GRM033R71E102KA01D are actually 0201 (imperial) or 0603(metric).
You need 0603 imperial.

You sadly got bitten by the metric / imperial issue: You bought them as 0603 metric (0201 imperial), which indeed are tiny:

The intended 0603 (imperial) would just be a little smaller than 0805 and are still somewhat usable with a soldering iron.

Edit: Oh well, too slow with my response 😄

So the card got soldered by a friend of mine and works perfectly now. Thanks for all your help.
Now to get a perfect 12MB SLI system i bought another identical card but sadly it's broken.
Mojo says "Bogus number of TMUs (57007)" and randomly reports 0 MB FBI memory.
Someone on another thread said one needs to resolder/reflow the FBI chip.
Now my question is: Which one is the FBI chip? Also the TMU is assumed, i guess it's the lower of the three 3dfx chips? It gets pretty warm when trying to start a game but then crashes.

The one near the PCI is the FBI.
The 0MB detected FBI error can be caused by:
-broken FBI traces
-broken FBI
-broken RAMDAC or 14.318MHz crystal
-broken ferrite beads that power the RAMDAC
-a specific FBI RAM IC broken in a specific way so that the FBI can't initialize the RAMDAC anymore and can't request the needed clocks required for counting the RAM
-broken PCB traces
-broken series resistors
-broken CLK buffer or resistors near it (the single IC or two ICs sitting between the RAMDAC and FBI, depending on the card model/revision).

Thanks. The 0MB detection is randomly, most of the time it reports 4MB.
The bogus TMU message however stays.
I'd give the card to my friend and let him reflow/resolder the FBI chip and check for the other problems you mentioned.

I received a non working Diamond Monster 3D II 12MB, it seems to be undamaged aside from what looks to be a fluid stain behind TMU 0.

I placed the card in a system to check what it does, detect.exe recognized the card and mojo.exe gave the following output:

download/file.php?mode=view&id=187319

The stain bothered me so i gave it a gentle clean with some IPA, after that the card didn't give the same output anymore and mojo hangs with the follwing message:

It doesn't look like i damaged anything, maybe the card just damaged more by powering it on after many years. As the first try failed to initialize the DAC is started checking the components which gave me some results. The schematic of the Creative on the first page of this topic seems to be identical to the Diamond Monster 3D II 12MB (at least it does for my rev. A). After some measuring around the DAC it gave me the following results:

C23 = infinite restistance, seems to be broken and is connected to the DAC
R37 = 2.0K resistance which should be 4.7K, not sure if this will cause trouble?
C5, C6, C7 and C8 = infinite resistance, seem to be broken all just above the PCI connector

The rest around the DAC seems to be fine, i also checked all the resistor arrays which seem to be fine too. The resistor at R153 is a fix made by Diamond as R153 shouldn't be a 0 Ohm resistor as i can see on the schematic of the Creative card on the first page. All rev. A card have this fix or should have. The pins from the TMUs and FBI seem fine after a first inspection. Any ideas if i'm on the right path here? Could this just all be related to some broken capacitors?

In most of the cases you can't get actual values of electronic components by measuring them while they're soldered to the PCB.

analog_programmer wrote on 2024-03-05, 19:06:

In most of the cases you can't get actual values of electronic components by measuring them while they're soldered to the PCB.

So measuring isn't going to get me anywhere? I ordered a test bench so i can measure voltages easier while running to see if the DAC gets powered at all. Maybe that is a better solution.

Yarda wrote on 2024-03-05, 21:20:

So measuring isn't going to get me anywhere? I ordered a test bench so i can measure voltages easier while running to see if the DAC gets powered at all. Maybe that is a better solution.

Of course you can measure voltages in many points while the board/PCB is energized, but if you want to measure actual values for some of the components, then you have to disconnect them from the circuits they're in i.e. desolder them from the PCB.

TheMobRules wrote on 2021-09-06, 17:23:

tkrn wrote on 2021-09-06, 17:06:

Instead of creating a new thread, I need assistance finding out two capacitors C50 & C81 on the STB Voodoo2 cards. I can't find an equivalent component map. I need help reading the values and packaging to get the right SMD capacitors. Thanks.

Both of those are 10uF 16V. In fact, all the SMD electrolytics on the card have the exact same value.

Thanks 😀 I used this info to replace C45, C46 and C48 missing on the STB 12MB card i've just bought "Untested"

tkrn wrote on 2021-09-10, 19:38:

I re-capped both of my Voodoo3 1000 videos card. Major problem on one of the cards, two of the SMD capacitors broke off C57 & C53. I was able to find one of them but I never noticed them being broken off. One of them, I did for sure. I don't know the value of these capacitors. Can you help with that too? Lastly, I had to do a trace repair on of the 10uF capacitors. It turned out okay. Also, on one of the cards two of the main chipset had bent pins. I got my magnifying glass out with a new razor to separate the smushed pins though I don't want to test it until I get replacement caps for C57 & C53...

I'm a bit confused with your reference to the Voodoo3, but the photos you included look like the STB Voodoo 2 card I have. I took C57 & C53 off my card and measured with a capacitance meter. They both measure between 240 and 250nF.

DonutKing wrote on 2021-12-04, 12:53:

Hoping someone here can help... I have two 12MB STB voodoo 2's. One is working, the other is detected in Windows device manager […]
Show full quote

Hoping someone here can help... I have two 12MB STB voodoo 2's.
One is working, the other is detected in Windows device manager but all Glide games and the 3dfx demos tell me that they couldn't detect a Voodoo card.

I discovered that C577 was missing off the back of the faulty one.
I had a spare 100nF SMD cap so soldered that on, but this didn't fix the issue.

Does anyone know what the value of that cap should be or if 100nF would be 'close enough'?..

I've just taken C577 off my card and it measures 133nF so 100nF is close enough but i'd probably put a 120nF in there. It seems like you have bigger problems though (Like me)...

My card has 4 ceramic caps broken off/ missing on the back which are C551, C582, C612, C614.. can any of you STB owning folks help me out with values?

I tried putting 100nF in each location and the card is detected, passes D3D tests, official drivers loaded fine, control panel shows with gamma correction, mini animated 3DFX logo etc... but when loading Games, 3DFX logo appears and up to about 2 to 10 seconds of everything seeming fine, the GFX corrupt big time and the PC locks up. Took some caps off nearby on similar circuit (near the RAM for example) and got some more accurate guesstimates (C551 is most likely 240nF) but again no real difference, maybe is even locking up quicker.

The card is Running in a Win 98se DX7.0a, ASUS P2B 1.10, 1014 Beta 003 BIOS, P3 733 (running at 100MHz FSB 550MHz) paired with S3 Trio 64 PCI and also tried pairing with an AGP Riva TNT2 M64, same probs.. I would like to get the capacitors correct to at lease rule them out, even though I think this card has much bigger problems . It seems the STB has quite a different layout to the standard one other manufacturers used for Voodoo 2 cards, I've tried to compare to the original diagrams by the OP but I can't see any similarities.

I've added some photos to show my board, and where freezing occurs in games.. Tomb Raider III, Wipeout 2097 and Ultim@te Race Pro