Spent a while trying to figure out an issue with my hard disk that first cropped up a couple days ago. Upon power up or rebooting, my 486 would sometimes say it could not find any system disk. Attempting to either reset or power off and wait a bit before trying again would change nothing.
When it first happened, I just assumed that the hard disk had gone kaput. On a whim, I tried unplugging and re-inserting the IDE cable and then booting the PC up again. For whatever reason now worked fine. Strange. However, the same issue happened again later. Decided to this time try the BIOS' IDE auto detect and see what came up (the BIOS previously had no problems doing this). Detected a drive on the correct IDE channel but with junk values that were quite off. Not entirely unexpected if the system is complaining that it can't find a system disk to boot from.
I'm using a Promise Technology 560C VLB I/O controller card that has two IDE channels + floppy. I had the hard disk connected to the primary (VESA bus) channel and the CD-ROM connected to the secondary (ISA) channel. Both drives configured as master. I had been using this exact setup with the same motherboard, IO card, drives and IDE cables with no issues since I got the IO controller card in early September. As a result, I again was suspecting that the hard disk was maybe not quite totally dead yet, but was perhaps on it's slow way out.
Swapped in another drive to test and it worked, but a couple reboots later it started to not detect this hard disk either. Argh! Ok, maybe the drive isn't dead / dying after all? Maybe the problem is with the IO card. Luckily I had one other IO controller card on hand to try out, a UMC SST-2946/8 VLB card. Had a little bit of a problem at first with this card. Even though it also has two IDE channels, my previous setup with both drives on their own channel (and cables), both configured as master did not work. Had to set the CD-ROM to slave for it to boot successfully. Then it pretty much was the same story as before... first couple of boots, it all worked fine. Eventually the same issue did crop up again. I also tried without the CD-ROM connected at all and eventually hit the same issue with that configuration also. Ugh. 🙁 Tried a different IDE cable too, but no luck there... eventually showed the same issue.
Decided this time to go back to the Promise Tech IO card and bypass the VESA bus entirely for the IDE devices (kind of wish I had a plain ISA controller card right now, but unfortunately I do not). The card has a jumper configuration printed on the card itself that seems to clearly indicate that it sets the ISA channel as the primary (the UMC card probably has this configuration too, but it's not as obvious to me from looking at the labels printed on the card itself and I cannot find a manual for either of these cards). Configured the CD-ROM to slave and connected both it and the hard disk to the same IDE cable.
So far -- knock on wood -- this configuration has been working... at least I've so far gone much longer (and through more reboots) without the issue reappearing then I had been able to before. Will have to give it a couple more days to see for sure I guess. However, I am quite a bit uneasy about the fact that this problem just spontaneously occurred without any other hardware changes. I guess the only thing I've not tried replacing so far is the motherboard, currently a FIC 486-PVT. I guess this is potentially a good excuse to finally try the NOS Eagle G486SLV board I got a few weeks ago that I've been really lazy about trying. Also, will need to pick up a regular ISA IO controller card and maybe another IDE cable in another attempt to rule out any problems there.