SodaSuccubus wrote on 2020-07-25, 03:22:

Simply can't get it to wanna play 100% with my CF cards. My LS486 wouldn't play fair with them either, but that was fixed by manually typing in the values for a 2GB HDD in the bios. Same trick doesn't seem to work here.

I had similar issues with an old Schneider Tower AT from Western Germany..
Re: Fixing a Schneider 386 SX System 40

To my understanding, 286 BIOSes are one of the earliest piece of software in regards of IDE/ATAPI.
In fact, they aren't even aware of IDE. What they know is the original WD1003 instruction set, which is what the controller of MFM/RLL HDDs often used.
So that code may or may not even predate ESDI (a short livd technology released in between of WD1003 and IDE).
IMHO, best bet is to use an external IDE BIOS instead of the internal HDD controller code, if possible.
https://en.wikipedia.org/wiki/Western_Digital#1980s
https://datasheetspdf.com/pdf/499114/WesternDigital/WD1003/1

If you're interested in the low-level stuff, have a look at:
http://www.os2museum.com/wp/how-to-please-wdctrl/

It describes the changes that came with PATA-2 and how they differed enough to make the original Windows 3.1 hard disk driver fail.

Also: Inj CMOS Setup, you can can enter any fake values just fine,
as long as they are within the limits of IDE+PC BIOS and the physical capacity of the flash disk.
Re: M321 BIOS HD limitation question

So i'd need an XTIDE then. This board is running an AMI-BIOS from 04/09/90. I guess i figured it would have had IDE by then.

Wouldn't XT-IDE cripple the performance though as most pre-builts are only 8bit cards?

SodaSuccubus wrote on 2020-07-25, 04:56:

So i'd need an XTIDE then. This board is running an AMI-BIOS from 04/09/90. I guess i figured it would have had IDE by then.

Well, I don't know this board in particular, but my PC's BIOS was from '87, '88 and came with an MFM/RLL card.
XTIDE Universal BIOS will also work with normal IDE interfaces for ISA bus (multi i/o cards) and on-board IDE.
All it does is doing the BIOS' job of accessing the HDDs (just set HDD type to "none" in CMOS Setup of your BIOS).
And since XUB has auto-detection, it's not even necessary to enter any drive geometry.

SodaSuccubus wrote on 2020-07-25, 04:56:

Wouldn't XT-IDE cripple the performance though as most pre-builts are only 8bit cards?

Yes, it would. You also would need an AT compatible version, because the PC/XT had slightly different i/o adresses for HDDs.
Of course, it should work anyway, but it may cause some buggy behaviour..
Here's an older video that I took years ago. It shows an XT HDD controller card running in a Pentium PC:
https://www.youtube.com/watch?v=irtDxYman88

Simplest way would be to continue using the IDE controller that you used before, I guess.
Just write XTIDE Universal BIOS onto an Flash/EPROM chip and use, say, an ISA network card (or Year 2000 BIOS fix card) as a host.
Anything can be used, as long as it puts the XUB code in the 640KB-1MB area where the BIOS looks for Option-ROMs.
Some XUB images that are known to work out of box (tested/dumped them myself, set to auto detect) can be found here:
Re: IDE disk-on-module + 386

Alternatively, you can also use a network card with a DoC module. A DiskOnChip (M-Systems DoC 2000).
I haven't really tried, but it can simulate a HDD and has a special bootloader code inside. Simpy put, this is a vintage SSD.
It was used in several embedded systems running DOS. That beeing said, it's 8-Bit again..
https://hackaday.com/2018/03/30/a-retrocomput … disk-on-a-chip/
https://elektronik-kompendium.de/public/arner … tel/doc2isa.htm
http://www.daupara.de/surfstation/

Sorry for the double post but i just had a bit of a revelation....

In my infinite wisdom i just realized i neglected to type in anything for "WP" and "LZ" in the BIOS HDD section.
You don't think that uhh..would have anything to do with it having troubles reading?

I got some default values for my card by plugging it into my DX4 rig and reading what that bios reports:
31045 Cyl. 16 HD. 65535 WP. 31045 LZ. 63 SEC. Total of 15280MB. How would i shrink these values down too the 504mb limit? Anyone know any good values for a 504mb drive?

SodaSuccubus wrote on 2020-07-25, 05:56:

Sorry for the double post but i just had a bit of a revelation.... […]
Show full quote

Sorry for the double post but i just had a bit of a revelation....

In my infinite wisdom i just realized i neglected to type in anything for "WP" and "LZ" in the BIOS HDD section.
You don't think that uhh..would have anything to do with it having troubles reading?

I got some default values for my card by plugging it into my DX4 rig and reading what that bios reports:
31045 Cyl. 16 HD. 65535 WP. 31045 LZ. 63 SEC. Total of 15280MB. How would i shrink these values down too the 504mb limit? Anyone know any good values for a 504mb drive?

I can't talk for WP, but LZ mean landing zone and refers to where a non auto-parking drive would park its head. For CF and most modern drives, this is unnecessary . WP refers to write precompensation and is also likely meaningless with modern drives and CF . See https://en.wikipedia.org/wiki/Write_precompensation

Just a little update on this:

I managed to scrounge up a 95' era 1gig (I think?)Maxtor drive from the depths of my storage and gave it a go.

Waddya know. DOS 5 actually installed and can write to the disk fine! :0

I had to limit down to 504mb in the bios and it was off to the races.

Absolutely no idea what this 286 board hates about CF cards that my other 486 boards don't. So I'm gonna claim a partial victory on this one. It's gonna be a PITA to transfer games over untill I can figure out how to get it to play nice with CF.

SodaSuccubus wrote on 2020-07-25, 05:56:

I got some default values for my card by plugging it into my DX4 rig and reading what that bios reports:
31045 Cyl. 16 HD. 65535 WP. 31045 LZ. 63 SEC. Total of 15280MB. How would i shrink these values down too the 504mb limit? Anyone know any good values for a 504mb drive?

Hi, you can try 1023 cylinders, 15 heads, 63 sectory.
That's a safe value, I think.

Depending on the BIOS, you can als try 16 heads.
It depends on how the BIOS works.
Some start with head 0, rather than head 1.

The cylinder value is most critical.
1024 is the maximum in theory, but using less is safer (1023).
That's because old MFM/RLL HDDs had a diagnostic cylinder at some position also.

There's also E-CHS, aka Large. It uses fake values (logical CHS).
Most 486 systems supported this before LBA became the norm.

Write precompensation and landing zone. Latter is to amplify the write strength change beyond certain cylinders and landing zone is reserved for inner cylinders for heads to land on at each power down, providing that you used parking head program.

Cheers,

Oh I have the same board as you !

I got mine working with a SD card so you should definitely be able to use it with a CF card !

Jp14-15 are indeed improperly documented by MTL ...
I believe "enabled" = jp14 and disabled = jp15 instead (same for the floppy disk drive)

Deksor wrote on 2020-07-27, 15:49:

Oh I have the same board as you ! […]
Show full quote

Oh I have the same board as you !

I got mine working with a SD card so you should definitely be able to use it with a CF card !

Jp14-15 are indeed improperly documented by MTL ...
I believe "enabled" = jp14 and disabled = jp15 instead (same for the floppy disk drive)

Hey Deksor! I tried that but I still wasn't able to get the onboard IDE to enable. Mind sharing some pictures of your board and jumper configuration if it isn't much trouble?

Maybe CF will work over the onboard IDE? Not sure why it would hate my generic Goldstar IO card

Mine has :
Jp15 set
Jp14 unset
Jp10 set
Jp9 unset

Basically the opposite of yours.

SodaSuccubus, the board looks pretty fresh and clean. What did you use to clean the corrosion? How did you clean that?

deksar wrote on 2020-07-27, 16:43:

SodaSuccubus, the board looks pretty fresh and clean. What did you use to clean the corrosion? How did you clean that?

Regular vinegar for the corrosion. Just dabbed some onto a paper towel and gently scrubbed away at the corrosion as best I could.

Afterwords it was Rubbing alcohol wipes+ dry qtips for everything else and a can of compressed air for the ISA slots.

Deksor wrote on 2020-07-27, 16:31:

Mine has : Jp15 set Jp14 unset Jp10 set Jp9 unset […]
Show full quote

Mine has :
Jp15 set
Jp14 unset
Jp10 set
Jp9 unset

Basically the opposite of yours.

Just tried that. Still couldn't get onboard to work 🙁
Maybe the corrosion got to it. No idea

If setting the jumpers as Deksor wasn't able to get it to work, I would check the traces around that area. The traces around the 245s and the PAL chip looks corroded, and some pins of the PAL and its socket look like having some green stuffs too.

kalohimal wrote on 2020-07-27, 17:02:

If setting the jumpers as Deksor wasn't able to get it to work, I would check the traces around that area. The traces around the 245s and the PAL chip looks corroded, and some pins of the PAL and its socket look like having some green stuffs too.

Ah yes. Looking at it now there is still a bit of gunk on the ide connector and the legs of some of those tiny chips. I might have to give it another scrub.

The actual traces look okay though. Except for that big blotch spot where you can see I did my major corrosion cleaning.
Il give it another scrub here and see if that changes.

Honestly I wouldn't be too worried about loosing onboard IDE if I could get my ISA IO card to play nice with CF. It is just this one board it has issues with. I sware.

I have no idea what in sees in that 1gb Maxtor it doesn't in CF