I tried but cannot find anything but dead links. The old Mitac website is archived at Archive.org but the page you need was not crawled.
this page has the jumpers according to www.peteweb.com.
http://www.mitacservice.com/visitor/fcabdl.asp?fid=23860557
https://web.archive.org/web/20021201140358/ht … isitor/main.asp
https://web.archive.org/web/20010411002358/ht … r/v_welcome.asp

Any news about this? Got the same board without manual.

There's a partial manual here https://www.ultimateretro.net/en/motherboards/9583

perhenden wrote on 2022-03-16, 19:53:

There's a partial manual here https://www.ultimateretro.net/en/motherboards/9583

Hi perhenden,

Do you happen to have the Mitac IH4077H BIOS?
The copy at https://www.ultimateretro.net/en/motherboards/9583 is unusable. It contains only the upper 64KB half from Shadow RAM, but the complete ROM dump should be a 128KB BIOS. 😉

Thanks, Jan

Yes, I'd be happy to help. I've got access to this motherboard. There is something to work out - romsavat.exe produces a 64KB dump, same as on UR. Another dump bios tool, dumpat, produces an 1MB file. Perhaps you're able to cut that file and extract the 128KB rom? See attachment. The first 64KB of the bios seems to read as empty.
Unfortunately, the bios chip on my board, a Micronix MX28F1000PC-15, when put in a TL866II-plus cannot be read. The Mitac board can read it, though.

Thanks for the DUMPAT output. A small improvement because the F-Segment of the BIOS seems to be a real copy from ROM and not from Shadow RAM anymore. Amazing that this old tool knows how to (temporarily) disable Shadow RAM on the UMC498 chipset!
But the DumpAT tool is obviously made for getting 286/386 and early 486 BIOSes and only produces a 64KB BIOS. The rest of the file is filled with FFh, so still no valid E-Segment.

Luckily we have DOS’s Debug tool and can do this the old fashioned way. 😉
I’ve made a Debug script named GETBIN.DEB, attached below, that copies both the F- and E-segment from memory (ROM or Shadow RAM). At the DOS prompt, run the command:

DEBUG < GETBIN.DEB

This produces 2 files, SEG-E.BIN and SEG-F.BIN that are 64KB each. With these files I should be able to construct the complete 128KB BIOS image.
Please send both files with your reply.

Jan

That went well. Here are the files attached.

Amazing, this is one coy BIOS, it keeps hiding itself!

The Debug copying went well indeed and I can see some BIOS code in the segment-E part now. However, only the top 8 KB of this segment is visible and the rest is still hidden. 🙁
The visible part contains the Interrupt 10h handler, the BIOS’s Video Display Services, which obviously should remain visible during run-time.

In the F-segment I see the call-table for all POST routines, which points to addresses in the E-segment. This means the hidden part of segment-E in this BIOS contains all POST routines. Essential code for a working BIOS copy, but only needed during POST so this BIOS hides this part at run-time…

I have an idea how to disable the shadow RAM of the E-segment (not possible via the BIOS Setup unfortunately) so the ROM code becomes visible. But I need some time to work that out.

Jan

Hi perhenden,

Okay, this is what I like you to try.

In the attachment you’ll find the DOS-based CTCHIP34 chipset configuration tool with a .CFG file for the UMC498 chipset.
As CTCHIP34 originally doesn’t have support for the UMC498 chipset and the databook is nowhere to be found, I wrote this UMC498GP.CFG file by analyzing another UMC498 BIOS. This particular UMC498 BIOS is 64KB and therefore only occupies the F-segment. So it allows option ROMs in the E-segment and has BIOS Setup options for Enabling/Disabling Shadow RAM in this E0000h-EFFFFh memory region.
By following the BIOS code, I found the UMC498 chipset registers and bit-definitions that control Shadow RAM. For the E-segment, this appears to be reg 35h bits 7-2.

At the DOS-prompt on your IH4077H system, type the command:

CTCHIP34 UMC498GP /35

and you should see the actual reg 35h setting as programmed by the BIOS.
The bits are probably 010111xx to indicate an enabled and write protected Shadow RAM E-region. By entering the command:

CTCHIP34 UMC498GP /35:=%000000xx

You will disable Shadow RAM of the E0000h-EFFFFh memory region, allowing the BIOS ROM to be read directly from the EPROM.

Now run my Debug script again to copy the BIOS to the 2 SEG files.

Member feipoa was so kind to test the UMC498GP.CFG file on his PCChips M912 UMC498 board and this worked fine. So I’m confident this will work on your UMC board as well. 😀

Jan

Thanks! I tried it, but the resulting seg-e.bin unfortunately has empty (nothing but FFH) contents. I made sure to delete the old .bin files first.

I tried the variations
CTCHIP34 UMC498GP /35:=%000000
CTCHIP34 UMC498GP /35:=%00000000
CTCHIP34 UMC498GP /35:=%00000000xx
with no difference in the result

I gave the bios chip a third run in the TL866II programmer. By disabling BOTH pin detect and check id in the Xgpro software, I was able to read the chip (id reads as 0xC211, but should probably have been 0xC21A according to datasheet). I don't trust this chip, and I'm happy about replacing it.
Modbin could read it (checksum OK), and when the rom is written to a new chip, the motherboard posts and boots to DOS ok.
I don't know if the read is 100% good (depends on how the checksum is made/verified).

Great that you were able to read the EPROM with your TL866II afterall. 😀
Yes, I agree that this chip was on its way out.

I've checked this new BIOS dump as well and it looks perfect. All code is where it should be and the various checksums are correct.
You can send this BIOS to ultimateretro, so they can replace the unusable one.

Thanks for trying my UMC498GP.CFG file. Good to see it works on this board, including changing that chipset register.
Too bad it didn't have the expected effect. 😉

Cheers, Jan

Thanks, new BIOS published at UltimateRetro.net