VOGONS


First post, by majestyk

User metadata
Rank Oldbie
Rank
Oldbie

First the pictures:

trio64_art.JPG
Filename
trio64_art.JPG
File size
157.5 KiB
Views
992 views
File license
Public domain
trio64_tot.JPG
Filename
trio64_tot.JPG
File size
1.07 MiB
Views
992 views
File license
Public domain

This artefact seems to not occur on lower resolutions like during POST or in the BIOS menue.
Maybe I should mention that I made the [VGA <-> 10-pin header] adapter myself using this diagram:

10-pin_to_VGA.JPG
Filename
10-pin_to_VGA.JPG
File size
47.69 KiB
Views
992 views
File license
Public domain

It seems to work, but I cannot remember having an artefact like this before...
Needless to say the problem is not existent when using a different PCI video card instead.

Reply 1 of 14, by Horun

User metadata
Rank l33t++
Rank
l33t++

Hmm... does it do that with the added ram removed ? Am wondering if a video ram may have an issue (a row/column address thingy)

Hate posting a reply and then have to edit it because it made no sense 😁 First computer was an IBM 3270 workstation with CGA monitor. Stuff: https://archive.org/details/@horun

Reply 2 of 14, by majestyk

User metadata
Rank Oldbie
Rank
Oldbie
Horun wrote on 2023-11-19, 20:49:

Hmm... does it do that with the added ram removed ? Am wondering if a video ram may have an issue (a row/column address thingy)

I tried with the soldered 1MB only - no difference. One of these chips could be flaky of course.

It´s funny that the bar is transparent, but I´m unable to draw any conclusions from that.

Reply 3 of 14, by majestyk

User metadata
Rank Oldbie
Rank
Oldbie

Video-RAM is not the problem here, as it seems. I replaced all 4 chips (both soldered and socketed ones), but the bar is still there.

A PCI-card with the same S3 "Trio64" chip works perfectly.

I wonder if some flaw in the video BIOS might be causing this bar at certain resolutions. The BIOS is integrated in the mainboard BIOS so replacing it would be difficult.

Reply 4 of 14, by Tiido

User metadata
Rank l33t
Rank
l33t

If it wasn't a RAM issue it suggests there is a fault internal to the chip, display line FIFO or something similar is faulty. To get such a bar only in one part of the image and stay in same place in different resolutions you need some active involvement and BIOS will not be able to actieve that, not when most apps write stuff to video memory directly and even if they didn't, it would take extra effort to single out some pixels per line to get the observed effect.

T-04YBSC, a new YMF71x based sound card & Official VOGONS thread about it
Newly made 4MB 60ns 30pin SIMMs ~
mida sa loed ? nagunii aru ei saa 😜

Reply 6 of 14, by Tiido

User metadata
Rank l33t
Rank
l33t

What other resolutions it does appear in besides 640x480 ? It looks like 70Hz text mode (720x400) is unaffected and I suppose 320x200 too ? One can try VBETEST from UNIVBE to try various resolutions via VESA and see what shows up.
I'm pretty sure it is a chip fault when there's no mechanical issues elsewhere (loose or shorted pins), you already swapped memories and nothing changed from that. This chip has internal RAMDAC so there isn't really more to blame either. It might be worth doing a finger test while machine runs, loose or otherwise problematic pins often react to presence or contact with finger, although care should be taken not to touch the two pins that connect to the nearby 14MHz crystal, that will disrupt clock generation and image is likely to disappear and other weird things happen. But ones connecting to memory are primary interest.

T-04YBSC, a new YMF71x based sound card & Official VOGONS thread about it
Newly made 4MB 60ns 30pin SIMMs ~
mida sa loed ? nagunii aru ei saa 😜

Reply 7 of 14, by majestyk

User metadata
Rank Oldbie
Rank
Oldbie

I have a couple of "donor-cards", but they all have some variant of this chip like Trio64 / + / V2 / DX. Would these be fully compatible? Or should I source some card with the original variant?
I will have to live with the stock BIOS, since there are no updates to be found and I also found no other mainboard with this VIA chipset and Trio64 onboard graphics.

Reply 8 of 14, by Tiido

User metadata
Rank l33t
Rank
l33t

I don't know the compatibility between different variations of the chip. I assume some will need specific BIOS too.
It shouldn't be too difficult to update/change the video BIOS, in many cases it is just before the main BIOS in the ROM chip, i.e in 128KB chip the first 32KB is video BIOS and rest is main BIOS, in which case you can just replace the first part with something new but only a dump will tell if that is actually the case.

T-04YBSC, a new YMF71x based sound card & Official VOGONS thread about it
Newly made 4MB 60ns 30pin SIMMs ~
mida sa loed ? nagunii aru ei saa 😜

Reply 9 of 14, by majestyk

User metadata
Rank Oldbie
Rank
Oldbie

I tried the tool "VESATEST" and it detects the following modes:

Trio64_modes.JPG
Filename
Trio64_modes.JPG
File size
39.02 KiB
Views
583 views
File license
Public domain

The vertical bar in Speedsys appears at 640 x 480, so I tested all the colour settings (8, 15, 16, 32) for this resolution with VESATEST and all the pictures are perfect.

Reply 10 of 14, by Tiido

User metadata
Rank l33t
Rank
l33t

It might be related to 4bit modes only then, Speedsys uses 16 color 640x480, like Windows desktop with the default VGA driver.

T-04YBSC, a new YMF71x based sound card & Official VOGONS thread about it
Newly made 4MB 60ns 30pin SIMMs ~
mida sa loed ? nagunii aru ei saa 😜

Reply 11 of 14, by majestyk

User metadata
Rank Oldbie
Rank
Oldbie

Here´s the vesatest output of some other PCI Trio64 card that displays the speedsys gui without any issues:

Trio64_modes2.JPG
Filename
Trio64_modes2.JPG
File size
48.15 KiB
Views
554 views
File license
Public domain

Reply 12 of 14, by Tiido

User metadata
Rank l33t
Rank
l33t

You should try to use the attached program (VBETEST from the freeware Scitech Display Doctor package) with any supported 4bit modes to see if they also produce the problem. It looks like the program you have been using doesn't support the 4 bit modes. It is worth to try Windows desktop in VGA mode (or just its setup as it also uses 640x480 in 4 bits).

Attachments

  • Filename
    VBETEST.zip
    File size
    79.1 KiB
    Downloads
    20 downloads
    File comment
    VBETEST from SDD
    File license
    Fair use/fair dealing exception

T-04YBSC, a new YMF71x based sound card & Official VOGONS thread about it
Newly made 4MB 60ns 30pin SIMMs ~
mida sa loed ? nagunii aru ei saa 😜

Reply 13 of 14, by majestyk

User metadata
Rank Oldbie
Rank
Oldbie

Thanks for uploading VBETEST!

I tested all the available modes today. It does not offer 640x480 /4 bit, but 800, 1024 and 1280.
The odd vertical bar appeared in none of the tests.
The "Page Flipping" test, where it says "should be no flicker" does flicker in most 4bit and 8bit modes. With a regular Trio64 PCI-card there´s no flicker at all.
The "Wide DAC" test is always o.k.

Here´s the current (original) BIOS file:

Filename
FIC_Aries_III.7z
File size
95.44 KiB
Downloads
17 downloads
File license
Public domain

Reply 14 of 14, by Tiido

User metadata
Rank l33t
Rank
l33t

Ok, interesting, so it is only the VGA mode that doesn't work right.
I see flicker on many cards I have tested, but I'm not sure what causes it. It isn't gonna be related to the problem you have though. It is certain that memory is fine, since everything else works properly and interface to the memory and the S3 itself from CPU side also works. That only leaves possible internal fault in the chip.

Looks like this BIOS actually needs Award BIOS editing tools to get video BIOS changed (or any other modifications done), it is all compressed modules type things. I don't have any experience with this sort of stuff unfortunately... There should be some people who have experience here on the forums, perhaps they can give the help on this matter.

T-04YBSC, a new YMF71x based sound card & Official VOGONS thread about it
Newly made 4MB 60ns 30pin SIMMs ~
mida sa loed ? nagunii aru ei saa 😜