Gateway 80486-DX2/50 Revival

[eccerr0r]

Re: Gateway 80486-DX2/50 Revival

ugh...can't read my old cd-r's anymore... sigh. bitrot...

[TechGeek]

Re: Gateway 80486-DX2/50 Revival

Tried Uniflash or Flashrom? I have used those on quite a number of mainboards (granted, most were 430VX/TX, VIA 580VPX and newer) and they usually work. Flashrom has quite a higher rate of good flashes - I've had cases where Uniflash didn't want to play nice.

Uniflash said that their software is intended only for PCI-era chipsets. I have not looked at Flashrom yet.
EDIT: Flashrom doesn't appear to have any mention of ISA chipsets, so I'm ruling on the safe side (not compatible).

[Dan81]

Re: Gateway 80486-DX2/50 Revival

Tried with external battery power, no change. The BIOS sees it as an IDE drive and forcibly resets it to auto-detection. Unfortunately, the BIOS also has halfway-implemented CHS translation. Ontrack Disk Manager actually fails to install due to the bugged BIOS, either with "Failed to install because of a severe error condition!" or a warning message saying that "This System's BIOS Does Not Fully Support Translation!"


There are newer BIOSes available, but I haven't found a flash utility that I feel comfortable running against it, and I don't have an (E)EPROM programmer.

Tried Uniflash or Flashrom? I have used those on quite a number of mainboards (granted, most were 430VX/TX, VIA 580VPX and newer) and they usually work. Flashrom has quite a higher rate of good flashes - I've had cases where Uniflash didn't want to play nice.

[stj]

Re: Gateway 80486-DX2/50 Revival

your supposed to install the chip driver first.

[TechGeek]

Re: Gateway 80486-DX2/50 Revival

what dos?
did you try freedos?

what chip is on the card? i can probably find the correct driver

MS-DOS 7 (hangs when loading ASPI drivers with the card installed). Can't determine the chip installed on it. (It's got this label on it that I can't remove. I could try to find the mfg. by looking up the MAC address.)

[stj]

Re: Gateway 80486-DX2/50 Revival

I do, but it's that ISA clone-of-a-clone card that causes DOS to hang up. I'll need an actually decent network card to make that happen.

what dos?
did you try freedos?

what chip is on the card? i can probably find the correct driver

[TechGeek]

Re: Gateway 80486-DX2/50 Revival

That's not making a whole lot of sense in relation to eachother. Did you replace the batt with something/anything that will make it retain memory (bug1)?

Bug2 would be technically negated by bug1, and that sounds like more of an issue with the IO controller. Try a different one? If these are known bugs, probably a BIOS update out there somewhere for it....

I wouldn't waste any time trying to hunt down a sub-528mb HDD; especially one that's actually GOOD.... Just get a 512mb CF card with an IDE adapter....and done.

Tried with external battery power, no change. The BIOS sees it as an IDE drive and forcibly resets it to auto-detection. Unfortunately, the BIOS also has halfway-implemented CHS translation. Ontrack Disk Manager actually fails to install due to the bugged BIOS, either with "Failed to install because of a severe error condition!" or a warning message saying that "This System's BIOS Does Not Fully Support Translation!"


There are newer BIOSes available, but I haven't found a flash utility that I feel comfortable running against it, and I don't have an (E)EPROM programmer.

[Topcat]

Re: Gateway 80486-DX2/50 Revival

Nope, it's a pair of bugs in the system BIOS that interact in a bad way.

Bug #1- BIOS reverts hard disk configuration to auto-detection on boot, cold or warm.
Bug #2- BIOS tries to implement CHS translation for >528MB hard disks, but doesn't implement it fully and the end result is spectacular corruption of whatever you write to the disk, not to mention that whatever you read from the disk also ends up getting thoroughly trashed if you wrote it to the disk before transplanting it into the 486 system.

So either I'll need a sub-528MB hard disk (in the works) or an add-in BIOS that completely bypasses the (critically broken) INT13h routines when a >528MB hard disk is installed.

That's not making a whole lot of sense in relation to eachother. Did you replace the batt with something/anything that will make it retain memory (bug1)?

Bug2 would be technically negated by bug1, and that sounds like more of an issue with the IO controller. Try a different one? If these are known bugs, probably a BIOS update out there somewhere for it....

I wouldn't waste any time trying to hunt down a sub-528mb HDD; especially one that's actually GOOD.... Just get a 512mb CF card with an IDE adapter....and done.

[TechGeek]

Re: Gateway 80486-DX2/50 Revival

You've done something wrong then.....

Nope, it's a pair of bugs in the system BIOS that interact in a bad way.

Bug #1- BIOS reverts hard disk configuration to auto-detection on boot, cold or warm.
Bug #2- BIOS tries to implement CHS translation for >528MB hard disks, but doesn't implement it fully and the end result is spectacular corruption of whatever you write to the disk, not to mention that whatever you read from the disk also ends up getting thoroughly trashed if you wrote it to the disk before transplanting it into the 486 system.

So either I'll need a sub-528MB hard disk (in the works) or an add-in BIOS that completely bypasses the (critically broken) INT13h routines when a >528MB hard disk is installed.

do you have a network card with a socket for an eprom on it??
if you do then i can point you to an eprom that gives you an LBA48 ATA driver extension!!

I do, but it's that ISA clone-of-a-clone card that causes DOS to hang up. I'll need an actually decent network card to make that happen.

[eccerr0r]

Re: Gateway 80486-DX2/50 Revival

The Weitek was not a 487 or a 486, that was something different. It's not a "performance penalty" rather it's simply slower due to the lower clock rate than the higher clocked 486DX2 or 486DX4.

[Topcat]

Re: Gateway 80486-DX2/50 Revival

haha, i like your antique 486 system already, because it has two components made in singapore, the wd hard disk and the creative sound blaster sound card. seems singapore was quite the IT manufacturing factory back during the day, then it all went away to other countries. same boat as u guys in the US but since this is not the vip room, i wont comment further.

I've also seen a lot of tech that era with 'made in usa' on it as well; quite common then....but not now....but I'll leave it at that.

[ChaosLegionnaire]

Re: Gateway 80486-DX2/50 Revival

Temporary hard disk. Due to BIOS bugs, it doesn't work properly in this system. In fact, no hard disk above 528MB will work with this system.

Sound Blaster Vibra/16 card.

haha, i like your antique 486 system already, because it has two components made in singapore, the wd hard disk and the creative sound blaster sound card. seems singapore was quite the IT manufacturing factory back during the day, then it all went away to other countries. same boat as u guys in the US but since this is not the vip room, i wont comment further.

[stj]

Re: Gateway 80486-DX2/50 Revival

do you have a network card with a socket for an eprom on it??
if you do then i can point you to an eprom that gives you an LBA48 ATA driver extension!!

[Topcat]

Re: Gateway 80486-DX2/50 Revival

Disk managers are a no-go. Multiple bugs in the system BIOS that combine to impose a hard 528MB limit.

You've done something wrong then.....

[TechGeek]

Re: Gateway 80486-DX2/50 Revival

If you want to make a bigger drive behave on this, you can use an overlay; like Ontrack Disk Manager; with DOS or NT it would be fine up to 2gb @ the point of the install. With NT4, you could go even bigger but the install partition can't be any larger than 2gb. Once installed, you can partition the rest from within the OS. I've used this on stuff as far back as 386's that pretty much had zilch for HDD support.

Disk managers are a no-go. Multiple bugs in the system BIOS that combine to impose a hard 528MB limit.

[Topcat]

Re: Gateway 80486-DX2/50 Revival

If you want to make a bigger drive behave on this, you can use an overlay; like Ontrack Disk Manager; with DOS or NT it would be fine up to 2gb @ the point of the install. With NT4, you could go even bigger but the install partition can't be any larger than 2gb. Once installed, you can partition the rest from within the OS. I've used this on stuff as far back as 386's that pretty much had zilch for HDD support.

[TechGeek]

Re: Gateway 80486-DX2/50 Revival

Windows? What's Windows?
The K6-233 is running Linux (and DOS...)

BTW seems a lot of older 486 boards have a coprocessor socket for 487's... which only makes sense if the main processor socket had a 486SX in it. Of course the 487 was an entire 486DX and simply disables the 486SX.

What would have been fun is that the board was SMP capable and ran two 486 integer cores and the FPU of the second CPU...alas, nope.

Not an entire 80486 in the upgrade chip, not to mention that it imparts a performance penalty when installed in tandem with an 80486-DX2. Looks like that upgrade socket is staying empty.


https://www.cpu-world.com/CPUs/4167/index.html

[TechGeek]

Re: Gateway 80486-DX2/50 Revival

Just photos with this post.

Starting off with a shot at the upper half of the case.


Temporary hard disk. Due to BIOS bugs, it doesn't work properly in this system. In fact, no hard disk above 528MB will work with this system.


Working I/O card pulled from the cheap-and-nasty 286 clone system.


A better view of the extent of the corrosion. Battery long gone by this point.


Damaged I/O card, not cleaned up.


Here's the cheap "NE2000-compatible" clone-of-a-clone network card that doesn't work with hardly anything I can throw at it.


Sound Blaster Vibra/16 card.


Lugged it home last weekend. Here's it all set up, running SpinRite 6.0 because I suspected hard disk problems.


As I said, this large of a drive just doesn't work on this system. This was the end result after trying to install MS-DOS 6.22.


And this is the end-result of trying to run Windows 98 setup because why not.


That's all I've got for now. More to come later.

[eccerr0r]

Re: Gateway 80486-DX2/50 Revival

Windows? What's Windows?
The K6-233 is running Linux (and DOS...)

BTW seems a lot of older 486 boards have a coprocessor socket for 487's... which only makes sense if the main processor socket had a 486SX in it. Of course the 487 was an entire 486DX and simply disables the 486SX.

What would have been fun is that the board was SMP capable and ran two 486 integer cores and the FPU of the second CPU...alas, nope.

[TechGeek]

Re: Gateway 80486-DX2/50 Revival

I have an old gateway p5/90 case that looks very similar to this 486dx2-50...full size AT case, board is long gone. It has a K6-233 (with 256MB SDRAM on a 430TX...) board in it now.

Alas no nostalgic value to me, it's still a bit too fast to run my old EPROM burner software...

"Black to black, back to back" ... if they didn't cut out the right keys on the board and plugs.

Watch out and make sure that you don't end up with CIH on that thing, as the Intel 82430 series of chipsets *WILL* be exploited by CIH, and the end result is a trashed BIOS. Or make sure that you're running an NT-flavored OS, as CIH's BIOS-wipe payload doesn't work on Windows NT.


I didn't even know that some boards had keying, not that I'd trust it anyways. Black to black and you'll be back! Red to red, motherboard dead!

I read your reply, but ATM, just have time for a quick reply...

re: power plugs. If the reds were together, +5 was grounded via the chassis, if metallic motherboard mounting was intact. Some of those old ones used nylon standoffs & fiber washers, so it wouldn't always trip the power supply- in which case "motherboard dead." And even if the short on +5 shunted that rail, the other voltages could still damage stuff during the brief period they "twitch."

re: HDDs- Maxtor 7xxxAT series were OK, and period correct.
My first real PC had a 7245AT- came to 230ish MB after formatting.
Great spinup sound, too!
The sound that meant "we're firing up the PC & doing serious work."
And they had "artwork" on the PCB; each capacity had its own "feature," IIRC.


ST3660A was OK too, had one of those alongside the Maxtor.
Was interesting setting the jumpers- these weren't simply "master-slave," etc.
You should be able to find charts to jumper them for operation on the same cable.

The drive that I'm looking at right now is a ~425MB Seagate drive from late-'94/early-'95. It reports a CHS geometry of 899/15/62, so it should work in this system.

That socket next to the CPU might be for a Weitek FPU. I don't know the best place to find information about these old motherboards.

According to THIS link, the socket is for a Weitek FPU. Not sure what kind of performance benefit that it would have, given that the system already has a full-fledged DX-series 486.

It seems weird that the I/O card would have its own battery, but I would guess that it was replaced at some point with whatever was available.

Damage from the motherboard battery. The same battery that inflicted damage on the 1st ISA slot.

If you get around to cleaning the corrosion off that motherboard and removing the model rocket igniters, you can still buy new ISA slots. I don't think it will last too long with that kind of corrosion, but at least it works for now. That's an improvement over some of the leaking battery repairs I've seen.
https://www.digikey.com/en/products/...WN-S420/927281

Battery got clipped off not long after I found it. I cleaned up the damaged ISA card, and if it's anything like the motherboard, it should clean up with little to no collateral damage, ISA slot included. I've tested it, and while the slot still works, it's flaky. Still has lots of tension left in the contacts, so I think it'll pull through. If it ends up destroyed during the cleaning process, I'll probably have TC drop a new one in.


P.S.: Eaten trace is what seems to have killed the I/O card. It and the ISA slot took the brunt of the damage, as the system was stored upright for all of its life. I don't think that I'm going to risk pulling the 8042 keyboard controller, unless it's an EPROM (the UV-windowed things). I've powered the external battery connector off of an external supply and it holds time and settings fine, which is a good sign. I'll lug the beast home this weekend and clean the motherboard up and put an end to the corrosion.