I do have a working TH7II system. If it is of any help, it also appears to be running revision EH. I have made a backup copy, as well as the bootblock with UniFlash; both are attached. To start with, you can compare that with the files you downloaded, as well as the dump of your old chip (assuming same revision). The file sizes may differ slightly, since your programmer may dump the bootblock as part of the same file. Have a look through that, and feel free to let me know if I can help with anything else.

Thanks! Downloaded it and I'll probably test it tomorrow.

There is quite a few of us, myself included that downloaded the entire FTP database of abit before they went down the drain. I have eight regular bioses for the th7ii and 3 beta ones. I dont know which ones are which, but if you want any or all of them, LMK.

Did the TH7II and TH7II-RAID use the same BIOS? Presumably the BIOS was actually for the TH7II-RAID and the previous model just ignored the RAID stuff.

no, there are separate folders for each, with different versions and sizes, so I'd say no.

I was a little suspicious and just checked the old system. I had forgotten that it is actually a TH7II-RAID, so you may be better off getting the file you need from luckybob.

DING!

All files I have for TH7-II NON raid.

Ok thanks, looks like these .exe only runs on older Windows. I think they unzip the bin.

7Zip can sometimes decompress .EXEs.

I had a quick look at the files, and can confirm that 7-Zip is able to extract the .bin file from the old (probably 16-bit) self-extracting archive. In comparing the file provided by luckybob to the EH revision on my TH7-II RAID, both are 512 KiB files, but the version for the non-RAID board has more unused space, as well as some other differences throughout the file. I didn't make any attempt to determine the nature of the differences, however.

If ABIT boards are anything like my old Asus boards were the BIOS chips are the same size as are files between the raid and non raid. With the Asus you could flash the non raid bios to the raid board and it would work fine but you lose the raid functionality. The reverse would work in that the board would post but often hang trying to detect a non exsistent chip.

Used to do this back in the socket Athlon days since some bioses had various performance differences or different unlocked options.

Okay, I might need some basic BIOS lessons here. First BIOS is the one @640K!enough posted. I assume the beginning is the boot block where it identifies BIOS/board type. Second BIOS is the one @luckybob posted. It seems to lack them. Both BIOS are exactly 512KB.

Is there a special way to flash the BIOS @luckybob posted? Writing the file to the BIOS chip will result in no POST, because I'm assuming it's missing that boot block? There is an option on the BIOS programmer to use a start address, but not quite where the boot block starts/ends (aside from the fact @640K!enough posted a 8KB boot block).

special? i don't think so. usually its a standard procedure. boot with a dos floppy, run the flash program, ?????, profit.

luckybob wrote:

special? i don't think so. usually its a standard procedure. boot with a dos floppy, run the flash program, ?????, profit.

I can't. My BIOS was corrupted for whatever reason. When I boot to floppy, run flash program, it will complain about the lockout jumper not being set. I searched the archives, and most people just got a new BIOS chip at that point.

http://www.overclockers.com/forums/showthread … 16-bios-install

One correction from what I posted previously: the bootblock is part of the 512 KiB backup file I posted earlier, as well as the file posted by luckybob. In this case, the bootblock resides at the very top 8 KiB of the chip, starting at address 7E000H (516096 decimal). If you look at the contents of your chip in a hex editor or via your programmer's software, you should see exactly the same contents (at that address) as the bootblock file I posted earlier.

If it were my system, the first thing I would do is ensure that I have a complete dump of the chip, saved to file via UniFlash or using your programmer. Then, re-install the chip, and clear the CMOS via the jumper (with power disconnected). Plug in and test. If the system boots at all, try starting with a DOS floppy and programming using AWDFLASH and the following command line:

1AWDFLASH <bin_file_name> /PY /SN /CC /R /CP /CD /QI

If that fails, try UniFlash. If none of that works, try programming the file I posted using your programmer and see if it will boot at all, or if you can then do a bootblock recovery to program luckybob's EH file.

EDIT: I'm fairly certain that if luckybob posted the revision EH file for the TH7-II RAID, it would not match the backup I posted. AWDFLASH doesn't erase and write the entire chip by default (for instance, the bootblock is generally not re-written without the /WB switch, I think). Some sectors are likely not contained in the file, but are left (partially) intact.

Ok that makes sense.

What happened was this. My TH7II was working fine, it had a 38 bios. I had a Northwood M0 stepping (Gallatin) CPU in it, so technically it wasn't supported... although it worked fine. I tried to update to EH bios. During the update procedure, it stalled and crashed. However, it would boot up to DOS (boot block was good), but using awdflash to flash BIOS would just result in 'lockout jumper' errors.

I saved the boot-only BIOS.

dexvx wrote:

However, it would boot up to DOS (boot block was good), but using awdflash to flash BIOS would just result in 'lockout jumper' errors.

The /QI switch has sometimes been reported to help with that problem. You may also need to add /F. If either of those aren't accepted, it may be worth making sure you have the latest version of AWDFLASH you can find.

On a related note, I have repeatedly noticed that AWDFLASH is rather fragile. When creating a backup with it, it simply aborted (no return to DOS, no error message; just stalled) when it encountered a floppy write error.

It is also now clear that AWDFLASH and other tools, such as UniFlash, don't work exactly the same way. I made backup copies of a system with each. The AWDFLASH-generated file looked rather like luckybob's file, while the UniFlash backup of the same system looked more like the file I uploaded. To me, this indicates that my previous suggestions are worth a shot.

Now this has me curious, and I'll be sure to pick up an extra flash chip or two next time I order parts to experiment with this.

thought:

If you had the original bios, you could "match up" the updated part in the correct spot and make a larger bios that will be complete. I can't image it being that difficult.

I've flashed the bios many times with many boards, and never had any issues.

Ok so here's my update. Work has been picking up lately (towards the end of our development cycle), so haven't had much time for fun.

1. Checked the boot block section of my original TH7 bios and @640K!enough. Aside from versioning, they look the same. Also checked the BIOS itself. The TH7 and TH7R likely share the identical BIOS. It probably does a HW init check to see if the RAID controller is there and is auto disabled if not detected. I've flashed @640K!enough's entire TH7R_EH bios onto my system and there is no indication of the RAID at all.

2. The first 8KB (0x0 to 0x1FFF) is the boot block. The award flasher will obviously skip this because even in the case a BIOS update goes bad (like my issue), you could still boot to floppy. This is meant so that you could still flash the BIOS. However, me getting a lockout jumper error just seems to be bad luck.

3. So if you want to flash the BIOS from the BIOS files as @luckybob posted, you basically need to copy the boot block address on top of the BIOS file. If you just write the BIOS file to chip, there is no boot block so board won't do anything. You will be one sad puppy without a boot block. So as always, Backup!

Anyways, I recommend this BIOS programmer (TL866A). My TH7II board would've been a complete waste otherwise. If anyone has a bad BIOS, I can help out for free if you mail me your chip with a prepaid return envelope... and as long as I don't have to go BIOS hunting. Hope that isn't against the rules. 😐