You can't use just a multimeter to determine which input goes to what output.
just No.

Ok thanx for the great detailed explanations in that case. I will wait for my chip to arrive

The chip came today and I installed it, the resistance values look good and looks like my previous shorts are gone.

Turned on the system and unfortunately no video but I can hear the system booting.

Before I gave up I tried starting with vga direct into 2d card, started a 3dfx game and then manually plugged vga direct to 3dfx (as you guys gave me the idea) and to my suprise got picture.

I don't have anymore time today to try do a side by side comparison with my other working card but what do you guys think is most likely happening.

1. got a semi dud chip? What are the chances it doesn't switch 2d but is doing 3d? (low I would say)
2. Another component carrying the 2d signal to the switch chip burned out.
3. "insert here"

At this stage if I was really desperate I could ofcourse plug my 2d card and the 3dfx into a kvm switch and manually switch them.

Yes I did check the soldering on each leg and I went over it twice with a flat tip to make sure its all solid, no bridges etc

at this point I'd say fuck it and grab an Extron VSW 2VGA A off ebay

it handles the auto switching just like the factory passthrough would and gives much better bandwidth

maxtherabbit wrote on Yesterday, 14:51:

at this point I'd say fuck it and grab an Extron VSW 2VGA A off ebay

it handles the auto switching just like the factory passthrough would and gives much better bandwidth

+1 to that

I disagree with darry and maxtherabbit. You are so close, you can't stop now! You either fix it completely or you destroy it! 😁
Seriously, this looks like a fun project. Let's not forget that most of us are not really in it for the "gaming" part. 😀

Hey, at least the Monster is drawing stuff again! That's great news, it means that the output stuff and business end of the GPUs is probably working at least.
Did you check to make sure you didn't accidentally bridge any of those tiny pins together on the chip you replaced? Also look for traces that may have burnt out from the surge.

Actually there is merit to the Extron switch idea too. By not running your VGA output through the Voodoo you avoid the usual signal degradation on your 2D/VGA signal, though it is a bit less convenient to use. It's your call whether or not you want to 'fully' fix the device though.

There is another way to avoid the image quality degradation in the passthrough : Image quality fix for Voodoo1 and 2 cards

@bloodem
I am not against the idea of trying to get it working again, as a matter or principle, though the extra degree of effort and possible risk involved would not entice me to do so if I was the one experiencing the issue. I personally used a dual-input monitor (Viewsonic PT771) with my Voodoo 1 and later Voodoo 2 for most of the time period when those cards were current tech to avoid using the passthrough and would probably do the same today, even with a fully working card . Consequently, if I purchased a card with the issue OP currently has, I probably would not even notice it . Anyway, what to do next is definitely a matter of personal opinion .

@Tiido
That's an interesting solution . If the OP decides to try to fully fix the card (and hopefully succeeds), this would be a great next step .

Now looking back in hindsight should of ordered 2 chips, could of tried both and even soldered 1 on top if they both worked to improve the pass through.

This is a second 3d monster I have so not currently in use but yes my aim is to have a backup/collection that works 100%

I am very happy that it is alive, I guess it will be a little side project when I have time to poke around, I may also order another chip again.

At the moment I have compared resistance readings on the chip, vga in and vga out with the working card.

They are all consistant in the way that all ground pins are grounded and all other pins show a similar resistance, however the working card has a slightly lower resistance on all tested pins of about 0.100 ohm but constant. if it was about 0.30 ~0.50 I would say that is normal to vary accross different hardware but .100 seems like there is some phantom component that seems to be open. Visually everything looks ok so that is the hard part