[Help Needed] Bricked Acer Predator Helios PH-315-53-76Z1 BIOS

Hi. My Acer computer failed during bios update and is now stuck. The computer turns on but only keyboard lids and fans turn on, no screen at all.
What are my options here? Is it possible to flash new bios on with ch341a?

So far I’ve found 2 possible bios chips? One under the hard drive and one near the cooling which requires to remove the cooling to flash.

XMC QH128AHIG under the ssd disk.
Winbond 25Q80EWSIG near the cooling.

Or am I searching for another chip?

Ive been doing a ton of reading of the procedure of flashing but any input regarding this would be awesome and especially if anyone done it to this laptop. all help appreciated.

Best regards, Tobias.

thumbnail_image0.jpg



thumbnail_image1.jpg



Edit by Fernando: Thread title shortened

Go for that one.

Do a dump/ backup of the original content, check it with UEFItoolNE for structure/ errors in parser.

Stock firmware doesn’t contain machine specific data (SN, UUID, MAC…) so these are lost if you don’t recover them from the backup/dump of the bricked bios.

Acer bios files are valid images, but contain padding in the beginning and the end.

Which versions for the original bios and which version were you trying updating to?

Oh ok. Been getting different opinions where people think its the winbond one, but whatever. Im not sure which version it was going to update since it was through windows update… should i do dump/backup through asprogrammer or any other program? Any way to confirm while i read that chip to verify im at the right place?

Im a new at this kinda task but not new to computers but im up for a challenge. Hit me with kinder garden info and help if you must ;D

Well, always good idea to check the prerequisites: Read the datasheets and compare to firmware size => Firmware is 16MByte/ 128Mbit, one chip is 128 Mbit, one chip is 8 Mbit… And check the voltages, too! (8Mbit chip is 1.8 V, don’t touch without 1.8V adapter- might be video firmware or EC firmware)

Unclear how much you’re used to using a SOIC8 clip, in any case: Make 2 or better 3 dumps of the exisiting chip, compare via "fc /b file1 file2" or in HxD or… . If all three are a 100% identical it’s highly probable that you have good contact to the chip, your program is at least working for reading, and the dumps are valid backups of the chip content.

Some chips are difficult to read and / or write, some hints in the forum regarding Asprogrammer and Neoprgrammer:
[REQUEST] Acer Nitro 5 (AN515-54) (4)
How to extract SN, UUID, etc from BIOS dump

Post the dumped chip either in the forum or via external link.

Ok. I have no experience in using that clip so will probably take a few tries to get it to read, is it ok to just unplug the battery or do i need to disconnect cmos battery too? If that exists on that board bcs havent found any. Ill give a go at the xmc one first, should i be looking at something specific in the dumped files if they are broken?

Ok Here the first dump comes. Did the 3.3v mod to the ch341a and it read fine.

https://easyupload.io/967xty

Can anyone help to proceed?

Ill put a link to the bios id like to flash:

https://easyupload.io/l8rgoj

Ok Here the first dump comes. Did the 3.3v mod to the ch341a and it read fine.

https://easyupload.io/967xty


Ill put a link to the bios id like to flash:

https://easyupload.io/l8rgoj


Can you help to proceed? =)

Take FH53Mx64.fd from update file, cut padding from 1131E50, and thereafter cut padding from start to 131E4F, this should give you stock update bios. File shoud be exactly 16.777.216 bytes and have exact same structure as your backup in UEFIToolNE

- Use MEA on your backup and the bios update- ME didn’t get updated.

- Compare both files in UEFIToolNE- both files have same errors in parser, actually update file has one more error because of empty NVRAM (expected)

- Compare both files in HxD (or another Hex Editor), FD identical, ME differs as expected (different version, configured vs. initialized), check bios region from end, first differences in NVRAM, FTW store and first padding in bios region (machine specific data)

Non updated ME wouldn’t normally brick, both MEs can be read in FIT, settings almost identical but ClkoutITPXDP value, was enabled, changed to disabled in later version. Bad NVRAM settings can brick, for example when trying to undervolt or in case unfinished transaction. There’s data in FTW store, which is used to support transactions in NVRAM, and is meant to restore NVRAM integrity in case unfinished transactions. That might point in this direction.

I would
- take ME region from update files, ‘extract as is’ in UEFIToolNE, ‘replace as is’ ME region in copy of your backup in UEFItool025 with extracted newer ME region from update file
- replace NVRAM and FTW store in bios region in the file you created from backup, meaning exchanging from 300.000 to 32F.FFF in the file with exchanged ME region with same area from stock update bios.

Flash the repaired file into the chip, read the content back from the chip in a different process and save the new content in a different file. Compare original file to newly saved file. If both files are identical the flash was good.

Good luck!

[Edit] Post your repaired bios/ a link to it if you’re want a confirmation before flashing…

Got some help with the new bios. Just flashed them and computer up and running again. Thank you so much for the help!

If someone is searching- seems to be there:
https://bios-fix.com/index.php?threads/a…745/#post-93293

Hi,i have the same problem,but with the Video Bios at Rtx 2060 chip,i want to ask you guy’s,which bios chip is the for the VBios? That one WINDBOUND which you mentioned în the first time?Wait for an answer, thank you!

20220410_172002.jpg

20220410_171938.jpg

20220410_165817.jpg