VOGONS


First post, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t

I'm not going to get into a ton of detail here because it looks like an issue that others have documented well over the years.

Basically, I have recently run into multiple S3 Virge cards from different brands that work fine in DOS and in Windows without drivers installed. After drivers are installed in Windows 98SE, the system will show the desktop and work fine for 10-15 seconds and then it just freezes solid, requiring a forced shut down.

I am pretty sure that I've seen this in the past with other variants of the Virge, but the particular cards I am dealing with at the moment are all the original Virge 86C325 models. The cards are all completely different as well... STB, ExpertColor, Biostar... some have electrolytic caps, but some do not have any at all so there are effectively no age-sensitive components on them. I have other Virge cards (DX, GX and others) that work fine in this same system with the same Windows install.

System specs:
Wintec W6BXA 440BX
Pentium III 850Mhz (100Mhz FSB) in Slotket (also tried a Pentium II 266Mhz with no change)
128MB SDRAM
IDE to MSATA adapter with 16GB MSATA drive
Windows 98SE (have tried multiple different fresh Windows folders from a backup with the same result)

I have found several other threads where people have had similar issues in the past. These guys know their stuff and have done all the same steps I have done (and more) with the same results:
Windows 95 - S3 Virge Issues
Re: Virge 325 locks up Win98 after drivers installed
Various games freeze in Windows 95

So, is there a really obscure compatibility issue with some Virge cards? Since they aren't worth much and are common it wouldn't surprise me if this has been an issue for a while and people just haven't bothered diagnosing it. I had written the cards off as bad until my "not working" pile contained mostly S3 Virge cards and I got suspicious.

I will try them in some other systems and post my findings here when I can.

Now for some blitting from the back buffer.

Reply 1 of 1, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t

Well, this is pretty weird.

I found 2 more Virge cards in my collection which I had tested 2-4 years ago (in this system). Both are Diamond Stealth 3D 2000 (one full size 2+MB, one smaller 2MB) and were found to be working perfectly last time. I just tested them again and they are still fine. I noticed they were using the default Windows 98SE Virge driver rather than the one from S3 that I had downloaded. I then tested the STB Virge card I had tested earlier. It seemed like it was working longer than normal so I went to device manager and saw it was using a different driver (one from S3), so I removed the device (hoping to try the other driver). I didn't reboot right away out of curiosity, and it never crashed. I rebooted, chose the default driver and it seems to be working totally fine now. -_-

So, I thought maybe there is something driver related going on after all, even though I have been through this before and have used a fresh copy of the Windows folder with a fresh registry and drivers. I went into regedit, removed any of the devices that referred to the driver from S3, and left any that used the default driver.

Then I put in the generic card (I thought it was Biostar, but it actually seems to be a different manufacturer), started it up, Windows detected it as new hardware and set it up with the default driver. I rebooted, everything looked normal and it seemed like it was lasting longer than before... then it froze. -_-

I also tried the ExpertColor and it is not freezing anymore. -_-

Anyway, I have no idea what was going on before. I forgot to mention that the STB and ExpertColor cards are ones I have had for years and were tested on this system and found to be and working as of 2020-2021. The problems seemingly all started when I got the generic Virge card a few months back and it was freezing despite having no other obvious flaws. To rule out a compatibility issue I tested the "known working" STB and ExpertColor cards and they both froze the same way, which started me down this rabbit hole.

Now, it looks like I have four working and one freezing, inexplicably. I guess that one is most likely just dead. But I have no idea why the issue seemed to temporarily affect the others I tested though. How strange.

... in other news, I did stumble across something else that was strange about these things. All of the cards currently have 2MB installed and display 800x600x24bit and 1024x76x16bit perfectly fine. Two of the working cards (ExpertColor and the larger 2+MB Diamond) give a sort of messed up and shaky\flashing (but still somewhat recognizable) display at 1280x1024x8bit. Why is there a difference between the cards in this respect? It seems weird, especially with the Diamond cards since they are the same brand. The working Diamond has BIOS version 1.01, the one with the garbled output at 1280x1024 is version 1.04. I tried swapping the ROMs and the problematic one still does exactly the same thing, so it doesn't appear to be VBIOS related.

This testing was done on a Dell 2001FP (4:3, 1600x1200 native res).

Sorry for the meandering thread. I seem to keep putting these poor cards on the back burner every time I run into issues and I figured I should make a post about it this time. 😜

Now for some blitting from the back buffer.