I'm assuming from you mention of the switch "on the PSU", and "fireworks"
that this is an older-style AT supply (power switch on mains - cabled directly
to the supply), and not an ATX (power switch on mainboard jumpers)

In this case, assuming the PSU casing is properly grounded (3-prong power
cord plugged into grounded outlet) -shorting mains to ground shouldn't damage
anything with the system - for all intentional the short would have been
effectively external to the system.

Most likely there has been a fault for another reason which happened about the
same time - possibly induced current spike or a gound loop - both very unlikely.

Does the floppy seek/home head during POST? Try moving the head all the way
out (inner tracks) and see if it homes to outmost track during POST.
This will tell you if the FDC selection logic and step signals are working
and addressing the drive.

A good tool for testing low-level floppy operation is ImageDisk - this will
let you perform basic seeks, see if the drive is reporting index pulses,
track-zero detect etc.

I suggest trying "Clean head" - this will seek back and forth as long as the
drive is seeing index pulses and not require any data from the drive.

Your BIOS screenshot shows both 5.25" and 3.5" drives - are both not working?
Does the 5.25" drive have a termintion jumper resistor pack?

DaveDDS wrote on 2025-07-26, 18:32:

I'm assuming from you mention of the switch "on the PSU", and "fireworks" that this is an older-style AT supply (power switch on […]
Show full quote

I'm assuming from you mention of the switch "on the PSU", and "fireworks"
that this is an older-style AT supply (power switch on mains - cabled directly
to the supply), and not an ATX (power switch on mainboard jumpers)

In this case, assuming the PSU casing is properly grounded (3-prong power
cord plugged into grounded outlet) -shorting mains to ground shouldn't damage
anything with the system - for all intentional the short would have been
effectively external to the system.

Yup, you are correct! I am pretty sure ground was in place then...

DaveDDS wrote on 2025-07-26, 18:32:

Most likely there has been a fault for another reason which happened about the same time - possibly induced current spike or a g […]
Show full quote

Most likely there has been a fault for another reason which happened about the
same time - possibly induced current spike or a gound loop - both very unlikely.

Does the floppy seek/home head during POST? Try moving the head all the way
out (inner tracks) and see if it homes to outmost track during POST.
This will tell you if the FDC selection logic and step signals are working
and addressing the drive.

A good tool for testing low-level floppy operation is ImageDisk - this will
let you perform basic seeks, see if the drive is reporting index pulses,
track-zero detect etc.

I suggest trying "Clean head" - this will seek back and forth as long as the
drive is seeing index pulses and not require any data from the drive.

Your BIOS screenshot shows both 5.25" and 3.5" drives - are both not working?
Does the 5.25" drive have a termintion jumper resistor pack?

Yes, I can hear both drives are seeking during POST. Also, they are both spinning when trying to access the drives (and giving the error message).

The 1.44mb doesn't have any jumpers, while the 1.2mb only has what to appear to be for device id selection (see attached images).

Linoleum wrote on 2025-07-26, 19:09:

The 1.44mb doesn't have any jumpers, while the 1.2mb only has what to appear to be for device id selection (see attached images).

I wasn't referring to "jumpers" .. 5.25" drives often have what looks like an IC socket where a 150ohm
Terminarion pack goes - this is only for the drive at the end of the cable. 3.5" drives (and some newer 5.25"
drives) doen't have this and have partial termination built in. As the drive was working before and
you've not changed it, this won't be the problem, it's just part of what I look for when testing out a
floppy drive.

If you have a scope, you could look at the data signals and see if the drive is sending good signals (If it works
on other systems, it most likely is - but always worth checking to see if the data line is excessively pulled
high or low on the mainboard side.

Well, I just tested one of the drives (the 1.2MB) on another system, and it worked right away—so the issue doesn't seem to be with the drives themselves.

Another possibility is that my original I/O controller (Kentech Computer IDE PLUS-V2) has gone bad, and I may not be configuring the replacement correctly. See the picture: most settings are left at their default (marked with an asterisk), except for the ones I changed, which are indicated by arrows.

After fiddling with AMISETUP, I used the Diagnostic utility and saw the "DMA Clock Speed" blinking in red (as in, it's bad) showing 6.24... I believe DMA is used for floppy disk drives. Could this be the problem?

Yes, The PC Floppy interface does use DMA - if the DMA channel is bad, that would certainly prevent the disks from working.

Sadly, this is kindof built-in the the mainboard, if you have (or can obtain schematics) you might be able ti figure out what exactly went bad.
Even if you can't find schematics for that board, you might fine one for a board that uses the same chipset and the DMA channels might be
"close enough" to work from.

Unfortunately, except for 8088s, DMA is likely built into one of the system chipsets, which will be hard to find, and difficult to replace.
If you are lucky it might be one of the smaller supporting components which are much easier to replace.

Do other things using DMA work (hard drive for example) - this would suggest some hope that the DMA controller itself
isn't completely fried - but you never know, there can be partial failures in the more complex parts of a chipset.

I see from your photos that you are running DOS - do you have other ways to get stuff on/off the system (network etc)?
If so, it might be worth trying my ImageDisk (available at "Daves Old Computers") it talks directly to the floppy hardware
without going through DOS/BIOS (I have Fdrive set to "None" in my IMD system BIOS - but it works well either way - I use
"None" so the system doesn't try to Identify non-PC drives which are often connected - like 8")
(If no network, does the system have a serial port? - might be another way to put stuff on)

It might be interesting to see what IMD reports.

IIRC the PC does support a secondary floppy disk controller - I don't recallif it uses the same DMA channel...
(It's been years since I got into such detail) you *might* be able to work around by installing an FDC configured
as secondaty

DaveDDS wrote on 2025-08-31, 22:43:

Yes, The PC Floppy interface does use DMA - if the DMA channel is bad, that would certainly prevent the disks from working. […]
Show full quote

Yes, The PC Floppy interface does use DMA - if the DMA channel is bad, that would certainly prevent the disks from working.

Sadly, this is kindof built-in the the mainboard, if you have (or can obtain schematics) you might be able ti figure out what exactly went bad.
Even if you can't find schematics for that board, you might fine one for a board that uses the same chipset and the DMA channels might be
"close enough" to work from.

Unfortunately, except for 8088s, DMA is likely built into one of the system chipsets, which will be hard to find, and difficult to replace.
If you are lucky it might be one of the smaller supporting components which are much easier to replace.

Do other things using DMA work (hard drive for example) - this would suggest some hope that the DMA controller itself
isn't completely fried - but you never know, there can be partial failures in the more complex parts of a chipset.

I see from your photos that you are running DOS - do you have other ways to get stuff on/off the system (network etc)?
If so, it might be worth trying my ImageDisk (available at "Daves Old Computers") it talks directly to the floppy hardware
without going through DOS/BIOS (I have Fdrive set to "None" in my IMD system BIOS - but it works well either way - I use
"None" so the system doesn't try to Identify non-PC drives which are often connected - like 8")
(If no network, does the system have a serial port? - might be another way to put stuff on)

It might be interesting to see what IMD reports.

IIRC the PC does support a secondary floppy disk controller - I don't recallif it uses the same DMA channel...
(It's been years since I got into such detail) you *might* be able to work around by installing an FDC configured
as secondaty

That's the weird thing! My Audician 32+ works without any problems (SB and WSS)...

That ImageDisk idea sound like an option! I've got a CF card attached to it, so doable. And I do have another I/O controller. That would be crazy!

I'll get to it and report back tomorrow.

PS: Polpo's post gave sent me in this direction (Strategies for troubleshooting low DMA not working on 386 motherboard?). He ended up swapping the chip... My is a CHIPS: https://theretroweb.com/chipset/documentation … d1992683703.pdf

Just to be sure, I fixed the DMA clock speed error with PRO CLK/5 and it's now showing 5mhz without any error message.

Now, I tried a second FDC (where only floppy drives were enable, of course) and I could hear the drives initiate and boot without any error... But, that did work.

I also tried with IMD and the TESTFDC utility, and gave me these error messages (see attached pic)... I tried with and without floppies at "none" and also with just one floppy at a time.

Linoleum wrote on 2025-09-01, 13:43:

Just to be sure, I fixed the DMA clock speed error with PRO CLK/5 and it's now showing 5mhz without any error message.

Ok, I'm not familier with "PRO CLK" - I'm not sure what that would do to the system... And I still wonder why things changed - it was
working before.

Btw (sorry - not re-reading the thread), did you try clearing the CMOS? It's possible that something got flipped thats causing problems...
unlikely as the CMOS checksum should have caught any corruption, but you never know...

Now, I tried a second FDC (where only floppy drives were enable, of course) and I could hear the drives initiate and boot without any error... But, that did work.

Ok... I'm assuming "drives ... boot without error" and "that did work" is a typo and booting didn't work? (Otherwise your drives are working)

I also tried with IMD and the TESTFDC utility, and gave me these error messages (see attached pic)... I tried with and without floppies at "none" and also with just one floppy at a time.

TESTFDC is more about finding out what non-PC formats your floppy controller can handle and not so much about noting problems with it,
It assumes your drives are working, and when it gets errors it guesses the format it's trying doesn't work on your setup, and moves on to
the next.

I was thinking more of IMD itself in trying to read a disk and also the Align/Test function - not expecting it to give you a whole
lot of hardware debugging information (it also assumes the drives work) - but it would at least tell you if the problem is hardware
(sounds likely in this case) or software/config as it ignores BIOS configuration and talks to the FDC hardware directly
- It's also possible it might report some little detail about problems it encounters that DOS/BIOS wouldn't (again, it's not focused
fixing hardware and assumes it has working drives - so don't expect too much - but it's worth a try)

DaveDDS wrote on 2025-09-01, 23:58:

Ok, I'm not familier with "PRO CLK" - I'm not sure what that would do to the system... And I still wonder why things changed - i […]
Show full quote

Ok, I'm not familier with "PRO CLK" - I'm not sure what that would do to the system... And I still wonder why things changed - it was
working before.

Btw (sorry - not re-reading the thread), did you try clearing the CMOS? It's possible that something got flipped thats causing problems...
unlikely as the CMOS checksum should have caught any corruption, but you never know...

No worries, that's a fair question! Yes I did, more than once I would even say...

DaveDDS wrote on 2025-09-01, 23:58:

Ok... I'm assuming "drives ... boot without error" and "that did work" is a typo and booting didn't work? (Otherwise your drives are working)

Correct! FDC settings and drives are not preventing the computer from booting and I even enabled "drive seek" with amisetup to see if it would detect something else, but no! Everything seem to work until I try to access the drive...

DaveDDS wrote on 2025-09-01, 23:58:

TESTFDC is more about finding out what non-PC formats your floppy controller can handle and not so much about noting problems wi […]
Show full quote

TESTFDC is more about finding out what non-PC formats your floppy controller can handle and not so much about noting problems with it,
It assumes your drives are working, and when it gets errors it guesses the format it's trying doesn't work on your setup, and moves on to
the next.

I was thinking more of IMD itself in trying to read a disk and also the Align/Test function - not expecting it to give you a whole
lot of hardware debugging information (it also assumes the drives work) - but it would at least tell you if the problem is hardware
(sounds likely in this case) or software/config as it ignores BIOS configuration and talks to the FDC hardware directly
- It's also possible it might report some little detail about problems it encounters that DOS/BIOS wouldn't (again, it's not focused
fixing hardware and assumes it has working drives - so don't expect too much - but it's worth a try)

I did give the Align/Test function a try. I was getting data and could hear the drives spinning. However, I am not too sure I am using the tool correctly...

Linoleum wrote on 2025-09-03, 23:42:

I did give the Align/Test function a try. I was getting data and could hear the drives spinning. However, I am not too sure I am using the tool correctly...

So... IMD was not reporting errors?
Seems very odd, as any bad data would at least cause a CRC error...

Can you read a disk into an image without error?

When you try and read track0 <> track30 - do you see/hard the drive seeeking?

If everything seems to work in IMD .. then maybe the problem is reading incorrect data from the drives..

If you can read an image, then look at that image with IMDV - can you see tet strings within the image?
(I'assuming whatever you have on diskette has text strings within it)

If no, and you have another DOS system can can write the disks, I could make up an imaeg for you that has
easily recoganizable data im the sectors which would let you see which track,head.sector it actually came from
and tell if that sector was read coffectly.

After swapping out the motherboard and leaving everything else unchanged, both floppy drives are now working perfectly. Turns out the issue really was the motherboard—how bizarre!