VOGONS


First post, by Biscuit Shaped

User metadata
Rank Newbie
Rank
Newbie

Hi all!

I'm building a retro pc from the bones of three machines from the early OO's and a few bits I recently found in the garage. I'd rather give too much info than too little, so if you don't care for the specifics, just page down a bit to the stupid questions...

Motherboards:
1: Shuttle AK37GT (non-RAID, so no SATA, just IDE) with the KT400 chipset "blue"
1: Biostar M7VIG-Pro , also w KT400 chipset "red"
1: Soyo SY-K7-ADA , Ali 1647 Magick chip set "brown"

Video Cards:
1: MX440 64MB
1: FX 5600 128 MB
1: 4200ti 128MB
1: Voodoo3 3000
1: Matrox IS Athena R2

Power supply: an otherwise generic unit that is *supposed* to have 50A on the 5v bus and I think 20 or 25 on the 12. The ad specifically stated that it was recommended for Socket A boards.

And a whole pile of assorted memory, mostly PC2100 and PC2700.

As a start, I'm setting up the AK37GT board with the FX5600 and 256M of PC2700, and I installed - or am installing - Win 98SE

First, I think I installed the nVidia driver first (45.23 iirc) and the machine kept hanging on the restart - at the pont where it was building the driver database for new hardware.

A couple restarts later - going in through safe mode and not loading drivers, then restarting again, I was able to get windows to boot up fully, but it got stuck in a loop saying that 640x480x16 is not available on your current hardware blah blah blah.

After a couple rounds of that I switched in the MX440, reinstalled Windows, and loaded the nvidia driver ver 30.82. That went through perfectly and I finally got a reasonable resolution and color depth.

Then I loaded the motherboard drivers (so Question 1: is batting out of order causing any problems here with the FX5600/45.23 driver that the MX440 isn't subject to?) and then the video drivers again, the MX440 continued to work fine. So I switched in the 4200ti and it was malfunctioning... dots and colons in spaces that should be blank, text running into each other and so on. It booted into Windows, but there was trash everywhere. (Question 2: does this sound like a 20+ year old board problem? any thoughts?). Then I put in the FX5600 again and once again it said 640x480 couldn't run and put me n the restart loop again. Grrrrr!

So do these problems sound like driver-out-of-order issues, or do you think two of my video cards are poop? Possibly a mobo issue? I haven't tried the other 2 boards but I'd prefer to use the Shuttle as I used to have that board years ago, oh, and I tried all three machines befor I tore into anything and only the M7VIG machine started, though I attributed that to bad power supplies on the other two.

Is it worth re-installing Win98 again, installing the chipset drivers first and then the nvidia drivers? Do you think that would make a difference?

The point of this build is mostly for 2d games, so whichever card I wind up using is probably more than sufficient, but there are a few old 3D games that I'd like to try out again: Hit Man, GTA maybe

Reply 1 of 5, by RandomStranger

User metadata
Rank Oldbie
Rank
Oldbie

In case of the FX5600 it could be bad caps. I had the same problem with my FX5700 on my XP test board. It ran fine with the Windows built in drivers, but once I installed the Nvidia drivers, it did the same thing as your 5600. After recaping, it works fine.

It's also common around here to recommend installing motherboard drivers first, and then the graphics card drivers.

sreq.png retrogamer-s.png

Reply 2 of 5, by Biscuit Shaped

User metadata
Rank Newbie
Rank
Newbie
RandomStranger wrote on 2023-05-11, 15:50:

... but once I installed the Nvidia drivers, it did the same thing as your 5600. After recaping, it works fine...

Could that be because the video card has one set of physical circuits to handle basic minimal tasks, and other dedicated circuits (the GPU?) that only come into play when higher commands are given? If the bad capacitor, or whatever else it may be, are integral to the higher commands circuit then we wouldn't see any problems until the drivers are in and we're asking for more from the card?

Sorry if that sounds insultingly rudimentary question, but I'm a complete zilch when it comes to circuit level electronics - or pretty much any other level for that matter.

Thanks for your advise, by the way!

Reply 3 of 5, by paradigital

User metadata
Rank Oldbie
Rank
Oldbie

It’s more likely that at idle or basic 2D tasks, the strain on the voltage rails is minimal, and the weak or failing capacitors aren’t particularly required for stable voltage rails.

Add some load into the mix and the bad caps will be causing voltage ripple and thus instability.

Reply 4 of 5, by Biscuit Shaped

User metadata
Rank Newbie
Rank
Newbie

Ok, so I checked the capacitors on the video card and they were all ok, actually they were really good. I haven't checked the caps on the motherboard, perhaps they're the culprit? That is where the card gets its power from.

Reply 5 of 5, by Minutemanqvs

User metadata
Rank Member
Rank
Member

Don't forget that to really check if a capacitor is OK you have to remove it from the circuit and measure it, which is a real hassle. Visual inspection for leaks is a good start but not a 100% guarantee that the capacitor is ok.

Searching a Nexgen Nx586 with FPU, PM me if you have one. I have some Athlon MP systems and cookies.