MikeSG wrote on 2025-12-05, 12:32:

Do none of the 1-15 settings in BIOS match what Speedstor / Ontrack says? I believe the BIOS needs something entered to detect the drive.

Do you have another MFM drive that you know works, so you can rule out everything else?

I can only enter a disk type from 1 to 15 in the BIOS, no way i can see to change disk parameters.
There was a page that lists what they are, but i haven't cross checked them yet in speedstor yet.
I know that info changes according to the disk type but will be interesting if speedstor lists them differently.
I have not tried running speedstor yet with only one drive connected, only ontrack which will hang.
The only mfm drive I know works is inside another computer stashed away, but ive got a ton of untested spares.
I might try another controller card first.

MikeSG wrote on 2025-12-04, 12:45:

The Ontrack disk manager (bootdisk) can identify drive data (cyls/head/sectors/etc) for entering into bios: https://www.philscomputerlab.com/ontrack-disk-manager.html

I need to re-enter the numbers each time I start my AcerPower 386/33w (also with a Ctrl+Alt+Esc Bios entry). Only two numbers need to match. If the whole row is not a match it's not a big deal.

OK, so some good news here, ive got one of the drives working, the other drive which is connected at the end of the cable presumably is drive 1 and the main boot drive, this drive appears to have had some issues and I noticed the motor making some noise and then it stops spinning until powered off and back on.
I disconnected it but still could not get anywhere with ontrack working, but when I shifted the second drive to the end connector on the ribbon cable, it worked. I also switched the data cable to drive 1 at this time.
I then ran ontrack and it picked up a partition and warned me it was going to wipe the drive if i proceeded the setup.
I managed to find some drive specs in ontrack, but its not providing everything the bios is asking for, and also the bios wont let me enter the correct specs.
See screenshots below, specs are as follows:
cylinders 924, heads 5, sectors 17, im assuming buffer size equates to sector size? (says 512 here)
What im having trouble with here is i cant go over 15 sectors per track here but this drive is 17.
The bios didnt ask me for the number of cylinders either, I also had to guess the rest of the data, so went with 40 tracks, 2 sides and made an educated guess on head settle time, step rates, etc(see screenshots)
I also selected type 14 as a drive type for the bios, i have no idea if this is correct, but is the closest corresponding size I could find from a table I found.
Anyway, I can access C: drive now, but the contents are garbled, see screenshot. I have not tried opening any of the folders yet.
My question is, is this due to incorrect drive specs, or is this just corrupted data?
I did a read test on the disk with success, but i got one failure on a random seek test.
I will give it a low level format if this likely just corruption/old age, etc.
I need to see if this controller supports RLL, or else i will find another controller to use, but im getting there now.
One other issue on this computer is some keys, arrow keys in particular will make the computer beep when pressed, but it still works, Im yet to find another keyboard to try, but its a bit odd.

I would do the same as you, selecting the 40.21MB model.

The blue Ontrack page for me says a lot more. "Device Identify CHS... 1978 x 16 x 63 = Size: 1.020GB"
CHS = Cyls, heads, sectors.
In BIOS I only match Head = 16, Sector = 63. And it works as a 1GB, despite cylinder count being wrong.

For yours, I believe Sector count needs to be 17 but could be wrong. Sector size seems right at 512. Not sure about the others.

Another controller might be better if the data is not readable after format.

MikeSG wrote on 2025-12-08, 14:45:

I would do the same as you, selecting the 40.21MB model. […]
Show full quote

I would do the same as you, selecting the 40.21MB model.

The blue Ontrack page for me says a lot more. "Device Identify CHS... 1978 x 16 x 63 = Size: 1.020GB"
CHS = Cyls, heads, sectors.
In BIOS I only match Head = 16, Sector = 63. And it works as a 1GB, despite cylinder count being wrong.

For yours, I believe Sector count needs to be 17 but could be wrong. Sector size seems right at 512. Not sure about the others.

Another controller might be better if the data is not readable after format.

Ontrack is reporting 17 sectors, so that looks right.
But the system bios config seems limited, it doesn't ask for cylinders or heads, I dont know if its an issue, as the disk controller has all this saved so does it override the bios on boot?

I don't think it overrides, I think it adds to drive information.

The MFM controller needs to know the difference between drive letters... A/B/C. In modern BIOS's, values are standardised and you are only selecting 1.2/1.44MB floppy drive types, and standard hard drive types.

I haven't used early 286's.

IMO, what is set up now is a good position to test format and read/write tests.

MikeSG wrote on 2025-12-09, 15:14:

I don't think it overrides, I think it adds to drive information. […]
Show full quote

I don't think it overrides, I think it adds to drive information.

The MFM controller needs to know the difference between drive letters... A/B/C. In modern BIOS's, values are standardised and you are only selecting 1.2/1.44MB floppy drive types, and standard hard drive types.

I haven't used early 286's.

IMO, what is set up now is a good position to test format and read/write tests.

Ok, so if I can access the C drive and see the folder structure, then its a good chance its all working right?
Given that the file names are garbled, I was concerned that incorrect disk geometry was causing this.
If this is likely corruption due to age then I wont worry about loosing anything and will give a low level format.

I was interested to see what software may have been on it as this machine was doing CAD work, but looks like the chances of finding anything is low.
This disk controller appears to be working ok, but i do have a bunch of others which some will have their own bootable bios.
Im assuming these cards must have flash memory on board to save all the configuration data?

Im pretty sure this card can support RLL, will ontrack ask me to choose between mfm and rll when I format the disk?

Made some progress tonight, seemed to display the filesystem fine this time round but only had a single log file visible this time, but was able to view its contents dated 1994.
I decided to give the drive a low level format and to my surprise it didnt take long.
I ran into issues with ontrack, it failed at 50%, see screenshot, so went with speedstor and it formatted it fine.
Unfortunately when the format completed, it asked me to press any key but was unresponsive, and failed to boot again after resetting.
Have been messing around all evening and reseated all the DIP chips i could find on the motherboard along with the cards. Really bizarre, but there was something wrong with this unit where some keys such as the arrow keys would cause the speaker to make a beeping sound when pressed. Makes me wonder if there is a bad capacitor on the motherboard or something?

Bad capacitors are possible. Check the 5V rail, if it's below ~4.8V it could mean worn capacitors, and spikes could dip below the minimum voltage threshold for some chips. Changing the capacitors around the PSU connector on the motherboard can bring the voltage up, more than changing other capacitors.

Ontrack is more for modern drives, so Speedstor might be the application to use..

Going back to the max '15 sectors per track' of your current drive controller.... This page says the most common number was 17, and RLL potentially doubles that number when low-level formatted.
https://www.redhill.net.au/d/10.php

Does the MFM controller provide some format command?

MikeSG wrote on 2025-12-10, 13:12:

Bad capacitors are possible. Check the 5V rail, if it's below ~4.8V it could mean worn capacitors, and spikes could dip below th […]
Show full quote

Bad capacitors are possible. Check the 5V rail, if it's below ~4.8V it could mean worn capacitors, and spikes could dip below the minimum voltage threshold for some chips. Changing the capacitors around the PSU connector on the motherboard can bring the voltage up, more than changing other capacitors.

Ontrack is more for modern drives, so Speedstor might be the application to use..

Going back to the max '15 sectors per track' of your current drive controller.... This page says the most common number was 17, and RLL potentially doubles that number when low-level formatted.
https://www.redhill.net.au/d/10.php

Does the MFM controller provide some format command?

I will take a look at the voltages this morning, I think it will be a tantalum thats failed.
Its going to be a bit of a nuisance getting to everything on this motherboard, its got so many things covering it, but when I first powered it on i did notice a bit of a smell but no smoke to be seen.

I looked through the manual for this disk controller and says to use IBM diagnostics to perform the low level format.
I think this controller is MFM only but others in this series do support RLL.
I have other controller cards here that have built in format programs on the ROM and can do RLL.
Do most older drives support RLL with no issues? I can check the model number when I get it out.

Edit.
I got the drives out, looks like they are very similar models from control data corporation.
The one that has issues spinning up had a chip taped to the back of it.
Looks like it plugs into the rear. Not sure why it was removed.
Has some post it notes stuck everywhere with something about removing a termination resistor if both drives are used.
Not too familiar with any of that. Also has some jumper diagrams.
See photos.

As far as voltages go, the motherboard power connector has a bunch of black wires that all go to ground, some yellow, white, blue, red and orange
All the yellow read 4.98v
Orange reads 4.63v
White is 5.13v
Blue is 12.27v
Red is 12.13v

Does this look normal?
The 5v rails seem ok anyhow, not sure about the others.
I can't see anything burnt out caps anywhere, im getting 4.98v and 12V across the ones near the power connector.
I measured a few of the yellow tantalum caps around the board and all those are 4.98v. I can't measure the ones that the power supply covers when fitted but I think they are ok

I also dusted the motherboard and was able to reseat all the chips.
I put it all back together and was able to get it to post but had no disk controller connected, I put that in and now nothing.
So I took it out and still nothing!
Going round in circles here, but intermittently i can get a single post beep but no image, or sometimes I will get 3 short beeps.
Possibly a memory error of some kind?
I had been getting the message CMOS RAM STAMP ERROR every time I switched it on but it still ran fine.
Im assuming this is due to no battery?
Still got to find a battery for this, its an unusually high voltage for a PC, this one is 6.8v

I assume most older drives support RLL from what I've read.

Is the 'terminator resistor' used on the drive if only one drive is connected now?

Voltages look fine, although the White cable on the AT connector should be -5V. Don't know if your system uses it for anything, most don't or only for some sound cards.

Maybe some conflict with the drive/controller creates some halt event. Memory can last some time after shutdown..

Does it boot with the controller connected but no drive..?

MikeSG wrote on 2025-12-11, 15:20:

I assume most older drives support RLL from what I've read. […]
Show full quote

I assume most older drives support RLL from what I've read.

Is the 'terminator resistor' used on the drive if only one drive is connected now?

Voltages look fine, although the White cable on the AT connector should be -5V. Don't know if your system uses it for anything, most don't or only for some sound cards.

Maybe some conflict with the drive/controller creates some halt event. Memory can last some time after shutdown..

Does it boot with the controller connected but no drive..?

OK, i had been reading a bit on how some drives didnt work with RLL well, so might be something specific or just limited to real early models perhaps?
I know there were some that were advertised as RLL or at least more optimized for it i guess.

Regarding the terminator resistor, I cant see anything anywhere, Im not even sure where they are supposed to go but i saw a picture online where it was pushed into the edge of the IDC connector, but I dont have the impression thats how its done.

Anyway, last night i stripped everything off the board gave it a bit of a dust off, reseated all the chips and got it working with just the video card installed, i thought great i had fixed it so installed the HDD controller and then nothing, no beeps or anything, so i took the HDD controller out and still nothing!
Pulled out the video card still nothing, however randomly if i try turning on at times I might get a single post beep even with no video card, i got it to beep with the video card but no image, then another time i got 3 beeps which i think means memory, I took out the expanded RAM and still nothing.
Put it back in and nothing. Have tried different video cards.
Makes me wonder if there is a dry joint somewhere?
I cant measure the voltage on every single capacitor, but i dont think they have failed. None of the tantalum caps have burnt out that i can see.
It looks a real pain to remove the motherboard on this but dont know if its going to tell me much.

I think the capacitors are fine. Tantalum's explode when they break, or voltage level will be lower but yours is 4.98v spot on.

After leaving it for CMOS memory to clear again, and RAM memory, I would try only the video card... Boot into BIOS, clear HDD settings, then try the controller on it's own, no HDD... find where the failure point is.

At any point where it's running normally, try jiggling everything... cards, cables...

Could also try another power supply to make sure.

Good possibility its the power supply.
I know nothing about repairing these, but its a unique power supply to fit in here. Could be a bad capacitor inside perhaps?
It looks like its using an XT style power connector, but dont think I have any compatible power supply to test it with.
I ended up taking the whole motherboard out last night, checked over everything and visually its OK, its had a fair ton of factory mods done with jumper wires soldered everywhere, seemed to be common for stuff in the 80s.

It should be AT compatible... AT has the second red wire as in this picture.

AT-ATX adapters are inexpensive

Ok, will just need to make some sort of jumper cable to test with.
There are a few differences on mine though, there is 3 yellow in a row from memory.
I don't know how easy these switch mode power supplies are to repair, but if it turns out its at fault, then its my only option as no other psu will fit in this case.

As long as the voltages match....

Capacitors are capacitors.... but tracking down which component isn't working might be the tricky bit unless you can buy internal replacements of the whole PSU. If that is the problem..

MikeSG wrote on 2025-12-13, 12:15:

As long as the voltages match....

Capacitors are capacitors.... but tracking down which component isn't working might be the tricky bit unless you can buy internal replacements of the whole PSU. If that is the problem..

I reckon it could be a capacitor, there was a hint of some smell when i first turned it on, but couldnt see where it come from or smell anything when i put my nose near.

Was able to wire up another AT power supply, but no luck unfortunately