VOGONS


First post, by The Pretender

User metadata
Rank Newbie
Rank
Newbie

Hi!

I have an old 3dfx Voodoo2 as my relic, and i recently have inpected the card and was found a damaged capacitor. I dont remember if this capacitor was damaged since 2 months ago as i have tested this card and has worked fine...

I belieave this capacitor is a decoupling capacitor, and not a passage capacitor. If is a decoupling, probably not have any effect into card, because some times, engeneerings add much more decoupling capacitors as necessary to guaranty the stability of the system.

Any of you have a model of this card to take the capacitance reference of C46?
Or a old reapair schematic or a list of components are good to.

Attachments

Reply 1 of 7, by pentiumspeed

User metadata
Rank l33t
Rank
l33t

To be sure, use a multi-meter and probe the solder pads with one probe on ground. If that is case, one of the pads is grounded then you will have your answer.

Then with both probes, touch the other pad (not the ground) and use other to probe few other capacitors to see if all are connected (this is voltage plane), this will also confirm you have decoupling capacitor.

Cheers,

Great Northern aka Canada.

Reply 2 of 7, by keropi

User metadata
Rank l33t++
Rank
l33t++

I don't think there was a capacitor there in the first place... sure the bottom pad looks weird but the upper one shows no signs of anything being soldered to it

🎵 🎧 PCMIDI MPU , OrpheusII , Action Rewind , Megacard and 🎶GoldLib soundcard website

Reply 3 of 7, by The Pretender

User metadata
Rank Newbie
Rank
Newbie
pentiumspeed wrote on 2020-04-11, 14:14:

To be sure, use a multi-meter and probe the solder pads with one probe on ground. If that is case, one of the pads is grounded then you will have your answer.

Then with both probes, touch the other pad (not the ground) and use other to probe few other capacitors to see if all are connected (this is voltage plane), this will also confirm you have decoupling capacitor.

Cheers,

After i posted this, i have the same idea and I tested witch a multimeter.
One of the pads are conected to ground, but, the other i don't have located into a PCI Bridge (3.3, 5V and 12V conectction).
I found the same NET at other capacitors around the PCB wich is connected to ground same as C46.

Finaly, i located a voltage regulator at front of PCB (AS2830), and the Vout is conected at the "unknow" pad of C46.

Resuming, one PAD are connected into GND and the other are connected into Vout of the AS2830 (voltage regulator).

I believe he was was a decoupling capacitor. But, what was his value?

Reply 4 of 7, by the3dfxdude

User metadata
Rank Member
Rank
Member

You say Diamond Monster 3D II Rev A. I have a Diamond Monster 3D II Rev A. But yours is 12MB and has different RAM, and my card is 8MB. The silkscreen seems basically the same as mine, except mine says 9808 and your says 9818 in the writing in the image.

Anyway my C46 is populated. But so is my C21, did you see that? And yes the above RAM is populated too. Warranty date is 8/24/98.

So there are differences, but nothing to prove anything.

On your close-up picture of C46, I see alot of vias that have exposed copper rings nearby. Did your board experience corrosion? I see nothing like that here. I think that the funny look is the lighting coupled with something on the pad. But you can look at C46 and C21 under a microscope and see if you see evidence of the caps pulled off.

The caps are probably so small, it's probably not worth the trouble.

Reply 5 of 7, by the3dfxdude

User metadata
Rank Member
Rank
Member

I pulled out my Orchid Righteous 3D II 12 MB. It is almost 100% the same layout (minus a couple things I guess that never get populated?), but the writing is in different places. Board # is 9815. The caps C46 and C21 are there.

Reply 6 of 7, by The Pretender

User metadata
Rank Newbie
Rank
Newbie
the3dfxdude wrote on 2020-04-11, 19:19:
You say Diamond Monster 3D II Rev A. I have a Diamond Monster 3D II Rev A. But yours is 12MB and has different RAM, and my card […]
Show full quote

You say Diamond Monster 3D II Rev A. I have a Diamond Monster 3D II Rev A. But yours is 12MB and has different RAM, and my card is 8MB. The silkscreen seems basically the same as mine, except mine says 9808 and your says 9818 in the writing in the image.

Anyway my C46 is populated. But so is my C21, did you see that? And yes the above RAM is populated too. Warranty date is 8/24/98.

So there are differences, but nothing to prove anything.

On your close-up picture of C46, I see alot of vias that have exposed copper rings nearby. Did your board experience corrosion? I see nothing like that here. I think that the funny look is the lighting coupled with something on the pad. But you can look at C46 and C21 under a microscope and see if you see evidence of the caps pulled off.

The caps are probably so small, it's probably not worth the trouble.

huuummm
At this evidence i believe there are a decoupling capacitor (C21 e C46) and the factory have no added this on this board for any particular reason (perhaps they beliave are not necessary for this layout).

I will test the card in future (when i have a time for do it), and post the result here!

Reply 7 of 7, by the3dfxdude

User metadata
Rank Member
Rank
Member

I looked at a few other pics online. It's a common layout. And in every picture the caps are present. Do you know the full ownership history of this card? I have had my 8MB Diamond for more than 20 years (still second owner) and almost as long for the 12MB one, and they are well taken care of. If you are skilled enough, go ahead and add them. If you have any doubts, I dunno what to recommend.

Yes in some design cases, more caps are added as a precaution, and the engineers later remove them since they were not needed to save component cost. But I'd doubt this one would have made any dent to savings.