timdog wrote on 2026-05-31, 19:51:

I have a strange problem that just appeared seemingly out of nowhere with a new 486 DX4 machine I just put together. Regardless of BIOS settings and whether or not a floppy drive is connected to the machine, it always tries to boot A: first. After 15 seconds it fails then if you hit a key to continue it will boot from C: properly. This just started yesterday with really no changes to any hardware that I can think of except adding a 9 pin serial mouse to the setup for the first time. I have a CF to IDE adapter as the primary master C drive with DOS 6.22 installed. No primary slave or secondary drive at all currently. I physically removed the A drive and disabled it in the BIOS but that doesn't make a difference. The boot order is C: A:. The system was booting properly from C: just a couple days ago. My motherboard is a Lucky Star UM8498 and the controller is a Pine Tech PT627B.

Any suggestions would be greatly appreciated!

Did you check you not have a overlay/similar etc on mbr?

theelf wrote on 2026-05-31, 19:53:

timdog wrote on 2026-05-31, 19:51:

I have a strange problem that just appeared seemingly out of nowhere with a new 486 DX4 machine I just put together. Regardless of BIOS settings and whether or not a floppy drive is connected to the machine, it always tries to boot A: first. After 15 seconds it fails then if you hit a key to continue it will boot from C: properly. This just started yesterday with really no changes to any hardware that I can think of except adding a 9 pin serial mouse to the setup for the first time. I have a CF to IDE adapter as the primary master C drive with DOS 6.22 installed. No primary slave or secondary drive at all currently. I physically removed the A drive and disabled it in the BIOS but that doesn't make a difference. The boot order is C: A:. The system was booting properly from C: just a couple days ago. My motherboard is a Lucky Star UM8498 and the controller is a Pine Tech PT627B.

Any suggestions would be greatly appreciated!

Did you check you not have a overlay/similar etc on mbr?

I'm not quite familiar with what that is. Seems to be a way to trick DOS into using larger drives? Is there a way to check? I didn't do anything special when I installed DOS. I used the cleanhdd.exe on the CF card first then used the installer to partition and format the 1GB CF card.

I'm still struggling with this. I thought it might be a problem with the Pine 627B VLB controller card but I completely removed it from the system so there are no drives attached and it is still trying to boot from a floppy drive first after about a 20 second wait even though A drive is removed in the BIOS, floppy seek at boot is disabled and boot order is C: A:. It's got to be motherboard related at this point. Is there something wrong with my AMI WinBios? I did a reset on the BIOS today and that didn't change anything.

CMOS battery good?

Can you set it to boot with just C: and not C: and then A?

BinaryDemon wrote on 2026-06-04, 19:08:

CMOS battery good?

I believe so. System time and all settings are all persisting after boot.

Law212 wrote on 2026-06-04, 19:09:

Can you set it to boot with just C: and not C: and then A?

Only options are “C: A:” and “A: C:”

I was having these kind of problems with a computer I have with AMIBIOS, even after resetting the BIOS settings with the jumper. Now I always load the fail-safe settings after playing with the jumper and reboot. After that the problems disappear. I can then also adjust all the other BIOS settings without problems.

AMIBIOS uses the BIOS settings you've entered to decide whether or not to do an ATA IDENTIFY to the drive. This is to be backward-compatible in case you're using an MFM/RLL controller and not IDE. Maybe there is some timing issue?
Since the CF doesn't support multi-sector anyway, it could be that if you've switched it from disabled to some number or Auto.
LBA could also trigger this, but you probably had it enabled the whole time, and if so, it's probably not this but something else.
You could try disabling 32bit, even though it should work fine.

timdog wrote on 2026-06-04, 19:48:

Only options are “C: A:” and “A: C:”

Does sound like CMOS battery -> going to defaults.

If you set it to A: C: or C: A: does it "stick" (try both) - just confirming that CMOS is retaining content.
If you set to floppies none and remove the drives (so it can't detect them) - it should definitely not be trying to boot from floppy.

Remove all cards you absolutely don't need to boot (if mainboard has built in floppy, hard drive, video etc. - use whatever you can)
I've seen add-in cards with "BIOS extensions" which can cause such problems.

Also, re-arrange your RAM sticks - I've seen weird things happen with a "hidden" RAM failure.

Do you have another hard drive to try? - I have seen weird/corrupted hard boot sectors which redirect boot back to floppy (I would expect this to happen even after failed floppy boot - but you never know what things BIOS is recalling as it tries to boot.

Just trying to narrow down where the problem happens.

jakethompson1 wrote on 2026-06-04, 22:00:

AMIBIOS uses the BIOS settings you've entered to decide whether or not to do an ATA IDENTIFY to the drive. This is to be backwar […]
Show full quote

AMIBIOS uses the BIOS settings you've entered to decide whether or not to do an ATA IDENTIFY to the drive. This is to be backward-compatible in case you're using an MFM/RLL controller and not IDE. Maybe there is some timing issue?
Since the CF doesn't support multi-sector anyway, it could be that if you've switched it from disabled to some number or Auto.
LBA could also trigger this, but you probably had it enabled the whole time, and if so, it's probably not this but something else.
You could try disabling 32bit, even though it should work fine.

Multi-Sector Transfer is set to Auto currently. There is no disable. Only auto, 2, 4 and 8. I’ve never changed this value.
32 but transfer has been enabled and disabled and it doesn’t seem to matter.
LBA has been enabled the whole time.

DaveDDS wrote on 2026-06-04, 22:28:

Does sound like CMOS battery -> going to defaults. […]
Show full quote

timdog wrote on 2026-06-04, 19:48:

Only options are “C: A:” and “A: C:”

Does sound like CMOS battery -> going to defaults.

If you set it to A: C: or C: A: does it "stick" (try both) - just confirming that CMOS is retaining content.
If you set to floppies none and remove the drives (so it can't detect them) - it should definitely not be trying to boot from floppy.

Remove all cards you absolutely don't need to boot (if mainboard has built in floppy, hard drive, video etc. - use whatever you can)
I've seen add-in cards with "BIOS extensions" which can cause such problems.

Also, re-arrange your RAM sticks - I've seen weird things happen with a "hidden" RAM failure.

Do you have another hard drive to try? - I have seen weird/corrupted hard boot sectors which redirect boot back to floppy (I would expect this to happen even after failed floppy boot - but you never know what things BIOS is recalling as it tries to boot.

Just trying to narrow down where the problem happens.

BIOS settings seem to be sticking when I change them as far as I can see. However, I did notice something strange with LBA mode on the primary master (see pictures). I turned of LBA for primary master once and noticed it was set ON during startup. Maybe this is a sign of something strange?

I removed all cards except the video card and no change. I don't have IDE or floppy connectors on the motherboard.

I only have a single 16MB RAM stick.

I don't have another HDD to try but I do have other CF cards for my CF to IDE adapter so I formatted and installed a clean DOS on one and that didn't change anything either.

It feels like the BIOS to me. For some reason certain settings aren't actually being changed.

Just had to resurrect this thread for some closure. I had been using a Cirrus Logic 5420 video card as a temporary solution and just recently swapped it with a new Trident card. Once I did that, the weird booting from A: issue went away! Crazy stuff. Happy things are working properly now.

Thats not a closure at all!
Especially the "This just started yesterday with really no changes to any hardware that I can think of" is hyper weird.
Did you also change CF cards/IDE adapters by any chance? CF card/adapter forcing IDE pin 28 to ground will mess with some but not all ISA graphic cards.

That’s interesting, if you swap back the problem re-appears?

rasz_pl wrote on Yesterday, 07:28:

Thats not a closure at all!
Especially the "This just started yesterday with really no changes to any hardware that I can think of" is hyper weird.
Did you also change CF cards/IDE adapters by any chance? CF card/adapter forcing IDE pin 28 to ground will mess with some but not all ISA graphic cards.

I haven't changed CF/IDE adapters. It is really weird for sure. It did start kind of out of nowhere after not touching the machine for a few days. The only thing I can remember was hooking up a new mouse which didn't work.

BinaryDemon wrote on Yesterday, 11:39:

That’s interesting, if you swap back the problem re-appears?

Maybe I can give that a try next time I get a chance.