If you have the file containing the bios and an eprom programmer or access to an eprom programmer and an eprom that is compatible with the boards on which you want to use it, then it is very likely that you will be able to program the file to the eprom and insert the eprom into the mainboard on which you want to use it and thus, hopefully be able to use the mainboard.

joebelter@hotmail.co wrote on 2025-02-09, 08:23:

The first board is an XT that's very clean, shows power on the Analyzer card, and I can't find a bad solder joint or cap anywhere.
The board shows power, no beeps, tried different memory configurations (pulled out one bank at a time and tested it).

I have that board. The BASIC BIOS is entirely optional and is missing on many XT clone boards. While the IBM BIOS is quite simple in functionality, so it could be legally clean-room cloned by third-party vendors, AFAIK no one over made a IBM/Microsoft BASIC clone for the XT. Clone machines sold into regions where Microsoft and IBM could enforce their copyright on the BASIC ROM usually came without the BASIC ROM. My variant of the board has a Phoenix XT BIOS instead of the DTK/ERSO one, though.

Are you aware that XT boards require configuration of the video card type using DIP switches? Your board is currently switched/jumpered to CGA support. If you want to use a video card with its own BIOS extension (like EGA or VGA), the fifth (counting from the keyboard connector) red jumper/switch in the block next to the keyboard connector needs to be set the other way (towards the ISA slots instead of towards the power supply connector).

I have the ROM files for the 486VLB but need to know what chip to get...
Any idea where or what I could order?
Thanks!!
Joe

joebelter@hotmail.co wrote on 2025-02-09, 19:35:

I have the ROM files for the 486VLB but need to know what chip to get...

Image 2 on the Retro Web site for your 486 mainboard shows an flash chip installed in the board. You clearly see the Intel "i" next to the sticker, so the chip currently installed is an Intel flash chip. The form factor of your socket is PLCC32. The BIOS Image I found on the Retro Web is 64KB, which is 512kBit. This narrows down the range to mainly one single chip, the Intel 28F512.

The Intel 28F512 uses the "JEDEC standard 32-pin pinout" in the PLCC package. Be aware that there also is a different "28-pin pinout" that distributes 4 "not connected" pins between the 28 actually used pins, and uses the same 32-pin PLCC package. An example of a "28-pin" 512kBit chip is the Winbond W27C512 (yeah, Winbond uses the "27" code that is typically used for UV-erasable EPROMs for flash chips, probably to indicate pin compatiblility). If the Intel chip on that TRW picture actually belongs to the board, it will not be compatible with the Winbond W27C512 chip (or any other "32-pin" PLCC flash chip that has pins 1, 12, 17 and 26 marked as "NC").

If you don't care about in-system programming and set the jumper to "flash write disable", likely any PLCC32 chips from the 28xxx or 29xxx series with 512kBit or 1MBit (program the same image twice to an 1MBit chip) should work, as long as the chip can operate at 5V (chips with a "V" or "LV" in the model name are often specified for 3.3V only).

I checked ebay, and it seems 512kBit chips are rare. I personally would take the risk to order a 1MBit chip like the AT29C010A or AM29F010 or MX29F001 and try to use that chip, and only bite the bullet to source a (nowadays) rare and expensive 512kBit flash that is not a 27E512 / 27SF512 / W27F512 if the 1MBit chip fails to work.

Looks like I could order :
MX29F040QC-12 Flash, 512KX8, 120ns, PLCC-32
or
AM29F010-70JC AMD Flash, 128KX8, 70ns, PLCC32

AT29C010A-12JC ATMEL Flash, 128KX8, 120ns, PLCC32

I assume the AM chip then?
Thanks!!!

joebelter@hotmail.co wrote on 2025-02-09, 21:22:

Looks like I could order : MX29F040QC-12 Flash, 512KX8, 120ns, PLCC-32 or AM29F010-70JC AMD Flash, 128KX8, 70ns, PLCC32 […]
Show full quote

Looks like I could order :
MX29F040QC-12 Flash, 512KX8, 120ns, PLCC-32
or
AM29F010-70JC AMD Flash, 128KX8, 70ns, PLCC32

AT29C010A-12JC ATMEL Flash, 128KX8, 120ns, PLCC32

I assume the AM chip then?
Thanks!!!

The AMD AM29F010 will be fine. The MX29F040 is a 4MBit (512 Kilobyte) chip, which might be incompatible with that board. The 1MBit chips most likely will work fine. But make sure to set the jumper to read only, because otherwise 12V might be applied to the programming voltage pin, which is fine for the Intel 28F512, but a terrible idea for the 5V 29-series 5V-only flash chips.

Super! Just ordered it. Supposed to be here in a week or so. I'd sure appreciate it if you could check back and see if I'm not having difficulties!
😀
Thanks so much,
Joe

Well I got the AMD29F010 chips in. I used my TL866 programmer to write the bios file from the Retro Web to the chip. I've never done ANY writing twice to make files bigger but I used a program called BINWIZARD to merge two of the bios files together to make one larger 128k file. (https://github.com/r1me/BINWizard)
When I start the computer the post card stops at A1 and the keyboard num, caps, and scroll lights stay on and I have no signal and NO beeps. Power supply is known good.
The bios seems to read ok as far as the Phoenix words and stuff, but I have little to no knowledge of how to write/combine bios files other than what I did or load and write a singe file.
The board looks EXTREMELY clean, no damage I can see from corrosion, or damage. Looks new. Just doesn't do much.
Any help would be so appreciated!
Joe

Before I get into reverse engineering all of the BIOS files, as I can't find anything official about post code A1 on Phoenix BIOSes, please tell me which of the three image files from the Retro Web you used.

I flashed the earlier one - G22-2LB.bin first...
The other two wanted things like working floppies and such... and can't get to that point... and I figured they'd be trying to flash the chip with 12v instead of 5v...

joebelter@hotmail.co wrote on 2025-03-05, 21:56:

The other two wanted things like working floppies and such...

Not really: You can use the G26-2LB.BIN or the GLB05.BIN from the other archives just like you used the G22-2LB.BIN. I will take a peek at the G22-2LB.BIN BIOS.

Did I do the right thing in combining the files that way? I sure do appreciate you taking the time to help me with this, I've been knocking myself out and getting nowhere. I'm so inexperienced with electrical troubleshooting it's silly, so I don't even know where to start there. I have the equipment - good multmeter, decent starter ocilloscope, etc... but I'm learning so slow I'll be dead before I get good at troubleshooting!

POST code A1 is, as it seems, a custom POST code by this BIOS that will be sent directly before RAM initialization. According to the Microhouse documentation, bank 0, the slot next to the CPU needs to be populated first, and only 4MB and 16MB SIMMs with parity are supported. Are you sure you have compatible memory installed? Do you have a speaker connected? If not, do so, and listen for beep codes.

joebelter@hotmail.co wrote on 2025-03-05, 22:10:

Did I do the right thing in combining the files that way?

It seems so. I don't see how the computer should even get to the location where it outputs POST code A1 if you would have programmed the chip incorrectly.

I would have combined the file using

1C:\BIOS>COPY /b G22-2LB.BIN + G22-2LB.BIN COMBO.BIN

but at the moment, I don't consider it likely that the combination is the issue. If the TL866 software lets you browse the data, you can verify that the region at address 10000(hex) looks identical the region at address 00000 to confirm correct combination.

Right, I tried it with no bios once and got no codes. Not sure why I'm not hearing any beeping, I usually do if there's a problem.
That combine command - just run it from command prompt in Windows 10?
I can check the address to make sure. Not good at hex stuff but I'll try.
Thanks!

At 10000 hex the exact same info is started again. In this case it's a copyright message that doesn't seem to start with what should be the real beginning, but it's identical at 0 and 10000.
Here's a pic.

joebelter@hotmail.co wrote on 2025-03-05, 22:47:

At 10000 hex the exact same info is started again. In this case it's a copyright message that doesn't seem to start with what should be the real beginning, but it's identical at 0 and 10000.
Here's a pic.

That's correct.

joebelter@hotmail.co wrote on 2025-03-05, 22:30:

Not sure why I'm not hearing any beeping, I usually do if there's a problem.

OK, so obviously a speaker is connected. I think we should focus on verifying that a compatible memory module is correctly installed. Can you take a photo of the board with the installed module? Preferrably one on which the type of the memory chips is readable?

I'm using 32mb in bank 0 that I used in a working 386sx board, but I can try others. I have lots of them.

OMG! Progress! Finally!
I put two known good 4mb sticks in and set the switches to 8mb and got a POST SCREEN!
One time, the speaker let out two long buzzing sounds (I think 2) and the screen said that there was setup problem and hit F1 to continue and F2 to go into setup AND IT WORKED!
I got into the Phoenix Bios, set the date and time, set the floppy to 1.44 and couldn't figure out how to save it. When I hit ESC it rebooted and now I don't get the F2 message anymore and it does keep trying to seek the floppy.
Then it just sits at the screen waiting for me to get into the setup screen, but F2 doesn't do anything. I've tried to take the battery off the CMOS terminals but it still has retained the settings I guess.
Should I go ahead and just makeup a new bios with the newest one you suggested?
Here's a screen shot of what I see.
WOW! After nothing for weeks, it's SOMETHING!!!
Joe

Oh, and I don't see a clear CMOS setting...
Here's the jumper sheet...