I wouldn't worry about the message from 3c5x9cfg unless the XTIDE ROM isn't picked up by the BIOS. The ROM on your 3com card is definitely no 3com boot ROM, so the diagnostics program is kind-of correct in indicating to you that "something isn't right about the accessing a 3com boot ROM at address D0000". The indication of an "unknown BIOS extension" at D0000 by CheckIt is encouraging that the ROM actually is correctly accessible by the processor. AFAIK, you should see some indication of the XT-IDE BIOS when you boot your computer, even if no IDE drive (or CF card) is installed.

Ok. Thanks so ive never seen any xtide bios when i first boot the computer on either motherboard.
Not sure what further i should try?
I also checked the physical socket for dry solder joints and my meter gets 5V on pin 1. Maybe its something simply. One MB has award bios and the other has amibios on it. Usually at boot my 486 complains of no floppy a/boot disk failure. So it doesn't execute anything on the network cards rom bios.

a simple test to see if the bios is present would be to use dos debug program, and simply display the memory region of the rom.

debug
d d000:0

The first two bytes should be 0x55 0xAA if i'm remembering correctly.

This is the debug section. It sees the xtide but is my rom not bootable?
At the moment im just booting the pc to a dos 6.22 bootdisk and running debug etc.

It seems the image you see is not what should be visible at that address. A PC BIOS ROM extension needs to start with 55 AA. So either you wrote some intermediate file of the XTIDE BIOS generation process, or the data you see in debug isn't what's written to the EPROM, because the 3com card does some funny stuff. Please compare what you see here with the data as shown by your EPROM programmer software. If they look the same, you didn't write a proper PC BIOS extension. If they look different, we need to compensate for strange stuff the 3com card is doing with the EPROM contents.

Wow so i see the 3com card is doing some silly things!
Exactly as you described. What can i do to alter this?
So the first 3 bytes are being truncated?

Many thanks

When you say "copied and pasted the bin file again at the 16k mark", are you sure you did that properly? The first 16 bytes seem to be missing so you might have either not included them when copying or pasted to the wrong address.

I feel i selected the entire block of data selected file copy and the paste at 3fff0 i think it was. I program view it and yes there is a second code set further down.
Regardless both methods yield the same result (filling with zeros) The first few bytes aren't there but present on the rom.

vacatedboat wrote on 2022-11-27, 04:15:

I feel i selected the entire block of data selected file copy and the paste at 3fff0 i think it was. I program view it and yes there is a second code set further down.
Regardless both methods yield the same result (filling with zeros) The first few bytes aren't there but present on the rom.

What jmarsh is trying to say: It is possible that the 3com card doesn't show you the first half of the ROM chip, but the second half of the ROM chip. That's why you made the copy: To make sure the second half of the ROM chip also contains the correct data. The second half of the ROM chip starts at 4000, not at 3ff0, though. So the copy is 16 bytes before where it should start when you pasted to 3ff0, and that would explain why you don't see the first 16 bytes.

Hi, please have a look at an old posting of mine.

Re: IDE disk-on-module + 386

It has some ready to use files that were tested/working.
If you've got a 2764 or 27128 or compatible chip. A cheap one-time programable EPROM will also do, XUB is in auto-detection mode.

The only major drawback is the age of the whole thing. A few years have passed since and XTIDE has made a progress since.

If you haven't a suitable chip, you can double size by doing a binary copy.
COPY /B ROM.BIN+ROM.BIN >BIGROM.BIN

Alternatively just make a simple EPROM adapter using one or two IC sockets.
By wiring the EPROM properly, you can select the different banks of it.

That way, you can halve the capacity and you won't waste memory in the UMA through multiple copies of XUB.

Here's an example. The 27256 acts as a 27128 if the switch modification is installed.

Depending on the switch position the address line (A14, PIN27?) is put high or low.
That way, the upper or lower half (16KB) is selected from those 32KB that are available in total.

If your EPROM content is doubled as mentioned before, you don't need to worry about pull-up or pull-down.

In fact, you could combine two different versions of XTIDE Universal BIOS if you like.
So the chance of wasting an one-time programmable EPROM is smaller.

Edit: https://www.douglaskrantz.com/ElecPullUpResistor.html

Hi again everyone!
I forgot to mention something.. The XTIDE Universal BIOS does not contain a checksum by default.

So you can't just download the ZIP file, extract the *.BIN file and write it onto an EPROM..

No, you have to use the configuration program (DOS executable) first and save the settings.

Then the checksum is added by the program and XUB becomes ready to use.

I've found out the hard way, years ago.
That's why I provided my personal EPROM backups in first place.

It's meant for all those of you who use a modern PC without DOS compatibility.

Good luck! 🙂🤞

PS: I don't blame the author, though. XUB was made with the XTIDE card in mind likely, which has an EEPROM.
So it made sense that the configuration program was meant to be used originally.

mkarcher wrote on 2022-11-27, 09:42:

vacatedboat wrote on 2022-11-27, 04:15:

I feel i selected the entire block of data selected file copy and the paste at 3fff0 i think it was. I program view it and yes there is a second code set further down.
Regardless both methods yield the same result (filling with zeros) The first few bytes aren't there but present on the rom.

What jmarsh is trying to say: It is possible that the 3com card doesn't show you the first half of the ROM chip, but the second half of the ROM chip. That's why you made the copy: To make sure the second half of the ROM chip also contains the correct data. The second half of the ROM chip starts at 4000, not at 3ff0, though. So the copy is 16 bytes before where it should start when you pasted to 3ff0, and that would explain why you don't see the first 16 bytes.

That's exactly what it does with 27C512s and the like, the card is only wired for 32kB chips

maxtherabbit wrote on 2022-11-27, 19:56:

mkarcher wrote on 2022-11-27, 09:42:

What jmarsh is trying to say: It is possible that the 3com card doesn't show you the first half of the ROM chip, but the second half of the ROM chip. That's why you made the copy: To make sure the second half of the ROM chip also contains the correct data. The second half of the ROM chip starts at 4000, not at 3ff0, though. So the copy is 16 bytes before where it should start when you pasted to 3ff0, and that would explain why you don't see the first 16 bytes.

That's exactly what it does with 27C512s and the like, the card is only wired for 32kB chips

Off by one (power of two). The 3c509B config screen shows D0000-D3FFF, which is 16KB. The chip in question is a 256kBit (32K x 😎 chip, like a 27C256.

mkarcher wrote on 2022-11-27, 20:02:

maxtherabbit wrote on 2022-11-27, 19:56:

mkarcher wrote on 2022-11-27, 09:42:

What jmarsh is trying to say: It is possible that the 3com card doesn't show you the first half of the ROM chip, but the second half of the ROM chip. That's why you made the copy: To make sure the second half of the ROM chip also contains the correct data. The second half of the ROM chip starts at 4000, not at 3ff0, though. So the copy is 16 bytes before where it should start when you pasted to 3ff0, and that would explain why you don't see the first 16 bytes.

That's exactly what it does with 27C512s and the like, the card is only wired for 32kB chips

Off by one (power of two). The 3c509B config screen shows D0000-D3FFF, which is 16KB. The chip in question is a 256kBit (32K x 😎 chip, like a 27C256.

Oh, well IIRC you can configure the card to provide a 32kB window which would eliminate all this BS

maxtherabbit wrote on 2022-11-27, 20:05:

Oh, well IIRC you can configure the card to provide a 32kB window which would eliminate all this BS

Depending on the card and the config program used you can pick which 16KB window of the ROM to use - obviously requires the card to control the high ROM address pins rather than just wiring them directly to either supply or ground. That seems unlikely here since the card is apparently defaulting to the second (misaligned) window/copy of XTIDE, and besides the multiple copies method gives you a ROM that you can plug in any card without worrying about configuring it first.

OP just needs to fix up the location of the second copy (to 4000 instead of 3FF0) and reburn the ROM.

jmarsh wrote on 2022-11-27, 20:16:

maxtherabbit wrote on 2022-11-27, 20:05:

Oh, well IIRC you can configure the card to provide a 32kB window which would eliminate all this BS

Depending on the card and the config program used you can pick which 16KB window of the ROM to use - obviously requires the card to control the high ROM address pins rather than just wiring them directly to either supply or ground. That seems unlikely here since the card is apparently defaulting to the second (misaligned) window/copy of XTIDE, and besides the multiple copies method gives you a ROM that you can plug in any card without worrying about configuring it first.

OP just needs to fix up the location of the second copy (to 4000 instead of 3FF0) and reburn the ROM.

The 3c509 does not allow you to pick which half (/quarter) of the ROM you use. If you choose a window smaller than the physical size of the chip you will always get the upper part of the data.

Yes in this case he needs to fixup his image, but simply 0 padding a 16kB image out to fill the chip and selecting a 32kB ROM window would also work (albeit wasteful of UMA)

maxtherabbit wrote on 2022-11-27, 23:09:

The 3c509 does not allow you to pick which half (/quarter) of the ROM you use. If you choose a window smaller than the physical size of the chip you will always get the upper part of the data.

The 3C509B (specified in thread title) does let you select the window.
https://www.janwagemakers.be/PIC18F452_3COM_3 … rnet/3c5x9b.pdf
See page 54 "ROM control register".

The size of the UMA used by the card is fixed at 16KB.

Thanks heaps. Yes my error was having the 3com cfg utility set at 16kb instead of 32kb. I will fix this.
I also used an online hex calculator and stuffed it up as i didnt half 32kb correctly. Will fix this also.

Here's a link to a blog post that might be helpful. It's written by another 3com card user.

http://sbfmdrv.blogspot.com/2015/11/retrofitt … old-pc.html?m=1

Anyway, good luck! 🙂👍

I'm confident things will work soon, needs a bit of patience maybe.
Fiddling with boot ROMs is a hobby in its own reign, IMHO.