VOGONS


First post, by Shponglefan

User metadata
Rank l33t
Rank
l33t
Orchid Righteous 3D - not working.jpg
Filename
Orchid Righteous 3D - not working.jpg
File size
1.41 MiB
Views
639 views
File license
Fair use/fair dealing exception

Updated: The video card itself is fine. The problem is I was testing it on a Pentium 4 which was clearly too fast for it to work reliably.

I have this Orchid Righteous 3D that is having issues.

Using MOJO.EXE and logging debug output, the results appear fine in comparison to what others have posted from working cards.

I then tested it with Whiplash (Fatal Racing) and Tomb Raider in DOS. Whiplash works fine. Game loads, plays fine, no artifacts or other 3D issues.

Tomb Raider freezes on the spinning 3Dfx logo. This is sometimes accompanied by visible artifacts or even a full-screen of multi-colored noise.

I can launch Tomb Raider if I restrict the card to only 1MB of TMU memory via SET SST_TMUMEM_SIZE = 1. It will run in-game, but attempting to bring up the menu causes the game to drop to DOS with an error about not being able to access the memory. Which I suppose makes sense since the card is being restricted to only 1MB of TMU memory.

Is there any way to know which memory chips are utilized when TMU memory is set to 1 MB? Or rather, which memory chips are not being used?

If I can narrow it down to only two of the memory chips, I can then try replacing one then the other.

Last edited by Shponglefan on 2024-03-04, 15:17. Edited 2 times in total.

Pentium 4 Multi-OS Build
486 DX4-100 with 6 sound cards
486 DX-33 with 5 sound cards

Reply 1 of 7, by dominusprog

User metadata
Rank Member
Rank
Member

I had a slimier problem with an S3 Trio64 card, a black screen and all sorts of artifacts. After inspecting the card I found out that one of the RAM chips has a broken solder joint.

Duke_2600.png
A-Trend ATC-1020 V1.1 ❇ Cyrix 6x86 150+ @ 120MHz ❇ 32MiB EDO RAM (8MiBx4) ❇ A-Trend S3 Trio64V2 2MiB
Aztech Pro16 II-3D PnP ❇ 8.4GiB Quantum Fireball ❇ Win95 OSR2 Plus!

Reply 2 of 7, by Shponglefan

User metadata
Rank l33t
Rank
l33t

I should mention I did conduct a physical inspection including pin testing. Everything seems solid. The card is in otherwise immaculate shape.

Pentium 4 Multi-OS Build
486 DX4-100 with 6 sound cards
486 DX-33 with 5 sound cards

Reply 3 of 7, by Morotz

User metadata
Rank Newbie
Rank
Newbie

You could try applying pressure on the memory chips while running to see if anything changes. The memory isn't super expensive, so you could just replace the entire TMU memory in one go. However, installing sockets might not work as the memory chips are so close to each other on the Righteous 3D.

Retro systems:
Athlon XP-M 1600MHz, Abit KT7A, 512MB PC133 CL2, Quadro FX3000, Vortex2 + SB AWE64, Win98SE
Athlon 1100MHz, Geforce2 Ti, Win2k
Pentium III 550MHz, Voodoo 5 5500, Win98SE
Xeon L5430, 4GB DDR2, GTS 450, WinXP

Reply 5 of 7, by Shponglefan

User metadata
Rank l33t
Rank
l33t

TL/DR: Card appears fine. Should have tested it in another system before reflowing all the TMU chips. 😓

Longer version:

After hours of additional testing, I can update on this. The card appears to be perfectly fine. The problem was how I was testing the card.

I should have mentioned originally that I was testing in a 2.8 GHz Pentium 4 setup. This is admittedly well past the normal spec of system with which one would use a Voodoo 1 card.

My original diagnosis appeared to be purely coincidental. When I restricted the TMU memory to 1MB and the card started launching Tomb Raider, I assumed it was the TMU memory chips. I further tried the suggested pressure test and after clamping the chips, the card subsequently launched into Tomb Raider with the full 2MB TMU just fine. All of this pointed to an issue with the TMU memory.

I figured it was a faulty solder joint. On subsequent tests the card started behaving intermittently. Sometimes it would freeze on the 3D logo as before, sometimes it would launch just fine.

I opted to do a full reflow of all the TMU memory pins. This did not improve things as the card still would fail to launch most of the time. When it did launch, I was able to use it without any issues in-game. Zero artifacts, freezing or other glitches.

Finally I did what I should have done in the first place: I put tested it in a more era-appropriate Pentium 200 system. In this setup, Tomb Raider would launch 100% of the time and worked with no issues.

My conclusion is that an Orchid Righteous 3D + Pentium 4 + Tomb Raider is just not a good combo. Curiously the Diamond Monster 3D works fine with Tomb Raider even with the P4. Not sure exactly what Diamond did, but I'm going to leave that latter card in this P4 system and save the Orchid Righteous 3D for something else.

edited: Upon further testing I'm now getting the same issues with the Diamond Monster 3D. Sometimes it works, sometimes it doesn't. Evidently I got lucky with my early testing, as it worked without issues upon several tries. But I clearly hadn't tested it enough.

Pentium 4 Multi-OS Build
486 DX4-100 with 6 sound cards
486 DX-33 with 5 sound cards

Reply 6 of 7, by meljor

User metadata
Rank Oldbie
Rank
Oldbie

It is well known that above a p2 350-400mhz cpu a voodoo1 can give problems ranging from glitches to some games not starting to even black screens.

Some people have them running fine on faster systems but it can give lots of problems.

So indeed, aways test them on slower systems first.

asus tx97-e, 233mmx, voodoo1, s3 virge ,sb16
asus p5a, k6-3+ @ 550mhz, voodoo2 12mb sli, gf2 gts, awe32
asus p3b-f, p3-700, voodoo3 3500TV agp, awe64
asus tusl2-c, p3-S 1,4ghz, voodoo5 5500, live!
asus a7n8x DL, barton cpu, 6800ultra, Voodoo3 pci, audigy1

Reply 7 of 7, by smola

User metadata
Rank Newbie
Rank
Newbie

Here is log from my old repair 3dfx VooDoo1 PCI 4MB (ORCHID Righteous 3D) , maybe you'll find there smth useful: http://3dfx.pl/cgi-bin/yabb2/YaBB.pl?num=1617221455/83#83.
Unfortunatelly to see links/images registration on forum is mandatory, also Polish text, but this shouldn't be a problem to auto-translate 😉

my repairs: mobo index :: vga index :: requests