Try the "normal" without any precomp and landz values. Basically forget anything LBA since the machine doesn't have auto-detect, including detection on a modern system.
Another thing you can try is using a utility like IDEINFO (search in the forum, I'll post it if you can't find it) and do the detection on the same machine you intend to use the card (correct BIOS values aren't required for the utility)

konc wrote on 2024-06-12, 13:42:

Try the "normal" without any precomp and landz values. Basically forget anything LBA since the machine doesn't have auto-detect, including detection on a modern system.
Another thing you can try is using a utility like IDEINFO (search in the forum, I'll post it if you can't find it) and do the detection on the same machine you intend to use the card (correct BIOS values aren't required for the utility)

Still no joy. I'm wondering if it has to do with the settings on the Winbond card i'm using as the IDE interface (and FDD interface - that FDD is giving me a drive seek failure the moment its activity indicator lights up) though I can't really see how. the one non-default thing I have at the moment is the IRQ14 jumper open (I'd tried it disabled, same results, HDD failure). I have the drives set for buffered, both FDD and HDD are using their primary port addresses, and are enabled. I have com3 for asyn1 and com4 for asyn2 (there are no IDE or Asyn ports on the SBC, jumpers or connectors). I can't use IDEinfo on the machine I'm intending to use the card on as I can't get it booted with either HDD or FDD. Once I can boot to an FDD I will try it though.

As a followup I tested:
the 1.44 floppy drive - works on the P3 SBC with the built in FDD controller. Gives nothing but errors on the DTI SBC via the Winbond mult-io card. I used the same ribbon cable.
the SD card drive - also works on the P3 SBC with the built in HDD controller. Gives errors on the DTI SBC via the winbond card. I used the same ribbon cable. I tried the "Normal" settings as shown above on the P3 SBC (which is where I got those values) and surprise, it didn't work on the P3 SBC. LBA does.
I tried further tests on the DTI SBC after disabling the serial and printer ports on the WInbond multi-io card, and no change in results.

Win10 told you it was 128GB. How did you come to the conclusion that it is 128MB? From the autodetect on a retro system?

If you already have it setup on one system using LBA, then I don't expect it will work on another one using CHS. If this is indeed a 128MB card, then I would try setting it up in CHS (normal) mode with 242 cylinders, 16 heads, 63 sectors, landing zone 242, and precomp 0. Repartition/reformat, add some files, then try it in the 386 with the same settings.

If it is a 128GB card like Win10 said, then you could probably go up to 1023 cylinders, which is likely the limit of the 386 BIOS.

bakemono wrote on 2024-06-12, 21:00:

Win10 told you it was 128GB. How did you come to the conclusion that it is 128MB? From the autodetect on a retro system?

If you already have it setup on one system using LBA, then I don't expect it will work on another one using CHS. If this is indeed a 128MB card, then I would try setting it up in CHS (normal) mode with 242 cylinders, 16 heads, 63 sectors, landing zone 242, and precomp 0. Repartition/reformat, add some files, then try it in the 386 with the same settings.

If it is a 128GB card like Win10 said, then you could probably go up to 1023 cylinders, which is likely the limit of the 386 BIOS.

Hi - thanks for responding.
I read the SD chip's label to determine its size, see the pic attached. I've seen SD cards come in at less (sometimes a lot less) than the labelled amount, but never more, unfortunately.

I've tried normal settings on both the 386 and the Pentium III in BIOS, and neither of them worked. Yes, even though the autodetect on the PIII BIOS gave me those Normal settings, they did not in fact work on that SBC. LBA did.

I think you're on the right track about the fact it is LBA being the reason. I'm looking into getting this doubleram card that I can put the XT-IDE BIOS on which I suspect will resolve the issue both with the SD card and HD floppy.
https://monotech.fwscart.com/product/doublero … le-rom-isa-card

I also have a Flashfloppy on the way, I've used one of these successfully on other old machines (I have one working fine in a Zenith Z-157 8088 system) and that might be configurable to overcome the issue too.

New question: I've been trying to put a new BIOS on this 286 board and it gives me an error saying BIOS checksum error check ROM (27256?) .. well I've been using 28c256 eeproms. The board starts to boot on the various bios but eventually all hit that error. Does this mean the bios will only work on 27(C)256 eeproms and not 28c256?

NHVintage wrote on 2024-06-21, 23:41:

New question: I've been trying to put a new BIOS on this 286 board and it gives me an error saying BIOS checksum error check ROM (27256?) .. well I've been using 28c256 eeproms. The board starts to boot on the various bios but eventually all hit that error. Does this mean the bios will only work on 27(C)256 eeproms and not 28c256?

No, it means the checksum stored in the ROM doesn't match the data.

Pretty sure when you used powershell on the win10 machine, you got info from a different drive instead of the SD card. Those parameters are nothing close to what a 128MB card would have and very obviously are from a 128GB drive.
What you want to use are the "normal" parameters you first posted, just ignore the precomp value.

How old is that 386 BIOS? Is it from the 80s?
If so, there might be other compatibility issues besides lack of LBA support.
ATA-2 and ATA-3 standards changed certain things the HDD routines in the 386 BIOS perhaps can't handle.

Personally, I'd try to use the floppy verson of XTIDE and see if it works.
If it does, I'd consider installing XT-IDE in an EPROM and use a network card as a host etc.
XUBDisk - floppy disk XTIDE Universal BIOS booter

In my opinion something is not working properly on your Winbond card.
I'd start from making floppy drive working and then work on HDD.
Since you've mentioned the FDD worked on P3 then cable and drive seems to be Ok.
Can you post photo of your current jumper settings on Winbond card? And also a photo of BIOS settings for FDD.

Or maybe it's the SD/IDE converter or SD card?
SD card technology is cheapest consumers technology.
CF cards have IDE mode, by contrast. They merely need a passive, mechanical adapter.

jmarsh wrote on 2024-06-22, 01:37:

No, it means the checksum stored in the ROM doesn't match the data. […]
Show full quote

NHVintage wrote on 2024-06-21, 23:41:

New question: I've been trying to put a new BIOS on this 286 board and it gives me an error saying BIOS checksum error check ROM (27256?) .. well I've been using 28c256 eeproms. The board starts to boot on the various bios but eventually all hit that error. Does this mean the bios will only work on 27(C)256 eeproms and not 28c256?

No, it means the checksum stored in the ROM doesn't match the data.

Pretty sure when you used powershell on the win10 machine, you got info from a different drive instead of the SD card. Those parameters are nothing close to what a 128MB card would have and very obviously are from a 128GB drive.
What you want to use are the "normal" parameters you first posted, just ignore the precomp value.

Hi - I did try the Normal values I got from the Pentium III machine's BIOS's autodetect, they didn't work. Nor did the Normal values work on the Pentium III machine itself that I got the Normal and LBA values off of (I plugged it and its IDE-SD adapter into the IDE cable of the P3 machine and went into BIOS and did the auto-configure to get those results, it would autoconfigure to LBA and that would work on the P3 machine, but I tried choosing the normal values on the P3 machine that itself generated and they didnt work. nor did they work on the other machine).

I've tried multiple BIOS with the AT28C256 chips and all that work at all give that checksum error. I got new chips in a delivery last night, they didn't work either, same error. That and the error message are whats leaving me wondering. the BIOS does start, I go thru numerous numeric values on the ISA diagnostic card as I watch it boot, then it stops with the checksum error. I could see one BIOS image being corrupt, but several? The programmer verifies the programming successfully after programming the chip too.

zami555 wrote on 2024-06-22, 09:15:

In my opinion something is not working properly on your Winbond card. I'd start from making floppy drive working and then work o […]
Show full quote

In my opinion something is not working properly on your Winbond card.
I'd start from making floppy drive working and then work on HDD.
Since you've mentioned the FDD worked on P3 then cable and drive seems to be Ok.
Can you post photo of your current jumper settings on Winbond card? And also a photo of BIOS settings for FDD.

I'll do this later today or tomorrow when I'm not tied up with other RL stuff, thanks! The BIOS setting for the FDD is a simple choice of capacity, and I've run through them all with multiple drives (Including a flashfloppy) with no joy. I have a pic of the settings for

this winbond card's instructions that I've attached here, I have the jumpers for the floppy set for Enable and the standard address (tried the other too... didn't work).

NHVintage wrote on 2024-06-22, 14:58:

Hi - I did try the Normal values I got from the Pentium III machine's BIOS's autodetect, they didn't work. Nor did the Normal values work on the Pentium III machine itself that I got the Normal and LBA values off of (I plugged it and its IDE-SD adapter into the IDE cable of the P3 machine and went into BIOS and did the auto-configure to get those results, it would autoconfigure to LBA and that would work on the P3 machine, but I tried choosing the normal values on the P3 machine that itself generated and they didnt work. nor did they work on the other machine).

Has anything worked when connected to that IO card in the 386? Sounds like it can't even boot floppy drives, so you can't assume the IDE parameters are the reason for it not working.

I've tried multiple BIOS with the AT28C256 chips and all that work at all give that checksum error. I got new chips in a delivery last night, they didn't work either, same error. That and the error message are whats leaving me wondering. the BIOS does start, I go thru numerous numeric values on the ISA diagnostic card as I watch it boot, then it stops with the checksum error. I could see one BIOS image being corrupt, but several? The programmer verifies the programming successfully after programming the chip too.

Maybe they're for a slightly different board that uses a different checksum.
You didn't say why you were trying to use a different BIOS, or even where this 286 board came from and what it has to do with anything else in this thread...

jmarsh wrote on 2024-06-22, 15:07:

Has anything worked when connected to that IO card in the 386? Sounds like it can't even boot floppy drives, so you can't assume […]
Show full quote

NHVintage wrote on 2024-06-22, 14:58:

Hi - I did try the Normal values I got from the Pentium III machine's BIOS's autodetect, they didn't work. Nor did the Normal values work on the Pentium III machine itself that I got the Normal and LBA values off of (I plugged it and its IDE-SD adapter into the IDE cable of the P3 machine and went into BIOS and did the auto-configure to get those results, it would autoconfigure to LBA and that would work on the P3 machine, but I tried choosing the normal values on the P3 machine that itself generated and they didnt work. nor did they work on the other machine).

Has anything worked when connected to that IO card in the 386? Sounds like it can't even boot floppy drives, so you can't assume the IDE parameters are the reason for it not working.

I've tried multiple BIOS with the AT28C256 chips and all that work at all give that checksum error. I got new chips in a delivery last night, they didn't work either, same error. That and the error message are whats leaving me wondering. the BIOS does start, I go thru numerous numeric values on the ISA diagnostic card as I watch it boot, then it stops with the checksum error. I could see one BIOS image being corrupt, but several? The programmer verifies the programming successfully after programming the chip too.

Maybe they're for a slightly different board that uses a different checksum.
You didn't say why you were trying to use a different BIOS, or even where this 286 board came from and what it has to do with anything else in this thread...

Fair enough! I'm actually working two different issues, on two different boards, and I wound up concatenating them into one thread. so for clarity, I'll explain... wait there is no time, let me sum up...

1) The 286 motherboard. This is a board that turned out to be from a CNC machine and likely made by Trang Bow (or Trangg Bow). I know its from a CNC because when I booted it I got a very basic BIOS screen that would only look for the CNC's floppy drive (picture attached). There were no options to break out for a setup and apparently it was looking for a manufacturer specific drive. So, I have been looking to try other BIOS that will let me use the card normally. I found a few, including one that specifically listed the same board configuration and peripheral chips as mine as well as the correct chipset (Heartland 12/a). On this one, I haven't gotten past the BIOS checksum issue. The board in question: https://theretroweb.com/motherboards/s/trang- … -headland-ht12a

2) The 386 SBC card-based project. This Diversified Technology CAT985 SBC board has an early Phoenix BIOS on it - the same BIOS as the board on this Necroware video: https://www.youtube.com/watch?v=sfbqZPLYNpI - and this vid states Phx BIOS has problems with SDIDE cards that XT-IDE may help me work around; I have a network card on order that I'll put the XT-IDE BIOS on to see. However, it doesn't mention that floppies would be an issue. This is the machine I'll be using the Winbond multi-io board on. This is because, unlike many SBC's, there are no IDE or FDD connections on it. I've attached a picture of the results I got when I had the bios of a P3 machine autoconfig-scan it with its bios (it then booted using the LBA settings, which this 386's bios doesn't support). Someone here happily shared with me the documentation for the board they got when they acquired one some years ago. The SBC in question, with documentation now posted there: https://theretroweb.com/motherboards/s/divers … logy-dti-cat985

As to whether anything has worked on the Winbond card (386 project): no floppies, even a 5 1/2 inch or flashfloppy - they all give a Drive Seek Failure almost immediately. For hard drives... maybe. The only small early 90s Hard Drive - a WD caviar 3300 - that I have is pretty much dead, platter and or head issues, it makes lovely crunchy noises, but it's enough to see if the machine reacts differently, and it did take much longer as it tried and failed to successfully detect this drive.

One common thread on the floppies - the ribbon cable. It is a standard 30-something pin on the PC end, and the other end has both those and the tab type connectors I needed on the 5 1/2 inch drive. I suppose it is possible that the cable is bad, but I will have to disassemble a working machine and borrow its floppy cable to see if that's the case.

Jo22 wrote on 2024-06-22, 04:38:

How old is that 386 BIOS? Is it from the 80s? If so, there might be other compatibility issues besides lack of LBA support. ATA- […]
Show full quote

How old is that 386 BIOS? Is it from the 80s?
If so, there might be other compatibility issues besides lack of LBA support.
ATA-2 and ATA-3 standards changed certain things the HDD routines in the 386 BIOS perhaps can't handle.

Personally, I'd try to use the floppy verson of XTIDE and see if it works.
If it does, I'd consider installing XT-IDE in an EPROM and use a network card as a host etc.
XUBDisk - floppy disk XTIDE Universal BIOS booter

This is very cool. Thanks! First I need to get the floppy working on the 386, or a BIOS to boot on the 286 and then a floppy working, and I'll give these a try.

jmarsh wrote on 2024-06-22, 01:37:

NHVintage wrote on 2024-06-21, 23:41:

New question: I've been trying to put a new BIOS on this 286 board and it gives me an error saying BIOS checksum error check ROM (27256?) .. well I've been using 28c256 eeproms. The board starts to boot on the various bios but eventually all hit that error. Does this mean the bios will only work on 27(C)256 eeproms and not 28c256?

No, it means the checksum stored in the ROM doesn't match the data.

I'm not so sure here - check this article out - its for a different system but the same chips are involved. I got a pair of 27256's on the way today from Amazon - if the storms don't cancel the delivery or keep me from going out to get them - so I guess we'll see:
https://electronics.stackexchange.com/questio … -how-to-wire-we

BIOS checksum errors could mean that the replacement ROMs are too slow. Or the motherboard is mapping part of the ROM into a different address range than what the code expects. Is there a dump of the original ROMs?

NHVintage wrote on 2024-06-23, 13:37:

I'm not so sure here - check this article out - its for a different system but the same chips are involved. I got a pair of 27256's on the way today from Amazon - if the storms don't cancel the delivery or keep me from going out to get them - so I guess we'll see:
https://electronics.stackexchange.com/questio … -how-to-wire-we

I'd assumed that you would have taken that into account already...
Bend pin 1 of the 28C256 upwards so it doesn't go into the socket, connect a wire from it to pin 27 (which still should go into the socket). A14 will be driven correctly and WE# occasionally being driven low won't matter - there's a command sequence required to enable programming and it will be impossible to be fed to the chip in that configuration.