VOGONS


First post, by Fläskboi

User metadata
Rank Newbie
Rank
Newbie

Finally decided to put the free 486 together, but it won't let me access devices and use it! Here's the situation:

- FDD won't work.
Different drives and cables from known working FDD's have been tested. Boot floppies were scandisked, repaired, thrown away and remade, confirmed working in more than one other PC. Tried all 4 possible cable directions, where the one working behaved differently depending on the diskette: non-system disk resulted in the message "improper boot device", and system disk lead to no message at all (and also, nothing more happening).

- HDD won't work.
Tried different HDD's - a known working Connor 2GB w/ Win98, that the 486 didn't recognize at all, and a Seagate 3.2GB, that it auto-detected and most times (but not always) found in BIOS auto-detect. Installing Win95 on the Seagate using another PC, I swapped it back at first restart prompt, and also after full install. Both times resulted in "invalid system disk, replace the disk and press any key". Pressing Enter resulted in short rattle from the disk, so it was obviously found and waiting. HDD was freshly formatted and fully scanned with no bad sectors, and IDE cable was replaced and redirectioned with no improvement to report (just the message "HDD controller failure" when put in the wrong direction).

The FDD, HDD and some other stuff is connected to a controller card, alongside an ISA riser on the board. Another riser slot has been tested, with no different result. The graphics card attached to the riser appear to work fine.

Together with initial overheating issues, freezes and suspected PSU/grounding problems, I've spent over 10 hours troubleshooting this rig, and taken four computers apart trying to solve this bullsh**. I don't get past this problem... Is the HDD controller toast? Is there a reason the HDD's are too large, even though they're found and formatted as FAT? Are jumpers improperly set, despite not even having IDE jumpers from my understandings? Is the ISA slot bad, although GPU works fine? Board is Elitegroup UC4914-G.

Thanks 😊

Reply 1 of 14, by weldum

User metadata
Rank Member
Rank
Member

about the hdd problem, it may be an incompatibility with the bios, most likely because of the 504mb bios limit, you should use an overlay software like ezdrive, or use the xt ide bios on a nic

Ohh, the humanity 😢
386SX 25-C3 800-P3 900-Atom 1.6-C-D 2.13-P4 2.26-P4HT 3.0-P-DC 1.6-AFX 2.6-FX 3.3

Reply 2 of 14, by The Serpent Rider

User metadata
Rank l33t
Rank
l33t

most likely because of the 504mb bios limit

Late VLB boards can understand 8gb.

Get up, come on get down with the sickness
Open up your hate, and let it flow into me

Reply 3 of 14, by weldum

User metadata
Rank Member
Rank
Member
The Serpent Rider wrote:

most likely because of the 504mb bios limit

Late VLB boards can understand 8gb.

really? i had some 486 mobos that had pci and didn't recognized 1gb hdd

Ohh, the humanity 😢
386SX 25-C3 800-P3 900-Atom 1.6-C-D 2.13-P4 2.26-P4HT 3.0-P-DC 1.6-AFX 2.6-FX 3.3

Reply 4 of 14, by Fläskboi

User metadata
Rank Newbie
Rank
Newbie
weldum wrote:

about the hdd problem, it may be an incompatibility with the bios, most likely because of the 504mb bios limit, you should use an overlay software like ezdrive, or use the xt ide bios on a nic

The Serpent Rider wrote:

Late VLB boards can understand 8gb.

Thanks for the quick responses 😊 I also thought of this, however I find it strange that BIOS correctly utilizes those 3099 MB and then can't boot from them. In a thread at Stackexchange, they say it apparently doesn't matter whether the BIOS supports LBA or not; DOS 6.22 won't use it anyway. Might just try to remove the two partitions (1.05GB and 2GB), set manual parameters, format a 504MB partition on the newer IBM and see what happens.

I'll post back with my findings 😀

P.S I think the board is rather old, according to manual the fastest it'll support is DX2/66. Plus that FDD interface doesn't work, meaning we're screwed loading drivers for a NIC 😒, although it would be a great solution.

Reply 5 of 14, by Intel486dx33

User metadata
Rank l33t
Rank
l33t

Does it have a NEW battery. Old batteries can cause havoc on the bios.
Also reset the ram and cache chips.
Bad IDE cables and floppy drives and cdroms.

Follow the Process of elimination.

Just connect the minimal at first
4 memory Simms. ( FPM ) in bank 0
IDE controller with all ports disabled except the mouse serial-1 and IDE channels and floppy drive.
Reference your motherboard manual for correct IRQ and DMA settings.
Hook up a video card.
Hookup a floppy drive 1.44
Hookup a hard-drive with jumpers set to Master/Single or Cable select.

Choose to boot from A,C drives in bios.

Hook up the CDROM drive only after you get these working.

Reply 6 of 14, by torindkflt

User metadata
Rank Member
Rank
Member

Have you verified that the cache settings are correct in the BIOS or jumpers? Also might be worth checking that the cache (if there are any) is even real, since Elitegroup AKA ECS is/was a subsidiary of PC Chips, which was notorious for releasing scammy ultra-cheap 486 motherboards with fake cache.

Reply 7 of 14, by derSammler

User metadata
Rank l33t
Rank
l33t
Fläskboi wrote:

I also thought of this, however I find it strange that BIOS correctly utilizes those 3099 MB and then can't boot from them.

I had a lot of 486s (and even one 386) that detected HDDs up to 8 GB, even though the controller wasn't able to support more than 504 MB. You normally need an E-IDE controller with it's own BIOS for this to work correctly, or use a modern IDE firmware like XT-IDE. Stock BIOS and a 16-bit ISA controller probably won't work. However, you first need to sort out your floppy issues. Try swapping the controller card.

Reply 8 of 14, by Intel486dx33

User metadata
Rank l33t
Rank
l33t

Newer IBM smart drives are not supported with old bios.
Also once you get it working fdisk and format the C:\ and install DOS 6.x
and run MSD and check your IRQ and DMA settings. Look for conflicts and FREE resources. You can take screen snapshots with you phone or tablet so you don't have to write everything down.

Reply 9 of 14, by mdog69

User metadata
Rank Newbie
Rank
Newbie

If you've fitted a new battery, don't assume that the BIOS has filled the CMOS with useful default values. If you can, boot into setup and then look for a "load default values" option. Do that, save and reboot then setup as you see fit.

Reply 10 of 14, by Anonymous Coward

User metadata
Rank l33t
Rank
l33t
derSammler wrote:
Fläskboi wrote:

I also thought of this, however I find it strange that BIOS correctly utilizes those 3099 MB and then can't boot from them.

I had a lot of 486s (and even one 386) that detected HDDs up to 8 GB, even though the controller wasn't able to support more than 504 MB. You normally need an E-IDE controller with it's own BIOS for this to work correctly, or use a modern IDE firmware like XT-IDE.

Uh...if you need a new controller with its own BIOS, then your board doesn't support 8GB. I've seen many boards that can detect large drives that don't actually support them. I don't think there's anything special about "EIDE" controllers. All the magic is in the system BIOS.

"Will the highways on the internets become more few?" -Gee Dubya
V'Ger XT|Upgraded AT|Ultimate 386|Super VL/EISA 486|SMP VL/EISA Pentium

Reply 11 of 14, by Fläskboi

User metadata
Rank Newbie
Rank
Newbie

Great news! Beauty's all up 'n running! 😊

Messed a bit with the jumpers, and noticed I had misinterpreted the manual, by connecting the power LED to ADS delay (whatever that might be). After correcting this, the FDD jumped to life, although a boot floppy had to be sacrificed 😒

Sadly, it wasn't much help for the HDD. Using the now working FDD, W95 found its way to two different disks. No luck, still "invalid drive" on boot. A total of 4 jumper settings on the HDD controller were possible, of which the previously untested all gave the "HDD controller failure" error. Gave up and went to sleep right there.

After sleeping on the problem, I realized we had it quite well surrounded. With both HDD's flawlessly completing Win installations, at least until first reboot; the HDD, cables and controller could be pretty much excluded from the tree of troubles. Also, the ancient Conner drive showed different errors depending if Windows was installed or not. Ideas left were colliding IRQ's, outdated BIOS, Windows incompatibility with the system, or the simple fact that I used a Windows Millennium startup floppy to format the disks, resulting in "Disk is compressed or there's a disk cache" warning message prior to Win95 installation (probably because of LBA; Win95 fdisk can't see more than 504MB as mentioned, whereas ME fdisk could see and partition the whole disk).

Attempting the last first, I went ahead and created a fresh Win95 boot floppy, used that to fdisk the partitions, and let the Windows installer initialize the drive. Without hoping too much, it turned out to work like a champ! So happy to have my first 486 in order 😊 Thank you for all the help here, finally time to get doomed 😎

BXRvCht.png

Reply 12 of 14, by Fläskboi

User metadata
Rank Newbie
Rank
Newbie
derSammler wrote:
Fläskboi wrote:

I also thought of this, however I find it strange that BIOS correctly utilizes those 3099 MB and then can't boot from them.

I had a lot of 486s (and even one 386) that detected HDDs up to 8 GB, even though the controller wasn't able to support more than 504 MB. You normally need an E-IDE controller with it's own BIOS for this to work correctly, or use a modern IDE firmware like XT-IDE. Stock BIOS and a 16-bit ISA controller probably won't work. However, you first need to sort out your floppy issues. Try swapping the controller card.

Seems like it's most meaningful to run the 540MB Conner for the OS then. Then buy a gotek for games 😀

Reply 14 of 14, by amadeus777999

User metadata
Rank Oldbie
Rank
Oldbie

There's the possibility that the board/connector or cable is slightly damaged. I had this 2 days ago. After connecting a floppy disk drive to a working system the ide ports would still "work"[1] but I couldn't boot - no matter which size the CF was. After havin tried many options I pressed onto the connectors and voila all was working well.

[1]The CF, which were working before, were detected, fdisk could be run and disks formatted, but bootup gave "read error from disk" or something similar.