VOGONS


First post, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t

This is kind of puzzling. I have a Viewtop S3 Trio64V+ which is unique in that it has 8 SOJ-40 sockets. This makes it a good card for testing memory chips, and an excellent all-around DOS\9x 2D card.

However, on my main test system which uses a Wintac/WinTech/Edom W6BXA 440BX board, this card will just lock the system up after being in Windows with the drivers installed after about 30 seconds. I can put the card into a Via MVP3 SS7 system with a K6-2 500Mhz and it works fine using the same OS and drivers. I have tried different Windows installations on the 440BX too, and it makes no difference. This is with only 2MB memory installed on the card, and I have tried it with different memory chips. There are no other PCI or ISA cards installed in either system. I have tried using it on an LCD and a CRT on both systems, and the results are always the same... totally fine on the SS7, but locks up on the 440BX (requiring me to hold the power button to shut down).

Has anyone ever run into this with an S3 Trio before? They are normally so easy to get running, this just seems weird. I have an STB Powergraph S3 Trio64V+ 2MB that I can drop into the 440BX system and it works exactly as expected. I can also run a huge variety of other cards in this board with no issues.

Maybe it's a BIOS issue? I have attached the BIOS file I dumped from the card. It does have Viewtop branding in the BIOS, so it isn't 100% a generic S3 BIOS, but it is actually more up to date than the STB Powergraph BIOS.

There is one slightly more up to date generic S3 BIOS version available online as far as I can tell (1.03-08 vs 1.03-06), so I might try that, but it kind of stinks to lose the Viewtop name at startup. Also, I'm not sure if there would be an issue using a generic BIOS since this card supports 4MB of RAM, which is somewhat uncommon for the Trio64V+.

Also, the card has a jumper block who's functions are seemingly undocumented. Mine has the jumper on pins 1-2 just like the card pictured at the Retro Web link above.

EDIT: Here is the Datasheet for the Trio64V+ chip.

https://bitsavers.org/components/s3/DB018-A_T … rator_Jul95.pdf

Last edited by Ozzuneoj on 2025-12-29, 17:47. Edited 1 time in total.

Now for some blitting from the back buffer.

Reply 1 of 4, by subhuman@xgtx

User metadata
Rank Oldbie
Rank
Oldbie

Try and check for an option called USWC write combining or write caching and set that to UC (uncached).

7fbns0.png

tbh9k2-6.png

Reply 2 of 4, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t
subhuman@xgtx wrote on 2025-12-16, 13:57:

Try and check for an option called USWC write combining or write caching and set that to UC (uncached).

Thanks for the tip. I checked the BIOS and didn't see a setting by that name. I think that FASTVID does something with that feature too, but I haven't been using that either.

Seems strange that this specific card would have a problem with it though since my other Trio64v+ doesn't crash. I even tried swapping the BIOS chips on the cards and only the Viewtop card has issues, regardless of BIOS.

I'm suspicious of those jumpers but I just have no idea what they're for.

EDIT: Did some more troubleshooting to narrow things down...

Installed these drivers just to change things up (even though the other ones work fine for other Trio64V+ cards and with this card on another board). At first it seemed like this made a difference, I was even able to change refresh rates... but when I went to device manager it froze when I clicked to open the properties for the video card. I thought it might be a CPU speed issue, so I swapped to a PII 266 (from a PIII 850) and froze at exactly the same spot... clicking to open the properties of the video card. 🤷

Its strange because I noticed this before too... that certain specific actions tended to make it freeze. Like, right clicking the desktop to bring up the context menu would reliably lock the system. Even though doing this works fine with the exact same Windows install on a different motherboard... 😕

I swapped SSDs from my other tester to this one to get the exact same Windows install and it froze on that one too once the drivers were installed. The card works fine on that same Windows install when hooked up to other system (VIA MVP3+ with K6-2 500).

I'm about ready to just start moving jumpers around but I'm always afraid that one is for "send 5v directly to the VGA port!" or something crazy like that. 🤣

Now for some blitting from the back buffer.

Reply 3 of 4, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t

Here are pictures of my specific card, just in case that ends up being relevant.

The attachment 20251218_101421 (Custom).jpg is no longer available
The attachment 20251218_101609 (Custom).jpg is no longer available
The attachment 20251218_101714 (Custom).jpg is no longer available

Now for some blitting from the back buffer.

Reply 4 of 4, by Ozzuneoj

User metadata
Rank l33t
Rank
l33t

Whelp... the plot thickens!

I was testing some more cards today and stumbled upon this one.

The attachment 20251227_203031 (Custom).jpg is no longer available
The attachment 20251227_203037 (Custom).jpg is no longer available

I think this the only other card I own with all socketed memory, so another good reliable candidate for testing memory, right?

Nope! It also freezes up my 440BX test system in exactly the same places as the Viewtop card! Blehhh...

So, I guess the good news is that it is very unlikely to be a physical defect with either of them. Right down to freezing the moment I click "Properties" on the video card in Device Manager... exactly the same behavior for both cards, and no visible artifacting.

I have attached the BIOS for this card to this post as well. I did a basic comparison with windiff and the two BIOS files are completely different. This one is version 1.02-02 and it is actually branded by Phoenix Technologies, which is a bit unexpected for a generic VGA card. Anyway, the BIOS isn't real important because I already tried swapping BIOS chips to no avail with the other card.

If I feel like obsessing over this any further I can probably compare the layout of these two cards to one of my working cards to tell if there is any obvious difference in the resistor layout.

EDIT: Also, the FCCID on the card says it was granted to National Advantages Computer Inc of Taiwan in March of 1997. Looking up that name I see that the company is also referred to as "N4". I'm not sure what the "Path 4II" is referring to... but I like the little scorpion, so yeah, that's probably part of how I ended up with whatever lot this card came in over the past year. 😅

Also, the FCCID includes "SCORP" so, I'm guessing the card was called a Scorpion or something.

Now for some blitting from the back buffer.