This is what’s on the chip: AJ1936 25Q256DYIG U8N282
The only revelant information I could find are this and this.
Well, I haven’t looked for them, since I thought I would use BIOS1 + finding them would involve completely removing the heatsinks – which I will do if necessary / sooner or later as I want to take some PCB pictures. I can do it at the earliest on the 19th/20th if you need it.
I would guess that all BIOS chips are the same. But if this clip is no good, could you link another one that should definitely do the job? (preferably from Amazon)
No, I just wanted to update you - since you’re here from the beginning, hopefully you will be here till I can solve the issue. Assuming, that I can fix it
I found another chip on the back of the board: Winbond25Q16JVSIQ
There is not much information about it on the internet, I only found datasheet of W25Q16JV, in which is explained that W25Q16JVSSIQ should correspond to 25Q16JVSIQ. 1
I tried to read this chip with the SPI programmer but didn’t succeed, though it should be possible since it’s a SOIC-8 208-mil configuration.
Any ideas?
No read at all.
Here is the command I used: sudo flashrom --programmer ch341a_spi -r backup.bin
And the output:
flashrom v1.2 on Linux 5.15.0-76-generic (x86_64)
flashrom is free software, get the source code at https://flashrom.org
Using clock_gettime for delay loops (clk_id: 1, resolution: 1ns).
No EEPROM/flash device found.
Note: flashrom can never write if the flash chip isn’t found automatically.
I’ve another board with which I could test the CH341A, and succeeded by reading & writing.
Yeah, I know - but as I mentioned I couldn’t attach the clip properly to this chip (and I was trying for about 30 minutes). But shouldn’t matter since it isn’t a BIOS chip. I’ll try to find the other 2 BIOS chips and get a replacement for the SOIC-clip
UPDATE:
I’ve found the other 2 BIOS chips, but they are the same package type as BIOS1 (as expected), so nothing interesting. I verified that FLASHBACK does in fact work - but that still doesn’t mean that it “actually” works
While experimenting with this, the flashes appeared to be successful. However, upon closer inspection, it became evident that the “damaged part” I had just removed was no longer present in the BIOS after a BIOS dump, despite being there before the flash. Therefore, the flashes were not truly successful, despite their initial appearance, much like the current situation – so there is still hope that it’s just the BIOS chip!
If FLASHBACK indeed works, it could provide an explanation for the damage to the other 2 BIOS chips, despite not having flashed the ‘incorrect’ BIOS onto them.
My theory is that the NVRAM chip has been damaged (since the modifications were in the NVRAM). Would it be possible to find the NVRAM chip and erase its contents/fix it somehow?
How did you find out if you can’t dump these chips?
Most probably there is no NVRAM chip, NVRAM is part of bios region in firmware, meaning that it will reside on these 3 ‘bios’ chips. This would be consistent with EVGAs update routines and that one had to make all settings again after an update.
I flashed a BIOS full of FFs and the board did not respond to anything, hence the flash was successful - FLASHBACK works.
Still, there is this possibility - though unlikely, because the last time only the given BIOS chip (e.g. BIOS2) was affected and not the others, in contrast to the current situation.
Older boards definitely had a seperate NVRAM chip, though yes, it’s possible that it’s different on modern boards (I couldn’t find reliable information on this) - but then I don’t understand how the other BIOS chips were affected from the flash.
The problem is definitely with the NVRAM region as the modification was done here, and the error codes are also right after NVRAM initialization.
Flashback does “something”. To say it works you should know how it’s supposed to work and have the ability to check if it does what it’s meant to do. Most boards don’t touch the NVRAM when updating. EVGA has an update routine
Their normal update would overwrite the NVRAM, look into the structure with UEFIToolNE
Every board I had, after a BIOS flash (built-in) all settings, profiles, etc. were gone, hence it did touch the NVRAM
But I don’t see how that’s relevant: that’s with FPT, FLASHBACK or the built-in flashing process could work completely different
Yeah, I know how the structure looks like…
But I still don’t understand how the other chips got affected, even though I flashed the BIOS only on 1 chip - so that’s why I thought there is a seperate chip for the NVRAM.
The problem is that some screws have been glued from factory, and as of right now I can’t unscrew them – tried heating them up, but didn’t help.
Well, I’ve ordered one of this - hopefully it will work
What else could have been damaged in your opinion?
And do you have any tips how to loosen the glued screws so I can remove the heatsinks?
