, doesn’t this happen anymore in 2023/2024???
If you have no other clues or indications of any automatic action or user action that can lead to a FW corruption, then yes it can be a hw issue on the board… in this case model everything is soldered, ram, sdd, gpu… just a tiny solder sphere from any bga chip is enough.
There is sometimes a mislead ideia from users that everything that breaks its now, due to bios or vbios firmware corruption… and there is no more hardware failures.
I found after hours of searching burned transistor not far from Bios chip on this machine. So it is hardware problem. I got another surface pro 5 which has similar problem. Logo appear for few seconds. I will upload bios of this machine shortly. Thanks so much for help.
There’s one 0x1000 Block missing in the beginning of the second padding. But this padding does contain _FIT_ and there’s an additional error message about parsing ACM:
Thanks for working your magic. It is booting now. Windows up and running. All good!! This is attempt to fix my daughters Surface pro as she uses it for drawing with pen. Thankfully she kept backup of all her drawings. Now its back to life I just need to put it back together. You made her smile and we saved anoter MS product going to bin and polluting our planet Point is Surface Pro is a crap machine, another Mobo I tried to fix, with HW problem, I cannot even find replacement transistor which was burned. Shame. Anyway, your help is much appreciated.
I wonder why MS keeps everything twice (despite NVRAM) in bios region but doesn’t use it? Anyway, that really seems to be an erased but not rewritten block in this case.
Can we speculate what caused this issue? There should be some safety protocol in place when anything gets erased and boot fails last known working copy is used. PC mobos got two bios chips for this purpose. Surface being quiet expensive piece of hardware when new should be more fault proof. I don’t get it.
Point is Surface Pro is a crap machine, another Mobo I tried to fix, with HW problem, I cannot even find replacement transistor which was burned. Shame. Anyway, your help is much appreciated.