Following this thread. I have two 12MB STB Voodoo 2's and one of them seems to have failed in a similar fashion - detected in device manager but gives the same GlideInitEnvironment error when trying to launch a game. So far I've been unable to troubleshoot it.

feipoa wrote on 2020-04-07, 01:10:

I don't have an SLI cable, but I recall reading that floppy cables work. Is there any modification I need to do to them?

yes you absolutely need to modify it!

https://www.vogonswiki.com/index.php/How_to_m … doo_2_SLI_cable

Lucky he asked! 😁

imi wrote on 2020-04-07, 01:16:

feipoa wrote on 2020-04-07, 01:10:

I don't have an SLI cable, but I recall reading that floppy cables work. Is there any modification I need to do to them?

yes you absolutely need to modify it!

https://www.vogonswiki.com/index.php/How_to_m … doo_2_SLI_cable

that guide is pretty good except for the bit about using a hammer, that's terrible advice, use a bench vise

pinch the cable between the two halves of the connector and then just crush it down

I recall the floppy cable needing to be modifed, just didn't recall how. I watched some Youtube video on it this evening and made my own SLI cable. I used a multi-meter to ensure each pin is connected through on the connector.

I used the STB as the primary Voodoo2. I have the DX7-required reference drivers installed. I think they are version 3.02.02. I no longer receive this error at boot:

But when I open Display Properties and click the Voodoo2 tab, the system pauses for about 10 seconds, then displays this error:

This is the same error I receive if I just have the Diamond installed and click on the Voodoo2 tab. So looks like the Diamond also isn't working in SLI mode.

Unfortunately, I could not locate the 500-0009-01 chip on seekic.com. If the QS5244 or S3L384Q IC's are bad, could they cause this issue? Are they known to fail?

Not sure what else to try. I could reflow the FBI one more time, or try 100 nF caps at c51 and c53 out of desperation.

The card was made by Creative. I use small diameter needle probes, no scraping is required.
I've never seen a QS5244 or equivalent IC go bad. But it is unlikely that IC is related to your issue. It buffers HSYNC/VSYNC for the SLI connector and a couple of other signals (FBI to TMU if I recall correctly).

If you try to run a glide app on a card with a missing QS5244, you won't get the "expected Voodoo2, none detected" error, you will just get a blank screen and the app will freeze.

If you can't find a FBI chip somewhere in your area, I can send you one, but you will have to cover shipping via courier (can't go to the post office due to the lockdown).

Thank you for the offer, but in my hast to quickly reflow the FBI chip (with a fine tipped iron, no hot air), I lost 4 pads, two of which have no visible trace. Given the circumstance, the cost:time:reward:risk ratio is leading me to give up on this board.

The one pad I lost for which I can follow the trace, I soldered on a 30 AWG wire. One of the pad/pins is GND, the 3rd and 4th, I don't know what they are supposed to be. I tested it again anyway, but the error was the same. Before reflowing the FBI chip, I replaced 3 of the back caps with 100 nF, and tested it, but it obviously didn't help.

Do you have the datasheet for the FBI chip so I can see where these lost pad/pins are supposed to go?

Sorry to hear that. I don't have a datasheet for the FBI chip. Can you post a photo of the affected pins/pads?

I'm not sure how much detail in the photo you are requesting, but here's a photo showing the location of the pads. It's actually only 3 pads, not 4. I counted a 4th in my previous post because I broke off one of the leads in its mid-section, but the pad is still OK. I have not marked this pin in the photo.

If I only have to solder one wire, I suppose replacing the FBI wouldn't be too horrific. But if I have to do 3 , I don't know. its pretty easy to get solder bridges when the spacing is this fine.

The photo you uploaded is fine. The marked pin on the top-left side of the FBI is not connected, you can ignore it. The marked pin on the bottom-left side of the FBI goes to the 3.5V rail. You could route a piece of wire from the FBI pin through the hole near C7 and connect it to C68 on the bottom side of the board.

Thank you for the information. I'll wire the 3.5 V, but I'm pretty sure I knocked the pad off on the 2nd reflow.

There's another lead that I accidentally severed fixing a solder bridge. The end connected to the FBI still goes to GND, so it is connected internally. I assume its safe to leave this pin severed. The pad is OK, just the lead is cut.

EDIT: I followed your routing path recommendation to get 3.5 V to that pin, but the error at boot persists. Soldering two wires isn't too bad, so I'm back to considering the replacement FBI chip. Is there any other component you think could be the culprit? There looks to be 5 fuses on the board, FB1 thru FP5 and FB7, and they all show a short. There's 3 transistors, which I assume are for the RGB signals. Would these transistors going bad result in this error? I suspect not because I'm guessing those transistors are used for switching the pass-through cable. I also wouldn't think that faulty memory chips would result in the error we are seeing, but would display artifacts.

If you have plenty of these FBI chips which are NOS and don't mind sending one, how much would shipping be to Canada? I'm assuming there are several people who have a vested interest to know if replacing the FBI chip will resolve the issue.

Pending no other ideas to test for, and confirming shipping isn't too $ for this chip, I would 1st proceed to cut off the FBI chip and remove the leads to ensure I don't loose any more pads.

EDIT:
I've found a few online shops which have the 500-00090-01, but they are expecting orders in quantity. https://www.cyclops-electronics.com/parts/3df … ctiv/500000901/

Here's one which may take less, but the website looks really sketchy, http://www.yxfic.net/english/3dfx/500-0009-01.html

I'll try this seller from alibaba to see how much, but shipments from China have come to a crawl. https://www.alibaba.com/product-detail/-ic-ch … .37a45679EAcVj0

EDIT2: heard back from that seller on alibaba. The price is ridiculous.

500-0009-01 3DFX QFP MOQ=5pcs*31.9usd/pcs=159.5USD
22usd by EUB to Canada in 15-20 days

You're right, the 3 tranistor looking components can't cause this error. Those are just some diodes on the RGB lines, the switching is done by the PI5C3384Q or equivalent IC close to those diodes. FB* are ferrite beads, all of them should read as a short.

I can send you a NOS FBI but shipping is somewhat expensive. The cheapest I found is about 60E. It should be significantly cheaper sending it via regular post, but due to the current situation it's not possible.

The price you got from the alibaba vendor is indeed ridiculous.

It would cost 60 euro to send this little QFP chip? Was thinking more like 6 euro. 60 euro is the shipped cost of a Voodoo2. Did the postal service shut down where you are? It's still running here, and I have something I need to post, but I have been reluctant to actually go to the post office. I know the one inside the pharmacy is still operating.

I know, it's ridiculous. 60E is the cheapest option I found (it's basically 60E for a letter) others are even more expensive (DHL or TNT was ~80 if I recall correctly) . The postal service is still up, but we're currently under quarantine and not allowed to leave our homes except for emergencies or groceries. If I need to ship something now the only option is having a courier pick it up from my house. Sending it as a letter via regular post would be 1-2E.

OK. I guess this project will be on hold until at least November then, unless you have post offices in grocery stores where you live. We have post offices in all sort of businesses here, supposedly to save on property costs. I've seen them in pharmacies, dry cleaners, and grocery stores so far.

I have never seen a post office in a grocery store or a pharmacy over here. The lockdown is in effect for at least 1 more month, in the meantime I'll see if I can find something cheaper. If not, I'll send it via regular post when I'm allowed to.

Welcome friends

I will share with you my observations and what I did.
The topic of repair Voodoo 2 Diamond 3D II has been raised many times in various forums around the world, however, without a victory. This indicates a high vulnerability to damage to Diamond 3D II cards.

Voodoo Diamond 3D II rev B dismantled the components for testing purposes 8MB RAM, TMU0, TMU1 - they do not take part in this damage. I desoldered 4 MB RAM FBI checked, soldered is OK. I replaced the FBI 500-0009-01 chip with a new one, but it didn't help, I replaced GENDAC ICS5342 and QS5244 also doesn't help.

win98 card detects you get such a message

Voodoo is good to test and make logs - it's easy and worth doing - the description is here Faulty Voodoo 2?

after the log is made, there is a problem reading Fill Device Info.
The attempt is repeated 4 times without success ..

for testing I took the second Voodoo 2 with ICUVGA-GW803 designations which only had mechanically damaged texture mapping processors TMU0 TMU1 500-0010-01.
I desoldered the remaining elements to get the same card as Diamond II.
Voodoo ICUVGA-GW803 is correctly detected by win98 and does not cause _GlideInitEnvironment error ...

the ICUVGA-GW803 log shows that there is a problem only with TMU 500-0010-01 and its memory.
However, the full device information is read correctly.
After entering the command to regedit SSTV2_TEXMAP_DISABLE = 1 and restarting the computer, it is possible to start the NFS3 game with texturing turned off.

I wonder where the fill device info data is stored and why it cannot be read in Diamond II.
I swapped the QS5244 chip with 74AC244 between Voodoo.
Diamond does not work, ICUVGA-GW803 plays ok. NFS3 ..
The problem probably lies in the connection 500-0009-01 or its circuit ... I soldered all smd components, capacitors resistors again, it did not bring victory ..
I suspect the damage to the board in the FBI pin to the PCI slot

Thanks for the information mario. I may try to check the logs when I work on this again.

Will you try changing your FBI chip to see if that will solve the problem?

Are the Diamond branded cards more problem prone compared to other brands of Voodoo2 cards? Perhaps something to do with their soldering process, or perhaps middle layer trace connection issues?