3dfx Archive - Print Page

3dfx Archive
http://www.falconfly.de/cgi-bin/yabb2/YaBB.pl General Section >> General Discussion >> 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700 http://www.falconfly.de/cgi-bin/yabb2/YaBB.pl?num=1370739590 Message started by trevormacro on 09.06.13 at 02:59:50

Title: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 09.06.13 at 02:59:50

Hi everybody,

My dream came true. I have found a 3dfx rampage card in dead condition.
I will do all my best to restore this card with the great help of one 3dfx engineer.
I have created a link for the card.

http://www27.brinkster.com/trevormacro/specter1000/rampage.html

nb. I don't have the card with me because I am trying to bring it to life. I hope I will succeed.

Here is a frontside picture of my rampy:

EDIT: 09/06/13

Video:

http://youtu.be/eARzJ8gd5Fo

Best Regards,
Trevormacro

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by rottentreats on 09.06.13 at 05:10:11

Whoa! :o Amazing catch! Congratulations! 8-)

I hope to see the card running in the future

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by nest on 09.06.13 at 06:50:30

9th Wonder of the World

Congratulation ;)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Quaker763 on 09.06.13 at 08:03:35

Hahaha nice find man :)

It's a tad bit disconcerting though that the socket and back of the chip are black :-/

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 09.06.13 at 12:55:58

Quaker763 wrote on 09.06.13 at 08:03:35:

Hahaha nice find man :)

It's a tad bit disconcerting though that the socket and back of the chip are black :-/

Not unusal, early chips usally have no markings.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by vik on 10.06.13 at 00:28:04

good luck! :)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Thema on 10.06.13 at 10:00:12

Nice one! I hope the processor is fine.
Good luck getting it up and running!

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 10.06.13 at 12:11:00

Very curious about what clean system will be used for clean the test socket (gold contact) and rampage chip, I am asking my self if not would be enough to use Ultrasonic cleaner with alcohol isopropanol.

ps. Nice catch Michel!!! 8-)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 15.06.13 at 21:00:58

Hi and thanks,

I have the card now.

The MSI 848P Neo-V motherboard doesn't boot with the card in agp4x slot.
Website updated.

Socket opened:

http://youtu.be/ZJPpOxVCnTc

The card:

http://youtu.be/SNNfblhWVts

I need some help now to repair it.
Perhaps this MB doesn't support my card!

Thanks in advance.
Regards,
Trevor.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 15.06.13 at 21:22:16

Thanks for the update Michael.

Did you try to start the card in another motherboard ?
( An agp2x )

Fit the rampage cip in only one Position in the socket ?

More Fotos please :)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 15.06.13 at 21:46:13

@Michel,
Just starts for using pci.exe SW and see if card is detected via 3Dfx ID vendor.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by ultima on 16.06.13 at 09:30:26

Sweet find indeed....but how did you get it?? Bought it from another collector I assume?!

Anywho me keeps his fingers crossed you will get this beauty up and running.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 16.06.13 at 14:28:03

osckhar wrote on 15.06.13 at 21:46:13:

@Michel,
Just starts for using pci.exe SW and see if card is detected via 3Dfx ID vendor.

Regards,
Oscar.

I would try a 1999-2000 vintage system (AGP x 2 is fine for testing) go into the BIOS and disable all video related shadowing and caching and use a NON-3dfx PCI video card to boot up with the PCI board indicated as the primary video adapter in the BIOS.

use PCI.exe and see if a device with vender id 121a is present, this will tell you if the Rampage board is live or not.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 16.06.13 at 15:11:47

@Gary: is possible to put the cip in the socket in a wrong position ?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 16.06.13 at 16:21:15

m14radu wrote on 16.06.13 at 15:11:47:

@Gary: is possible to put the cip in the socket in a wrong position ?

I can't say not having a sockected Rampage board, the number one position is clearly marked on the chip and board though.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 16.06.13 at 16:54:19

@m14radu:

I don't have any others MB with agp2x. I just have one AGP pro slot of my Tyan K7xpro but not compatible with the rampy.

I have fitted the chip in the right position. There is a mark on the rampy corner (rearside) and was putted on the lower left corner. If for some reason chip is placed in bad place, it will be burned!

Some others pictures:

@ultima:

It's private. Not from a collector.

@osckhar:

When I used my IR lasergun on each DDR, the temperature increases but on cooler NOT. Perhaps the socket and chip need to be cleaned with your Ultrasonic cleaner in alcohol isopropanol (recommanded by the 3dfx engineer).

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 16.06.13 at 17:29:59

Why is the Tyan K7xpro not compatible with the rampage ?
Tyan K7xpro support AGP2x/4x right ?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 16.06.13 at 17:51:50

m14radu wrote on 16.06.13 at 17:29:59:

Why is the Tyan K7xpro not compatible with the rampage ?
Tyan K7xpro support AGP2x/4x right ?

I have an AGP Pro 110 slot; supports 4X/2X AGP.
I need to put in this place, right?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 16.06.13 at 18:13:56

It should work Michael.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 16.06.13 at 18:26:12

m14radu wrote on 16.06.13 at 18:13:56:

It should work Michael.

Is this agp pro support 1,5 V for rampy?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 16.06.13 at 18:27:17

You schould check that before with another agp card, a non 3dfx one ! :)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 16.06.13 at 18:38:12

m14radu wrote on 16.06.13 at 18:27:17:

You schould check that before with another agp card, a non 3dfx one ! :)

Yes I have a ATI Radeon 7000 VE Family agp4x 1,5V. I will let you know.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 16.06.13 at 18:51:34

gdonovan wrote on 16.06.13 at 14:28:03:

osckhar wrote on 15.06.13 at 21:46:13:

@Michel,
Just starts for using pci.exe SW and see if card is detected via 3Dfx ID vendor.

Regards,
Oscar.

Perhaps I need to put the jumper like you on the card to boot it?

http://www.3dfx.ch/gallery/d/48735-2/DSCN5650.JPG

On what pins is your jumper?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 16.06.13 at 20:37:14

Quote:

Perhaps I need to put the jumper like you on the card to boot it?

http://www.3dfx.ch/gallery/d/48735-2/DSCN5650.JPG

On what pins is your jumper?

If you look close at the picture, the jumper is in the off postion, its just being stored on one pin.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 17.06.13 at 12:22:28

Quote:

@osckhar:

When I used my IR lasergun on each DDR, the temperature increases but on cooler NOT. Perhaps the socket and chip need to be cleaned with your Ultrasonic cleaner in alcohol isopropanol (recommanded by the 3dfx engineer).

I only hope the rampage chip always have been placed on the right position since it would be burn. Chip only has ONE position. if some one 'playing' to put/unpunt chip can place it wrong position.

Test socket works with 'springs' pins in each side of the socket. If for some reason any 'spring' is locked can cause fails on the board for bad contact.

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 17.06.13 at 17:20:32

Tyan MB tested with the rampy without luck too :(

Rampy in route at Oscar's home. Stay tuned. Surprises not ends here. :)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by ultima on 17.06.13 at 19:34:40

I have a load of different motherboards on which the card can be tested, ranging from socket 7 till socket 939 and almost all in between, so if needed, just let me know, ok??

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 17.06.13 at 19:54:15

Surprises not ends here ?
Comme on Michael, what is the next 3dfx card ? :)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 20.06.13 at 15:40:00

Hello,

Package came in today.

Yupe, I also own a SPECTER 1000 agp test socket card. More than 3 years to close the deal. Card came in just 1-2 month ago (crazy and hard deal). 8-)

My Rampy works ok just some problems in the interfce memory.

At the moment Michel's Rampage has some problems.

STATUS:

Card is bended (perhaps it can be the cause the test socket does not make good contact).
Rampage chip has 2 missing balls.

Add those 2 balls will not be poblem since pad is still present. I need to check if size ball is 0.76mm.

More test later.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 20.06.13 at 16:17:50

My website is updated.

I have found one other rampage chip and Rampage test vehicles chips.

News suprises inside:

http://www27.brinkster.com/trevormacro/specter1000/rampage.html

Take a look and enjoy :)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 20.06.13 at 18:57:35

Great Card Osckhar !
Congratulation for it :)

i want one toooooooo :D

Both rampage Card look just amazing !
I hope that you can bring Michael Card to live :)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 20.06.13 at 19:17:10

@Oscar: Try using copper-tape under the cover of the socket to increase the pressure on the chip. This might help to get a plain contact to the pins of the socket. If it doesn't help no problem, the tape can be easily removed.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by val_parthen on 20.06.13 at 20:11:23

Holy cow what a great find! So that's 3 Rampage cards in collector hands now? I am supremely jealous.

Good luck with the card! Congratulations!

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 20.06.13 at 21:35:04

Thanks at all!

I added the 2 new balls but sadly chip does not work yet! :(

CHIP WITHOUT 2 BALLS

BALLS FIXED

CLEANING CHIP

At the moment chip does not work.

@Dominik, I am thinking to use a 'clamp' for getting more pressure.

Tonight more test.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 20.06.13 at 21:47:19

The Clamp tool should look like this one :

If you look closely on the first picture, the heatsink is used in combination with the clamp.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 20.06.13 at 22:21:35

@Radu,
For me that kind of clamp is very agressive. I have checked the balls under microscope and they are really deteriorated even some balls has some missing part. I think the cause is for using such clamp. I will try to take some pictures from the rampage balls. I am thinking to use some tool less agressive though.

Test socket uses a kind of golden test in U shape, it is not plane. So, it is very agressive for the balls.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 21.06.13 at 00:00:29

Oscar, try planing the balls of the chip with very fine (600 grit) sandpaper on a sheet of flat glass or a mirror.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 21.06.13 at 00:20:54

gdonovan wrote on 21.06.13 at 00:00:29:

Oscar, try planing the balls of the chip with very fine (600 grit) sandpaper on a sheet of flat glass or a mirror.

Good idea Gary!

Thanks,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by paulpsomiadis on 21.06.13 at 01:02:43

Is it leaded solder or unleaded? :-/

It REALLY should be leaded, as unleaded is what causes micro-fractures (i.e. the cause of RROD in X360) ;)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 21.06.13 at 07:57:54

Maybe re-ball the whole chip?!

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by RaverX on 21.06.13 at 09:30:51

Loeschzwerg wrote on 21.06.13 at 07:57:54:

Maybe re-ball the whole chip?!

I agree, that would be the best choice: take all the "balls" from the chip and replace them with fresh balls (try to find leaded balls, *not* "lead free", those sucks)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 21.06.13 at 11:17:23

Reballing- I would need to find a right stencil 40x40mm pitch 0.76mm. I think with a standard stencil would be enough for doing the reballing but before I will try to leave the balls surface as straight as possible with sandpaper. Yes, balls are leaded.

Rampage- At the moment Rampy does not want boots up. Seems as the chip does not get hot once connected. Now trying to find if chip takes voltage from agp connection or internal power supply but the first to do is check if really all the gold pins are in touch with PCB pads.

Here some pics.

Cleaning with ultrasonic.

Engineers did some relief on the test socket with dremel tool for avoiding resistor pack.

High size:
https://dl.dropboxusercontent.com/u/4383540/3Dfx/fotos/rampage/rampy/relief_done_dremel.jpg

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 21.06.13 at 11:32:51

Nice pics, please make one of the PCB and without the socket :)

From what I know the Rampage should be powered only by AGP, connecting the molex power will damage the card.

Does Michels Rampage chip work with your PCB?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 21.06.13 at 11:53:30

Loeschzwerg wrote on 21.06.13 at 11:32:51:

From what I know the Rampage should be powered only by AGP, connecting the molex power will damage the card.

Hank said VERY bad things would happen if you tried to plug that in.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 21.06.13 at 12:05:32

Yes, but I am talking about INTERSIL SWITCHING POWER SUPPLY HP6012CB 25A continous current. There are 4 on board to power various power rails. Not sure if Rampage chip takes the current from those or all from agp connection.

Michel's Chip does not work on my card either.

Yes, I will take some pictures without test socket onto the card.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by goriath on 21.06.13 at 12:20:07

The dark appearance of the Rampage BGA makes it look like really a bad guy.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 21.06.13 at 13:28:11

Osckhar, your chip is working in the Michael board ?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 21.06.13 at 13:33:19

m14radu wrote on 21.06.13 at 13:28:11:

Osckhar, your chip is working in the Michael board ?

I'm not sure I would want to do that, perhaps something in the board caused the chip to stop functioning. Better to make sure board is ok first.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 21.06.13 at 14:00:14

With my chip Michel's card does not work either. Sometimes I am very impatience.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 21.06.13 at 14:18:45

osckhar wrote on 21.06.13 at 14:00:14:

With my chip Michel's card does not work either. Sometimes I am very impatience.

Regards,
Oscar.

I would have been tempted too.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 21.06.13 at 17:12:06

I just have a good new about the rampy chip and it come from 3dfx engineer :)
The rampage chip take voltage for boot up from agp connection and not INTERSIL SWITCHING POWER SUPPLY HP6012CB 25AMP.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 21.06.13 at 18:14:11

trevormacro wrote on 21.06.13 at 17:12:06:

I just have a good new about the rampy chip and it come from 3dfx engineer :)
The rampage chip take voltage for boot up from agp connection and not INTERSIL SWITCHING POWER SUPPLY HP6012CB 25AMP.

Great news since it means test socket is not doing good contact if not at least chip would take heat and it is not so. Time to try with a clamp.

Card is little bended and can be the cause. Not sure if try to put it in his original stated with a preheater tool. I use this system when a PS3 /Xbox motherboard is bended but I dont like to use preheating tools on 3Dfx cards since PCB color can change of tonality. At least it happens with VSA-100 family...Not sure how can react Rampage's PCB.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 21.06.13 at 18:43:21

Yes, it is dangerous.
I would try first of all with the clamp.

Can you measure wih a multimeter if there is a proper connection betwen the agp socket and the rampage socket ?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 22.06.13 at 00:12:27

A second Ex-3dfx hardware engineer give me other information about the rampage chip.
Not the same as the other, strange.

"It get's power form the intersil chip there should be voltage on the purple side of each large silver cap."

The two ideas are not the same. We can try both of them, no?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 25.06.13 at 12:42:54

Today more tests.

I already have different clamps to test it:

@Dominik,
Here some pics

NOT TEST SOCKET:

Big size:
https://dl.dropboxusercontent.com/u/4383540/3Dfx/fotos/rampage/rampy/no_test_socket_1.jpg

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 25.06.13 at 13:04:12

Thanks :)

Before you dismounted the socket I thought it would've been solderer to the PCB. But now I can imagine why there are contact problems between PCB and Rampage chip.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 25.06.13 at 16:46:54

Today, the others chips are arrived to Oscar's home and we have a good new and bad new. Let's see with Oscar :)

My website updated.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 25.06.13 at 21:11:42

Lets hope that the chip and card are alive :)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 26.06.13 at 10:49:00

Yesterday came in another package with different rampage chip evolution.

ASAT & ASE labeled chips are test vehicles 2 layers tape based substrates while the rampage chip was converted 1 layer substrate.

The good new is the Rampage chip works perfect even with 6 missing balls

I have noticed all the chips comes in very bad condiction on the BGA balls and even all have different missing balls. My guess is they used different clamp to make perfect contact but this system is very agressive and deteriorate the bga balls.

I have tested Michel's rampage with different clamps without luck. Card continues without work. I have checked voltage and all seems ok, in intersil and condensator as directly on the bga test socket but at the moment not luck.

Question- Hank stated the resistor packs used on the card, for example RN112 was custom component for 3Dfx. It is an array composed by 5 resistors. Hank only added 4 0402 25ohm and lack one but this one is traced and connected one VIAS. I guess since Hank did not add it means it is connected to ground but I have tested it and it does not connected to gnd.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 26.06.13 at 12:13:06

osckhar wrote on 26.06.13 at 10:49:00:

Question- Hank stated the resistor packs used on the card, for example RN112 was custom component for 3Dfx. It is an array composed by 5 resistors. Hank only added 4 0402 25ohm and lack one but this one is traced and connected one VIAS. I guess since Hank did not add it means it is connected to ground but I have tested it and it does not connected to gnd.

Regards,
Oscar.

Adding a resistor there doesn't have any effect, cause there is a trace between both solder spots. Even if you solder 25ohm resistor to it, it will remain 0 ohm ;)

Did you check the damaged trace that lead from one of the AGP connections? (backside of the card)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 26.06.13 at 19:10:15

According to 3dfx engineer Oscar need to put a sheet of copper between the rearside of the cooler and the rampage chip to make a very good pressure. All the pins of the chip will be in good contact with the golds contacts of the test socket.

nb. same idea as Loeschzwerg some days ago!

EDIT: But about the tickness size 0,5mm 1mm? I don't know.

I hope it will be the solution for the card... :(

According to Oscar, the chip on my rampy now get hot while before no.
Possibly because the clamp helped and now test socket makes much better contact on the pcb.

We have a better contact on pcb now let's try to have a better contact on the other side with the chip.

@Loeschzwerg:

According to 3dfx engineer, if you look much closer, trace is fixed. No damage here.

Better RN112's pictures:

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 26.06.13 at 20:20:14

trevormacro wrote on 26.06.13 at 19:10:15:

Yep, I hope this will help.

trevormacro wrote on 26.06.13 at 19:10:15:

@Loeschzwerg:

According to 3dfx engineer, if you look much closer, trace is fixed. No damage here.

Ok, thanks for the better picture. I also checked the AGP pinout for Pin A59 -> "Reserved"

According to the AGP 1.0 spec it does have no function but within AGP 2.0/3.0 it is used for an address strobe signal. This signal is used for clock synchronization between transmitter and receiver (board and card). The strobe signal is only used on speed setting 2x, 4x and 8x within spec 2.0/3.0.

So it shouldn't prevent the card from working even if trace would be damaged.

Edit: About the copper sheets... I would go for normal "copper tape". If one layer isn't enough you can add several more layers and find the right thickness by yourself. => something like that for example: http://www.amazon.com/Copper-Conductive-Adhesive-Thick-Length/dp/B009KB86BU/ref=sr_1_1?ie=UTF8&qid=1372271318&sr=8-1&keywords=copper+tape

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 26.06.13 at 20:33:29

@Loeschzwerg:

Thanks about the agp 1.0 informations.

If Oscar doesn't have copper sheet he can use aluminium sheet (perhaps it's more easy to find in the kitchen)?

It can be really agressive for the chip or not dangerous?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 26.06.13 at 20:53:56

Aluminum sheets or tape is fine also, but I wouldn't go for aluminum foil from the kitchen. It is very very thin and you will need more layers of it (and folding of aluminum foil sucks).

http://www.amazon.com/Duck-240225-1-88-Inch-50-Yard-Aluminum/dp/B0000DI82J/ref=sr_1_3?ie=UTF8&qid=1372272689&sr=8-3&keywords=aluminum+tape

Just a few dollars like the copper tape too.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Dexter on 27.06.13 at 08:38:11

I've read all 5 pages and all I can say is Respect !
There's a noble 3dfx fan right there.

Now I don't want to barge in rudely on this sensitive electronical issue, but has anyone ever thought of the oven trick? Solely for the card, without the chip.

Can't it repair eventual "cold solderings" which are possible ?
Or is it too risky ?

...just saying.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 27.06.13 at 09:46:42

The oven trick mainly works for soldered BGA components. Beside this it can fix the connection from DIE to substrate.
Both isn't the case when it comes to the Rampage card.

So no, I wouldn't recommend this trick.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 27.06.13 at 11:07:00

Loeschzwerg wrote on 27.06.13 at 09:46:42:

I'm surprised that no one has suggested doing away with the socket yet. Just attach the chip to the board and be done with it.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 27.06.13 at 11:14:00

If the "copper sheet"-trick doesn't work, soldering the Rampage to the board should be considered, yes :)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 27.06.13 at 11:42:56

gdonovan wrote on 27.06.13 at 11:07:00:

Loeschzwerg wrote on 27.06.13 at 09:46:42:

I'm surprised that no one has suggested doing away with the socket yet. Just attach the chip to the board and be done with it.

LoL, yesterday I recommended it to Michel soldering the chip onto the card. He must think on that.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 27.06.13 at 12:01:16

@Loeschzwerg:

Oscar tried with double paper sheet. He already had problems to close the heatsink but card continues without working. Even used the clamp but the card does not want to boot. With double paper sheet heatsink is doing a lot of pression and even balls are more deteriorated. Did you think it's enough or he need to buy copper sheet?

Check rampy chip picture:

It looks bad but it works perfect though. The pression is deteriorating some balls.

@osckhar:

If you put the chip on the rampy and after that the card doesn't work, is it possible to come back to original state without damaged both parts (balls traces on pcb and pins balls for the chip)?

Soldering the chip will be done by you?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 27.06.13 at 12:13:33

trevormacro wrote on 27.06.13 at 12:01:16:

If you put the chip on the rampy and after that the card doesn't work, is it possible to come back to original state without damaged both parts (balls traces on pcb and pins balls for the chip)?

It is not 100% possible to return to the original state. After soldering the chip to the PCB, the solder spots one the PCB are tinned forever. This mean if you unsolder the chip and mount back the socket it won't have contact to the gold surface it now has, but to a tinned surface => the conductivity isn't that good compared to gold.

Edit: Yes, I think the pressure was enough then :(

Edit2: Btw. I would rework the missing solder balls on your 2nd Rampage chip. It mainly works cause the traces lead to other solder balls with the same trace too. But it might cause loading problem (ground area and voltage supply, as well as signaling).

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 27.06.13 at 12:24:29

Loeschzwerg wrote on 27.06.13 at 12:13:33:

trevormacro wrote on 27.06.13 at 12:01:16:

If you put the chip on the rampy and after that the card doesn't work, is it possible to come back to original state without damaged both parts (balls traces on pcb and pins balls for the chip)?

When Oscar tried my test socket and my rampage chip, it worked on his card but on my card no.
Perhaps the problem come from another part?
Did you think it will be a good idea to solder the chip on my card?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 27.06.13 at 12:29:47

trevormacro wrote on 27.06.13 at 12:24:29:

Well it's still possible the socket doesn't have a good contact to the PCB. You mentioned the PCB is warped a little bit, right? Soldering might help here, but we can't say for sure. Still a problem on other components possible (vga bios, smd components, voltage supply).

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 27.06.13 at 12:31:23

trevormacro wrote on 27.06.13 at 12:01:16:

@osckhar:

If you put the chip on the rampy and after that the card doesn't work, is it possible to come back to original state without damaged both parts (balls traces on pcb and pins balls for the chip)?

Soldering the chip will be done by you?

Always there is a risk. I cant tell you 100% not risk but I have reemplaced several bga chip from 3Dfx cards without problems. I have done some reworks (vsa100 / tsops) for Gino and all perfect. No cards damaged.
Lately I had to fix a Daytona agp with problems in memory ddr and bga (reemplaced them) and card is perfect now.

About If is possible once soldered the chip come back his original stated. Yes, it is possible.
Is possible reballing the chip with new balls? Yes, it is possible. Anyway I should build custom stencil from a standar stencil 40x40mm for rampage chip.

Now the bga pads from rampage card are not leaded they are in his orginal copper stated.. Once I Solder the chip onto the card if I remove the chip pads willbe leaded but that is not bad since it can help makes better contact.

Yes, I would do all the process with my BGA reworks station.

Michel, is your choosing. You must think it and even talk with Hank about this. Always there is a risk and if something goes bad I dont want any responsability.

If you ask me what would I do? You know me and I would try but the main problem is if really the problem comes from there.

I know very difficult decission.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 27.06.13 at 12:42:25

osckhar wrote on 27.06.13 at 12:31:23:

Now the bga pads from rampage card are not leaded they are in his orginal copper stated.. Once I Solder the chip onto the card if I remove the chip pads willbe leaded but that is not bad since it can help makes better contact

Conductivity goes down:
http://metals.about.com/od/properties/a/Electrical-Conductivity-In-Metals.htm (compare lead to copper)

But I assume it won't have an impact :)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 27.06.13 at 13:41:46

@Loeschzwerg:

According to Oscar PCB is warped a little bit. Bios is tested by him on his card and it work.

@osckhar:

This afternoon, I will try to contact Hank and ask about it.
If I have an answer from him I let you know because it's hard to speak with him. He is very busy.
You can confirm and post a screenshot of the bended part please?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 27.06.13 at 14:14:28

What I would do is reballing the first rampage chip 'dead' and check how the traces and substrate react. I guess the traces and substrate come with a special protection and I imagine I can work there since I must use the soldering iron to take out the leaded and clean the zone with solder wick copper.

@Dominik- PCB is little warped but not sure if the cause for bad contact but soldering the chip is the best option to know if is or NOT a contact problem.

About adding the missing balls 2ND RAMPAGE CHIP would be not a problem since I can add them and keep the presents with patience.

@Michel- keep me informed about what Hanks tells you.

NOTE- All the parts from Michel's cards works ok on my card. I mean test socket, chip and bios.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 27.06.13 at 18:20:30

If all other parts are working, then there should be something with the board or with the chip connection.

Pressing more the chip would cause damaging !

Michael: if you ask me, I would solder the chip !
There is nothing to lose in my opinion.
Unsoldering is always possible.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 28.06.13 at 11:02:05

Hi,

Yesterday I am trying to contact Hank without luck.

@osckhar:

I take my decision now.
I accept you put the chip directly on the pcb.
I will take some risks but I want to take them to progress :(
Even if the card doesn't boot with that you can leave in this state and search others parts to find the problem.

Thanks :-/

@m14radu:

Thanks for your support.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 28.06.13 at 11:08:37

trevormacro wrote on 28.06.13 at 11:02:05:

Oky! This afternoon I will start to work in your card.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 28.06.13 at 11:14:19

Good luck!

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 28.06.13 at 11:18:54

[OFF TOPIC]

Talking about bad contact.

Here with my Rampy card.

Bad contact- NO DONGLE.

Good contact with clamp- NO DONGLE. Rows present since one memory connection is broken.

Driver loaded- HW ACCELERATION OFF - DONGLE

Locked HW ACCELERATION NOTCH 3 (SLIDE)

Why rows are no present once driver is loaded... driver is messed up as Hank stated.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 28.06.13 at 12:00:15

Could you please do a comparison shot of 640x480 @60hz and 800x600 @60hz (windows display setting). Thx :)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 28.06.13 at 12:14:14

osckhar wrote on 28.06.13 at 11:18:54:

Why rows are no present once driver is loaded... driver is messed up as Hank stated.

Regards,
Oscar.

I think you have either a contact issue, a bad ram module or compatibility with the motherboard.

The ONLY time I had an issues with any rows on the screen was with the KT-133 or BX with USB on and driver loaded or the ASROCK BX motherboard where it would do it all the time.

Rampage is VERY picky about motherboards.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 28.06.13 at 13:26:36

@Dominik- yes, no problem. This night I will do a comparison.

@Gary- I have tried many motherboards in all motherboards but those vertical lines are present.

Seems my Rampy card comes with an old bios.

MICHEL'S BIOS DATED 07/12/2000

OSCAR'S BIOS DATED 05/12/2000

Gary's Rampage bios is dated 07/12/2000 as well. So, my bios is only 2 days old than yours.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 28.06.13 at 14:53:15

@Michael: In my opinion is the best decision in that case !
Good choice mate !
Lets hope that the card is running - I keep my finger crossed :)
Osckhar: try to swap the bios on your card with the one from Michael card :)
Do not flash you bios yet.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 28.06.13 at 15:18:41

m14radu wrote on 28.06.13 at 14:53:15:

Yes, Radu I already did it. I used my bios on Michel's card and vice versa with not changes.

Soon we will know if Michel's card is working.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 28.06.13 at 15:20:48

I keep my finger crossed Osckhar !
::)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 28.06.13 at 16:27:27

m14radu wrote on 28.06.13 at 15:20:48:

I keep my finger crossed Osckhar !
::)

Thanks Radu!

At the moment all perfect. I added 6 new balls rampage chip. Now little ultrasonic cleaner and ready for soldering the chip onto SPECTER card.

Later pics.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 28.06.13 at 16:52:57

Once added new balls rampage chip works perfect in my test socket specter card. Now ready to starts soldering chip. Nervous here!

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 28.06.13 at 17:25:17

Chip soldered onto the card. Process has gone perfect. Now card need to be cooled. 8-)

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 28.06.13 at 17:33:12

I cross my fingers too.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 28.06.13 at 17:36:12

You make a great work Osckhar :)
im nervos too ::)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 28.06.13 at 17:53:45

Bad news guys! Card continue without work... >:(

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 28.06.13 at 17:55:33

damn.... :o
is the chip hot ?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 28.06.13 at 18:18:08

osckhar wrote on 28.06.13 at 17:53:45:

Bad news guys! Card continue without work... >:(

:( I am sad.
What is the next step?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 28.06.13 at 18:22:15

Checking again the card. Balls seems very good soldered. Anyway I will try a reflow on the chip. I wish it helps!

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 28.06.13 at 20:01:48

Nothing! Very dissapointed here! :'(

Seems as chip does not wake up. It is getting hot but not too much.

For example in epox 8k5a2+ error postcode 25.

It always gets same error with test socket and now with chip soldered onto card.

Would be very important to know more details about the card. If it stoped to work because RN112. If card was fully working once left 3Dfx. If we had at least the scheme internal Power Supply... :(

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 28.06.13 at 20:32:30

osckhar wrote on 28.06.13 at 20:01:48:

Would be very important to know more details about the card. If it stoped to work because RN112. If card was fully working once left 3Dfx.

Can RN112 part be harvested from another board like a 4500? Or does the three resistors that replace it properly do the job?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 28.06.13 at 20:51:31

What about measuring the input & output values on RN112 part on both cards ?
If they are the same, then the problem is somewhere else.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 28.06.13 at 20:54:52

gdonovan wrote on 28.06.13 at 20:32:30:

osckhar wrote on 28.06.13 at 20:01:48:

Would be very important to know more details about the card. If it stoped to work because RN112. If card was fully working once left 3Dfx.

Can RN112 part be harvested from another board like a 4500? Or does the three resistors that replace it properly do the job?

No since those resistor pack are custom for 3Dfx and no other card uses such resistor pack. According to Hank with the 4 smd resistor soldered at the card should do the work as well.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 29.06.13 at 18:47:08

Website updated with news pictures of the Oscar's rework.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 01.07.13 at 14:50:47

If you look more closely on this picture you will see one important thing, a numbered chip.

http://img15.hostingpics.net/pics/137446chip24.jpg

It means at least there were 24 rampage chips.

Must clean the others chips and later check if there is a number on them.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 02.07.13 at 11:05:23

osckhar wrote on 28.06.13 at 13:26:36:

Seems my Rampy card comes with an old bios.
Gary's Rampage bios is dated 07/12/2000 as well. So, my bios is only 2 days old than yours.

Regards,
Oscar.

Just noticed this post Oscar- e-mail me your BIOS and I'll reinstall my hex editor program which will let me compare the two. This will give us vital information about how and where to make BIOS changes.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 02.07.13 at 11:19:19

I was lucky, Hank answer me.

According to him,

Check Y1, the big silver component, with a scope you should see 14.3MHz clock. This is important or the chip core will not run.

Check voltages across the capacitors marked bellow. Record them and compare them to the working Rampage.

3 things are necessary to get the chip going.

The Power, the clock and the reset. Look at these first.

According to Oscar,

Checked voltage:

Michel's rampy / Oscar's rampy

C252 - 1.280 - 1.265
C276 - 2.360 - 2.373
C210 - 1.915 - 1.922

EDIT:

C232 - 3.275 - 3.290

Voltage seems correct.

About crystal Y1 - He doesn't have any tool here for MHz mesure. His multimeter dont have such option and he doesn't own any osciloscope though. Any tip? a cheap osciloscope or multimeter with such option.

Regards,
Trevor.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by ultima on 02.07.13 at 11:48:53

seems to me that the C components are operating within their respective ranges, so close to eachother.....it would suprise me if those 3 were the problem.
So, asuming the chip itself is in working order, the 14.3Mhz part would be the problem.

As for getting a reader for it, depends, what is cheaper..getting one locally yourself or ship one to you from someone elses place and probably pay a load on shipping, which in the end might be just as expensive.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 02.07.13 at 11:54:04

ultima wrote on 02.07.13 at 11:48:53:

COULD be the problem.

It might be easier swapping with a known good part.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 02.07.13 at 12:51:29

Voltage is fine. An oscilloscope can be very expensive, especially if you want to measure frequencies above 10 MHz (and that is the case).
Unfortunately my multimeter is also only capable of measuring up to 10MHz :-/

BUT, I've got one or two NEW 14.3MHz crystals left from the repair of exxe's Barracuda card: http://abload.de/thumb/after_319lvq.jpg

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 02.07.13 at 12:59:03

@Dominik- Are you interested to part with one crystal? Please, talk with Michel.

Just I dont like the idea to change smd parts without know if it works or not but... I think it is the cheap way but possibly this is the problem since internal power supply works ok.

@Gary- I sent you a copy bios.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 02.07.13 at 13:34:54

osckhar wrote on 02.07.13 at 12:59:03:

Just I dont like the idea to change smd parts without know if it works or not but... I think it is the cheap way but possibly this is the problem since internal power supply works ok.

Me too, but it really is the cheapest way unless someone you know is lending you an "osci".

Sure, you can have the crystal.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 02.07.13 at 13:39:37

Loeschzwerg wrote on 02.07.13 at 13:34:54:

osckhar wrote on 02.07.13 at 12:59:03:

Just I dont like the idea to change smd parts without know if it works or not but... I think it is the cheap way but possibly this is the problem since internal power supply works ok.

Me too, but it really is the cheapest way unless someone you know is lending you an "osci".

Sure, you can have the crystal.

Thank you so much!!!

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 02.07.13 at 15:35:40

http://abload.de/thumb/14.3-crystal_137s5o.jpg http://abload.de/thumb/14.3-crystal_2hgshl.jpg

Btw. there is no crystak with 14.3MHz, they are always 14.31818MHz ;)

I can't find the datasheet of the original crystal, so there might be a little tolerance when it comes to the frequency (0.00xMHz) and the load capacity. This will have no negative impact if you ask me.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 02.07.13 at 16:09:21

Loeschzwerg wrote on 02.07.13 at 15:35:40:

High res. pictures of my Y1 with ref. A143G0Y:

If it's not dangerous for the rampage card, Oscar can put it on my card.
You have a PM.
Can you send him today?

Thanks for help us.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Komponent on 02.07.13 at 17:43:21

Voodoo5 cards are using something very close: A143E0Y
http://upload.wikimedia.org/wikipedia/commons/6/61/KL_3DFX_Voodoo5_5500.jpg
Just take one off a working card, put in on the Rampage and see what happens. That's what I would do anyway...

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by ultima on 02.07.13 at 17:57:18

Before doing that I would check it's specs first if they are also close in operation, not just in number

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 02.07.13 at 18:18:16

Komponent wrote on 02.07.13 at 17:43:21:

Hi Komponent and thanks for your help. you have a good idea.
No risks to damage the card with the Y1 of the Voodoo5?
What are the differences btw A143E0Y and A143G0Y?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Komponent on 02.07.13 at 18:40:57

The only thing the crystal does is to create a reference clock signal at 14.3Mhz. They are more or less generic parts, you'll hardly find datasheets for them. Replacing is very easy and I can't think now on any scenario that will damage a card when mounting a working crystal on it.
Better on, if you only want to test the crystal, and not use the Rampage, unsolder it, and take a common working video card that has a similar part on it, unsolder the crystal on that card too, put the one from Rampage in it's place and test that card. Yeah, put the crystal from Rampage on a Voodoo5 and see if that works.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 02.07.13 at 19:14:55

Quote:

What are the differences btw A143E0Y and A143G0Y?

I now know what the marking means :)

A TUH ZY R

A -> stand for the Abracon Corporation
TUH -> Frequency
   T = First "ten" digit of frequency
   U = First "unit" digit of frequency
   H = First "tenth" digit of frequency
Z      -> Month letter ( A to L )
Y      -> Last digit of year
R      -> Traceability code ( character A to Z )

So A143G0Y means:
14.31818MHz July 2000

So if you ask me: The crystal of every Voodoo 1/2/3/4/5 works as well as the one I still have at home.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 03.07.13 at 14:39:24

Is a Smart Tweezers can measure the Crystal Y1 frequency?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 03.07.13 at 14:52:27

Depends on the smart tweezer, it must be capable of measuring more than 10MHz (a common limit on cheaper multimeters an tweezers).

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 03.07.13 at 16:28:52

Loeschzwerg wrote on 03.07.13 at 14:52:27:

Depends on the smart tweezer, it must be capable of measuring more than 10MHz (a common limit on cheaper multimeters an tweezers).

This one:

http://es.farnell.com/ideal-tek/st-5/meter-lcr-2-5x14x3mm-53g/dp/2064744

Is it good?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by goriath on 03.07.13 at 16:41:30

It's an LCR meter, it means is suitable only to measure inductance capacitance and resistance.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 03.07.13 at 21:02:51

@Osckhar: did you change the crystal ?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 04.07.13 at 01:50:05

m14radu wrote on 03.07.13 at 21:02:51:

@Osckhar: did you change the crystal ?

Oscar tried the Rampage crystal on a Voodoo4 and it work.
After that, he tried the Voodoo4 crystal on the Rampage and it doesn't work. :(
The crystal Y1 is not the problem.

My website is updated for pictures of the rework.

Next step:

Oscar wait the smart tweezer to test all components and compare them to his card.

If anyone have any solutions feel free to post them please :(

How is possible the orignal chip was dead. It can be a clue why the card does not work.
Perhaps chip does not work when it was placed bad in the socket and for such reason burned some components even the rampage card.

History of the RN112 on the card:

RN112 was missing because it was used to repaired other rampage card. The 4 resistors added instead of RN112 are good for the card. It's not the problem.

Regards,
Trevor.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by goriath on 04.07.13 at 14:23:42

trevormacro wrote on 04.07.13 at 01:50:05:

How is possible the orignal chip was dead. It can be a clue why the card does not work.
Perhaps chip does not work when it was placed bad in the socket and for such reason burned some components even the rampage card.

History of the RN112 on the card:

RN112 was missing because it was used to repaired other rampage card. The 4 resistors added instead of RN112 are good for the card. It's not the problem.

Regards,
Trevor.

Honestly I have some difficulties to understand what you tell time by time. I may be wrong but I remember your BGA worked on the Oscar's board, isn't it? So, nobody can be sure at the moment but your BGA should work.

Since PCB is warped, that could mean some internal electrical traces are broken, or maybe not, you never know.

RN112 was missing because it was used to repaired other rampage card - This statement could be a clue about the condition of your board. Maybe that card was unfunctional already when left 3dfx facilities, so they might have used it for spare parts.

But, if RN112 can be replaced by resistors, why they didn't use resistors to repair the other Rampage?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 04.07.13 at 15:31:50

Quote:

RN112 was missing because it was used to repaired other rampage card - This statement could be a clue about the condition of your board. Maybe that card was unfunctional already when left 3dfx facilities, so they might have used it for spare parts.

I was thinking the same thing, logic would lead one to that conclusion.

It did pop in my head that the pack could have been used on an unsocketed board as one would imagine that they were a bit more stable than the socketed units.

Just a passing note before I have to head back to the hospital, Hank mentioned during a phone call I had with him that all the Rampage chips made tested live, there was no failures. That doesn't mean though that after 12-13 years they have not been hit by static electricity once or twice.

But if the chip tested good on Oscars board then the problem lies somewhere else.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 04.07.13 at 16:45:55

goriath wrote on 04.07.13 at 14:23:42:

trevormacro wrote on 04.07.13 at 01:50:05:

The first rampage chip who came with the card doesn't work for now.
The second chip I had works on oscar's rampy.
Good question about the RN112 but no answer for now.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 04.07.13 at 17:48:58

Why we don't measure the RN112 Input/Output voltages on Osckhar Card, and compare them with the RN112 Input/Output voltages from Michael Card ?
If there are differences we have a clue :)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 05.07.13 at 01:28:11

m14radu wrote on 04.07.13 at 17:48:58:

Why we don't measure the RN112 Input/Output voltages on Osckhar Card, and compare them with the RN112 Input/Output voltages from Michael Card ?
If there are differences we have a clue :)

Oscar have checked it in parallel mode to compared with the others resistor pack. The value is different and in the others is the same (the only different is the individual resistence Hank added).
He also checked with his card and all the resistor pack have same value.
According to Hank it's not the problem because it can be bypassed with individual 25ohm 0402 resistors and tested. But since this is the first card in the lab, he think it would be inappropriate to modify it.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Tim on 05.07.13 at 10:17:28

Keep up the good work guys. Congrats on the Rampage trevor. Hope you get it to work in the end.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 05.07.13 at 11:17:56

If the values are different between the cards, swapping the RN112 from a runnig care to yours could reveal the thruth.
I know they are ptototypes an a swap is not recommeded :)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 05.07.13 at 12:00:56

Yupe, many questions without answer.

- Why chip is dead? ESD problems? Yes, it could be but I guess if chip is placed in bad postion since a lot of traces on the substrates are connected 'possibly' some PCB smd part could be 'burned' for NOT right voltage. It could explain why chip is dead and why card is NOT working.

- RN112- Odd about such resistor pack. Why 3Dfx ordered a custom resistor pack when with individual smd resistence has same effect? reduce costs? I either understand why 3Dfx requested a custom resistor pack when with individual resistence gets same results.

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Quaker763 on 06.07.13 at 09:33:17

goriath wrote on 04.07.13 at 14:23:42:

Since PCB is warped, that could mean some internal electrical traces are broken, or maybe not, you never know.

I guess it would depend on how badly the card is warped. I agree with goriath in that it actually may be the board not the chip or components. Does anyone have any idea how the card got warped in the first place?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 18.07.13 at 16:14:50

@Quaker763:

No idea.

@All:

Website updated!

The chip is now unsolded from the pcb.
Need to reball now.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 24.07.13 at 17:09:16

My Rampy card is return to his original state and in good condition.

Website updated with a picture.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Tim on 25.07.13 at 09:29:55

trevormacro wrote on 24.07.13 at 17:09:16:

My Rampy card is return to his original state and in good condition.

Website updated with a picture.

So what's next? Or is that it?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 25.07.13 at 11:19:54

Idea is the following:

Across an agp riser, an estructure and with a multimter to measure voltage pin by pin (gold test socket pin) and later to compare it with my Rampage. I still thinking the Rampage Chip does not get the right voltage. If I find what pin is not good I can follow traces and check what V-rail is doing the problem and work there.

Main problem- I dont have enough time free to work there. Summer is very busy here.

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 03.10.13 at 11:42:10

Rampage CHIP re-balled with great success. It works perfect with new balls. Rampy #B001 comes back his original stated.

Not too much time for working on Rampy but the most important is 4 POWER RAILS gives the right power, one for CHIP the others 3 for DDR memory. This is the most important. Only I wish the card has not broken any internal trace.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 31.10.13 at 12:45:43

Arrived the moment to take an eye at this baby.

First test:

Checking if pads /socket gets right voltage.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 31.10.13 at 18:35:28

OMG- I checked all the test pads and voltage are perfect. Now, I need to find where is the SWITCH makes the card wakes up.

Now testing values Ohm across of each pin/pad hoping to find an alterated vaule that can show me the way.

Regards,
Oscar

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 02.11.13 at 08:56:19

any news Osckhar ? :)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 11.11.13 at 11:10:36

m14radu wrote on 02.11.13 at 08:56:19:

any news Osckhar ? :)

Yes, seems I found the problem. :D

I need to test it today.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 11.11.13 at 11:57:15

osckhar wrote on 11.11.13 at 11:10:36:

m14radu wrote on 02.11.13 at 08:56:19:

any news Osckhar ? :)

Yes, seems I found the problem. :D

I need to test it today.

Regards,
Oscar.

Fingers crossed!

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 11.11.13 at 12:16:13

I have detected an anormal value using the multimeter setting ohms across the test socket pins compared #B009. This afternoon I will follow the trace from the pad and I will see where it goes. I hope to find the problem there.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 11.11.13 at 14:09:07

osckhar wrote on 11.11.13 at 12:16:13:

Good start with this find :)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 12.11.13 at 22:00:37

I have a good feeling...!

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 12.11.13 at 22:23:07

8-) 8-) 8-)

FOUND THE PROBLEM- I checked pad by pad with a multimeter set in ohms until I detected an anormal value. Following the trace I detected one resistor 100ohm soldered in wrong place. I unsolder and solder it in right place and voila!!! Card waked up!!!! We have antoher 100% functional card!!!

Really I still cant believe it!!!! :D :D :D

ps. My friend you already own a 100% fully functional Rampy card. EnJoy it and take care of the QUEEN!!! 8-)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Komponent on 13.11.13 at 05:17:43

Superb work Oscar, congratulations Michel!

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by ultima on 13.11.13 at 07:16:20

AWESOME!!!! I have no other word for it.

So how many functional Rampies do we have in the world now as far as we know??

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 13.11.13 at 08:13:19

Wow, amazing 8-)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Jean-Pierre Subtil on 13.11.13 at 08:32:30

Awesome job osckhar!

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 13.11.13 at 08:44:42

Thanks my friend. Even Hank Semenec didn't found this problem.
You do a very good Job. The Rampy is now in working condition.
Need to test it now with 3dmark2001se and others games to see this baby running very well.

Thanks again to take your time.

My dream came true :)

EDIT:

ps. can I see a picture of the modification please?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by EMPEROR on 13.11.13 at 10:42:57

Great work Oscar!!! Hands down to everything you do!

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Thandor on 13.11.13 at 12:13:45

Very nice job! Resistor is futile (well... sort of)!

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 13.11.13 at 13:08:46

Great work Osckhar !

Now Michael the second rampage must wake up.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 13.11.13 at 14:35:10

m14radu wrote on 13.11.13 at 13:08:46:

Great work Osckhar !

Now Michael the second rampage must wake up.

If Rampy A0 wake up it will be AMAZING!

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by paulpsomiadis on 13.11.13 at 14:37:31

Job's a good 'un! 8-)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by vik on 13.11.13 at 16:09:42

congratulation :D

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 13.11.13 at 23:00:16

ultima wrote on 13.11.13 at 07:16:20:

AWESOME!!!! I have no other word for it.

So how many functional Rampies do we have in the world now as far as we know??

5 Functionals Rampys:

- Two for Osckhar.
- One for Nightbird.
- One for Garydonovan.
- One for me.

with so many others hidden in the world. Try to search them :)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by ultima on 14.11.13 at 11:21:34

so many is a big word, by last official count, no more then 20 were ever produced in total, 10 A0 and 10 A1 revisions.

The chances of one of those remaining 15 cards even being close to me is astronomical, let alone the chance that the person who has it, doesn't know what he has and therefor making if financially more available to me :)

I mean, I don't have the means to shell out 15K for a card, my girl would kill me LOL. I can't even get a 6000 for that matter, so.....

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 14.11.13 at 19:47:57

and how many nonfunctional ?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 14.11.13 at 20:31:14

m14radu wrote on 14.11.13 at 19:47:57:

and how many nonfunctional ?

One with me (Rampy A0).
And for the others I have my secret idea.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by ultima on 14.11.13 at 21:21:16

Care to share the secret idea?? :)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 14.11.13 at 21:31:02

ultima wrote on 14.11.13 at 21:21:16:

Care to share the secret idea?? :)

A secret is a secret :( I can't.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by ultima on 14.11.13 at 21:44:09

So you know someone who has one or more people who have one and have asked you not to tell?? Ok, then I will not ask anymore. :) secret is secret afterall. :D

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 14.11.13 at 22:16:39

Come on Michael. Whats the secret ? ;)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by ultima on 14.11.13 at 23:13:25

The secret is he won the lotto and actually owns as many rampages as he does 6000's. :)

Am I getting close?? :D

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 15.11.13 at 00:00:16

I close my mouth lol.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 15.11.13 at 05:33:42

@Michael: how many ? :)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 15.11.13 at 16:24:05

@m14radu:

More than one.

@all:

I have at home a TYAN S2469UGN S2469 Thunder K7X Pro Dual with 2 Athlon MP 2800+ motherboard.

http://img11.hostingpics.net/pics/635235First3.jpg

Is it compatible with my Rampy001? without make some damages when the card running.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by ultima on 15.11.13 at 16:34:08

Damn!!! That is one big ass board. Lol

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 15.11.13 at 17:16:34

Maybe he find the Rampage Nest :D

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by ultima on 15.11.13 at 22:44:07

trevormacro wrote on 15.11.13 at 16:24:05:

When someone says more then one, it is always at least 3. :)

Seeing as to what you keep buying, I%m starting to get the impression you must be almost a millionaire if you got the cash to buy that many 6000's and what seems to become 5 rampies in total at least......you couldn't make me more jealous, and I think I am not the only one. LOL

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 17.11.13 at 09:26:32

@Michael: you already get the repaired rampage board from Osckhar ?
We want pictures and benchmarks results :D

Btw: is now the gpu soldered on the card ? or like it originally came, with socket ?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 17.11.13 at 11:37:38

m14radu wrote on 17.11.13 at 09:26:32:

@Michael: you already get the repaired rampage board from Osckhar ?
We want pictures and benchmarks results :D

Btw: is now the gpu soldered on the card ? or like it originally came, with socket ?

Hi, I am waiting to receive it.
If you look my website you will see the card is in original state with socket.
For sure I will make a video when i will have the card.
Have a good day radu.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 19.11.13 at 15:14:32

Before to be shipped, here is the first Rampage A1 built:

Here is the working chip numbered:

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 21.11.13 at 11:11:30

I have received the package.

I have tested the Rampy in 2 motherboards without luck. NO BOOT :(

If anybody can help me? or perhaps my two MB are not compatibles with Rampy and perhaps it's explain my Rampy A0 doesn't work too. It's just an idea.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 21.11.13 at 11:54:38

Hi Michel,

Are you sure?

What motherboards are you using?

Rampy #B001 worked as rock here. Totally stable in all the motherboards I used.

Did you check if I inserted the right Rampage chip on the Rampy?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 21.11.13 at 12:31:33

osckhar wrote on 21.11.13 at 11:54:38:

I use two MB:

- Tyan K7x pro (first)
- MSI 848P Neo-V (second)

I didn't removed the chip inside the rampy I just connect the rampy on my motherboards. I don't check inside the socket. I just see in other small box the original chip.
Next to the rampy there is a rampage 001 chip original.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 21.11.13 at 12:39:34

The original Rampage chip comes with Rampy #B001 is in dead condition.

Test #B001 in other motherboard if you can. If not Take out the Rampage chip from Rampage #B001 and place the Rampage chip with open die and test it again.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 21.11.13 at 12:44:12

osckhar wrote on 21.11.13 at 12:39:34:

I don't have others motherboards.
Open die mean what?
Did you think my MB are not compatible with the rampage?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 21.11.13 at 12:51:55

Open die means Chip de-capped.

What I know for sure is RAMPAGE CARD is in fully condition- so not worry for that. Or motherbards are not compatible or Rampage chip is dead. I dont think your motherboard 'kill' Rampage chips. ;)

You dont own there a normal EPOX board based KT333 chipset?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 21.11.13 at 13:13:51

osckhar wrote on 21.11.13 at 12:51:55:

Why my full rampage chip can be dead? if you had tested it before the shipment.
What is the best MB for Rampage? give me the full name please.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 21.11.13 at 13:36:25

I still dont know what is the best motherboard for Rampage. I have used serveral here. Since an ABIT SA7, to Elitegroup based KT333 board as DFI LANPARTY NF3 WITH AGP support, EPOX and ASUS.

When a Rampage does not boot here. I only needed to be sure test socket makes good contact tightening the 4 nuts for making perfect contact. But I dont think it is needed since card worked ok here... Anyway you can check it.

ps. I think the next time I will send you the motherboad as well.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 21.11.13 at 20:09:11

I have removed the chip and reput it again.
Now the card boot!
I think a bad contact between socket and chip.

But it's very dark for the screen why? perhaps i need to increase the brightness?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 21.11.13 at 20:58:05

Excellent!!! :D

I never needed the dongle for installing Win98SE with my LG LCD. I was enouh clear for reading the instructions.

Did you try to install win98se?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 21.11.13 at 23:20:00

osckhar wrote on 21.11.13 at 20:58:05:

Excellent!!! :D

I never needed the dongle for installing Win98SE with my LG LCD. I was enouh clear for reading the instructions.

Did you try to install win98se?

I have LG lcd too.
Sometime I get black screen sometime I get a very dark screen.
Not yet.
I will upload some picture tomorrow.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 22.11.13 at 08:56:11

Install Win98SE using another 3D VGA. Once installed, plug in the Rampy card. Once in windows image will be more clear and you will not need dongle.

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 22.11.13 at 12:30:54

Good morning everybody,

Here are some screenshots:

Hardware:

Booting before setup bios:

I had one HDD with Windows 7 already installed (just to test the color but very bad no?):

Here are some videos:

Listen the fan :)

http://youtu.be/tLmecBsUgLE

Booting...

http://youtu.be/SE926_1ADKQ

Setup Bios

http://youtu.be/EDWG8WvcwGM

Start Windows Seven

http://youtu.be/FzxbotI9SaU

I have some questions:

- Why in Bios the rampage look good without dongle and in Windows look very bad?

- Is it better with a Abit VA6 instead of my Tyan K7x Pro?

- Sometime when I push on the PC the rampy doesn't boot, why? (when i reput the chip and push on the heatsink it works)

Thanks in advance.

Trevor.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by EMPEROR on 22.11.13 at 12:35:53

Why are you turning on the system without the external power supply for the card turned in? Doesn't the Rampy require the extra molex to work?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 22.11.13 at 12:46:42

EMPEROR wrote on 22.11.13 at 12:35:53:

Why are you turning on the system without the external power supply for the card turned in? Doesn't the Rampy require the extra molex to work?

From Garydonovan website:

http://www.thedodgegarage.com/3dfx/rampage.htm

" The Molex connector for extra power is being blocked by two capacitors! The reason for the Molex connector was that since the chip was 100% modeled in software it was unknown if the AGP connector could provide enough power.

As it turned out the extra power was not required which is just as well since a snafu in the PCB layout ended up blocking the connector anyway.

NOTE: I'm told by an engineer that hooking power to the connector will cause something very unpleasant to happen to the board. So if you come across a Rampage board, don't try plugging power in here, the board runs just fine without it. "

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 22.11.13 at 12:51:55

Quote:

I have some questions:

- Why in Bios the rampage look good without dongle and in Windows look very bad?

- Sometime when I push on the PC the rampy doesn't boot, why? (when i reput the chip and push on the heatsink it works)

Boot process uses different INIT VGA screen. Some helps to get better image quality and others not.

Try to put a second copper or aluminium tape on the base.

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 22.11.13 at 12:55:25

osckhar wrote on 22.11.13 at 12:51:55:

Quote:

Boot process uses different INIT VGA screen. Some helps to get better image quality and others not.

Try to put a second copper or aluminium tape on the base.

- Oscar.

Ok for the tips.

Do you think the color in Windows Seven are good without the dongle? comparing to your PC (without drivers installed).

If I install on Windows 98se the rampage drivers it will be better for the color? always without dongle.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by EMPEROR on 22.11.13 at 12:57:09

trevormacro wrote on 22.11.13 at 12:46:42:

NOTE: I'm told by an engineer that hooking power to the connector will cause something very unpleasant to happen to the board. So if you come across a Rampage board, don't try plugging power in here, the board runs just fine without it. "

Wow... good to know... Thanks for the info!

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 22.11.13 at 13:02:25

Quote:

Ok for the tips.

Do you think the color in Windows Seven are good without the dongle? or not.
If I install on Windows 98se the drivers it will be better for the color? always without dongle.

We will not know it until you do it! Install Win98SE!!!!! :D

So is how looks Win98SE with Rampy #B001 with my Sony LCD.

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 22.11.13 at 13:27:48

I think on this picture,

Motherboard is an Intel 'Seattle' SE440BX-2 Motherboards.

Look this picture, right?

Regards.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by EMPEROR on 22.11.13 at 14:16:37

Yes, this is the correct motherboard.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by ultima on 22.11.13 at 15:46:18

That is not the same mobo, connectors at the back for audio and such are in different places.

Btw emperor, did you read my.pm??

Edit

I am blind it seems, is the same, just different color. Lol

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 22.11.13 at 16:29:13

ultima wrote on 22.11.13 at 15:46:18:

That is not the same mobo, connectors at the back for audio and such are in different places.

Btw emperor, did you read my.pm??

Edit

I am blind it seems, is the same, just different color. Lol

I think the shadow let you think it's not the same color.
95% it's the same motherboard.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by EMPEROR on 22.11.13 at 16:33:18

It's not 95%, it's 100% the same board. What I was thinking though, was if there is a proof they actually managed to get Rampy stable on this board. They might just made it to compatibility testing, before 3Dfx was sold.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 22.11.13 at 16:52:30

EMPEROR wrote on 22.11.13 at 16:33:18:

I think 3dfx used this MB to test the first Revision A0 (like the one I have ULTRA RARE).

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by EMPEROR on 22.11.13 at 16:59:35

Yes, I know, but did they confirm Rampy was stable on it?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 22.11.13 at 17:23:03

EMPEROR wrote on 22.11.13 at 16:59:35:

Yes, I know, but did they confirm Rampy was stable on it?

No 3dfx engineers said me it's stable on this MB.
But I think it's stable on this motherboard because the engineers used this one to test the rampage A0 and A1.
But in 2013 we need to test it.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 22.11.13 at 17:48:54

I have one question:

The SE440BX-2 is AGP 2x only (not 4x).

is it support the rampage agp4x anyway with good voltage 1,5V?

EDIT:

I read that on internet:

The Intel(R) SE440BX-2 Motherboard complies with AGP revision 1.0. It
supports only +3.3V AGP-66/133 devices, therefore, it will not support
newer AGP cards (revision 2.0, +1.5V)

Perhaps it's for that in the lab (picture) the engineers used an AGP riser card up to the agp slot of the MB? to change the voltage.

Is it true?

Here is the pdf:

http://downloadmirror.intel.com/17782/eng/motherboards_desktop_se440bx2.zip

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 22.11.13 at 19:32:52

Glad to see that the board is running and arrived safety.

Did you manage to run some games and benchmarks on it ?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 22.11.13 at 20:14:25

m14radu wrote on 22.11.13 at 19:32:52:

Glad to see that the board is running and arrived safety.

Did you manage to run some games and benchmarks on it ?

Not now I don't have enough time to install Win98se.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by goriath on 23.11.13 at 03:29:17

trevormacro wrote on 22.11.13 at 17:48:54:

It doesn't make sense, I think that riser is used to test AGP cycles. If they needed 1.5V for Rampage they would have used a 1.5V motherboard.

I could be wrong but, thinking at the period in which the Rampage would have come out it should follow 2.0 specifics of the AGP standard. This mean that Rampage should work as the Voodoo4, 1x/2x@3.3V or 1x/2x/4x@1.5V

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 23.11.13 at 12:21:48

Agree with goriath- the AGP riser you see in the pic is used for reading AGP register from Rampage.

Rampage can be used in 3.3V or 1.5VAGP motherboards.

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 23.11.13 at 12:27:20

So if i put the rampage on the intel MB with 3,3V, there is no risks to damage the card?

Thanks Goriath and Oschar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 23.11.13 at 12:33:34

trevormacro wrote on 23.11.13 at 12:27:20:

So if i put the rampage on the intel MB with 3,3V, there is no risks to damage the card?

Thanks Goriath and Oschar.

Nope, NOT DANGERUOUS. Not worry. It is safe motherboard for Rampage.

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 26.11.13 at 15:24:53

I have found another LCD: Fujitsu LCD 17 inch.

I need to test it with the rampage.

nb. It's not in 16/9 so perhaps better than my LG LCD.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 27.11.13 at 12:24:09

Installation of Win98se, Rampage drivers, DirectX 8.1, 3dmark2001se : (all without dongle :( )

16 colors:

before 16 bits:

In 16 bits, DxDiag:

In 32 bits, 3dmark2001se installation:

Oh yes the Rampage is running...

Running 3dmark01se:

Videos:

In desktop:

http://youtu.be/MHQvw85XpS8

http://youtu.be/MxvtPW_vDCU

Installing the Rampage drivers:

http://youtu.be/GvOmHlb7T6I

Installing DirectX 8.1:

http://youtu.be/Ptj2zMkdhJQ

Bonus:

http://img4.hostingpics.net/pics/198330bonus.jpg

Trevor.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 27.11.13 at 21:14:29

Excellent to check you handled to install correctly the drivers even without Dongle. Good work!!!

Very glad here to see the baby runing again.

Next step some games SShots.

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 28.11.13 at 09:56:00

Here are some news pictures and videos:

3dfx Tools installed:

3dmark2001se:

High resolution:

The Rampy:

Quake 1:

Videos:

Quake 1:

http://youtu.be/kJNSSLaR-3k

3dmark01se:

http://youtu.be/GkOIgHCQ1IQ

Trevor.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 29.11.13 at 11:51:22

Interesting picture (when the chip is not very good connected to the socket):

Half-Life:

The Rampy :) (The Magic Graphic Card)

Videos:

Some others games (tests) in 3dmark01se:

http://youtu.be/F3eJtih2AQs

Half-Life:

http://youtu.be/pNE0B-sgmjY (first)

http://youtu.be/90DnQ6he3AQ (second)

http://youtu.be/aUS1WUOjmEo (third)

You will see in the third video of Half-Life, when I want to change in OpenGL mode the game crash and return to the desktop.
Perhaps the directory search one .dll file? Any idea?

Trevor.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 29.11.13 at 12:32:51

Nice videos and pics!!! :)

Quote:

You will see in the third video of UT, when I want to change in OpenGL mode the game crash and return to the desktop.
Perhaps the directory search one .dll file? Any idea?

Not way trying to play UT with OpenGL. The only way would be across the Geri's SW. Very busy here and I dont find one moment for trying it.
In D3D UT locks up in few seconds.

Contact with Geri, he will explain you how do it.

ps. Anyway the most important now is find a solution for inverter the colors correctly or via HW with Dominik's dongle or via SW driver as EMPEROR recommended me. I think is the most important now even before to work on driver since why we want a driver stable if we can't enjoy of image quality.

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 29.11.13 at 12:33:38

trevormacro wrote on 29.11.13 at 11:51:22:

You will see in the third video of Half-Life, when I want to change in OpenGL mode the game crash and return to the desktop.
Perhaps the directory search one .dll file? Any idea?

Trevor.

Rename 3dfxogl.dll to opengl32.dll and put into the folder where the hl.exe is located.

You can also putting the renamed opengl32.dll into system32 folder.

Just one more test and your dongle is ready.

Edit: Inverting the colors via driver or software does not work, you will have the same strange "lines" in the image just with inverted color. HW dongle is the only way.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 29.11.13 at 12:37:47

One mistake: for the video it's not UT but Half-Life. It's fixed.

@Loeschzwerg:

Nice the dongle will be ready soon :) Thank you

I will try your tip.

@Osckhar:

I will ask Geri.

EDIT: what others games can i try? if someone can give me a list please ( i will download demos).

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 29.11.13 at 12:49:11

Unreal (not UT)
Quake 2
Forsaken
DeusEx
System Shock 2
Turok 2
Final Reality Benchmark
RealWorld OpenGVS Benchmark
Need for Speed Porsche
Sin

Try every game/benchmark from 1997 to 2001.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 29.11.13 at 12:53:48

Loeschzwerg wrote on 29.11.13 at 12:49:11:

Unreal (not UT)
Quake 2
Forsaken
DeusEx
System Shock 2
Turok 2
Final Reality Benchmark
RealWorld OpenGVS Benchmark

Try every game/benchmark from 1997 to 2001.

Ok. But I need to know how to run all benchmarks in all games :)

I have download this file (from Geri):

http://legend.uw.hu/TitaniumGL/TitaniumGL_for_windows98_2013_08.zip

Inside there is a file named opengl32.dll

I will try to put in every game to know if it work good in D3D or OpenGL mode. TitaniumGL is a good plugin to emulate the OpenGL/D3D mode. Follow me here ;)

What would be the best version of DirectX for the Rampage?

Can I download this OpenGVS version?

http://web.archive.org/web/20060312194149/http://www.quantum3d.com/support/opengvs/downloads/441/gvsp441.exe

Or this version is better:

OpenGVS Real World Benchmarks 2.4.2?

@Osckhar:

Did you run your game: THE LONGEST JOURNEY in D3D with Geri plugin or not?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 29.11.13 at 13:30:22

trevormacro wrote on 29.11.13 at 12:53:48:

Can I download this OpenGVS version?

http://web.archive.org/web/20060312194149/http://www.quantum3d.com/support/opengvs/downloads/441/gvsp441.exe

Or this version is better:

OpenGVS Real World Benchmarks 2.4.2?

Get the version from here:
http://www.xin.at/thrawn/files/opengvs-realworld/Q3D_RWB_242.zip

This is the version from GrandAdmiralThrawn's server, which we all used for benchmarking.

You will be most successful with Direct3d7 games and direct3d8 games that do have a fallback to direct3d7.

But it is possible that some early direct3d8 games will run. You can try Morrowind, UT2003, MaxPayne, Warcraft 3, NfS Hot Pursuit 2

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 29.11.13 at 13:34:21

Loeschzwerg wrote on 29.11.13 at 13:30:22:

trevormacro wrote on 29.11.13 at 12:53:48:

Get the version from here:
http://www.xin.at/thrawn/files/opengvs-realworld/Q3D_RWB_242.zip

This is the version from GrandAdmiralThrawn's server, which we all used for benchmarking.

Already downloading it: 164 Mo.
Where can I download the Unreal Demo?

EDIT:

I have downloaded these demos for now:

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 29.11.13 at 13:43:24

There is no official demo version of Unreal, there were only few demos that came with 3d hardware.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 29.11.13 at 13:48:03

Loeschzwerg wrote on 29.11.13 at 13:43:24:

There is no official demo version of Unreal, there were only few demos that came with 3d hardware.

Sorry but I don't have one :( if somebody have one and want to upload it, feel free :)

EDIT: a list of games from 1998 to 2001 :)

http://www.jeuxvideo.com/telecharger/jeux-et-demos/action-pc-date.htm

I download these games too:

- Expendable
- Tomb Raider 2, 3
- Carmageddon II : Carpocalypse

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 29.11.13 at 16:28:54

Loeschzwerg wrote on 29.11.13 at 12:33:38:

trevormacro wrote on 29.11.13 at 11:51:22:

You will see in the third video of Half-Life, when I want to change in OpenGL mode the game crash and return to the desktop.
Perhaps the directory search one .dll file? Any idea?

Trevor.

For Quake 3:
When I install the rampage drivers there is no file named 3dfxogl.dll in windows/system or system32 folder.
when I follow your 2 tips the errors is the same as gdonovan had in the past. opengl system not found. why? do you have a solution?
nb. i only tested with geri opengl32 dll file because i don't have 3dfxogl.dll in my windows directory.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 29.11.13 at 17:32:45

You do have the drivers from rashly, right? When you unzip them you should have three .exe files. Execute rampage-OpenGL.exe and it will unzip in a folder. Now you've got the 3dfxogl.dll.

Get this file and rename it to opengl32.dll and copy it in a game folder where the game-exe is located.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 29.11.13 at 18:58:10

@Loeschzwerg:

I will give a try.

Here are some pictures of games and Final Reality benchmark: (all in software mode for now)

Some games don't run.

Videos:

Final Reality Benchmark (1):

http://youtu.be/3uebNJsB4PQ

Final Reality Benchmark (2):

http://youtu.be/09DDaX_7iHE

Final Reality Benchmark (3):

http://youtu.be/vrmlwplEMMI

Bonus:

http://youtu.be/ljoEnQGZ78k

Try to find the game?

Trevor.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 29.11.13 at 19:09:13

Nice pictures Michael :)
Do you have a game that on the regular voodoo5 won't work to show us ?

How hot get the chip ?

As i recall corectly you have more than 2 rampage chips, right ? ....are they functional ?
What about the second rampage card ? :)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 29.11.13 at 19:23:23

m14radu wrote on 29.11.13 at 19:09:13:

No.
I need to measure with my IR Laser Gun.
Only one in full working condition. the decaped chip need to be repaired before.
The Rampy A0 + decaped chip are in transit to Oscar. I hope he can repair it.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 30.11.13 at 10:15:02

Some interesting informations:

The temperature on the gpu cooler reach 37/38 max.
The temperature on RAM go to 43/44 max.

I have taken some pictures with fraps (and they are better than in game for the colors). :)
Will upload this morning i think.

a bad new: quake 3 doesn't boot with your second tip :(
I have copy the dll file everywhere: in quake 3 exe folder, in system and system32 folder. but always the same error message. do you have an idea? perhaps it's my demo?

Trevor.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 30.11.13 at 10:15:30

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 30.11.13 at 12:25:14

As I promise:

Tomb Raider III crash even in software mode.

Worms 2 run very well!

Other picture.

I have removed the sheet of aluminium between chip and cooler and the card work nice now :)

Some FRAPS screenshots:

As you can see the colors are nice with fraps.

Half-Life:

Unreal Tournament:

Small resolution:

Better large resolution:

In game:

Videos:

GPU cooler temperature:

http://youtu.be/xP0izhq2Q7w

Tomb Raider III:

http://youtu.be/AKxIEaSZQHw

Worms 2:

http://youtu.be/wqL2bKZ0Cho

Carmageddon II:

http://youtu.be/J3z8iEghPTc

nb. When I run quake 1 and fraps before I can't take picture inside the game with hotkey :( if anybody have a solution for that?

Trevor.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 09.12.13 at 12:04:48

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 09.12.13 at 12:21:05

Seems this time Dominik found right Resistor value. Excellent work!!!

@Michel,

Take one picture using sidebar_trip pattern but not zoomed.
Like this one, with dongle and not dongle.

Added. Take another one when pc is booting up and you get info about info CPU, memory, bios, so on... Are letter shifteds?

Thanks,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 09.12.13 at 12:53:19

Yes, image is looking quite good.

Please also some pictures 800x600@60hz and 1024x768@60hz.

Try using 75hz.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 09.12.13 at 13:55:03

osckhar wrote on 09.12.13 at 12:21:05:

Already taken look my previous reply (sidebar_trip pattern).

640*480:

http://img4.hostingpics.net/pics/196536DSC07607.jpg

With the dongle in this resolution we have a very good display no?

ScanDisk:

BSOD:

Seems to be good no?

Without dongle (BIOS):

http://img4.hostingpics.net/pics/389589DSC07433.jpg

@Loeschzwerg:

I will try tomorrow.
but in 1280*1024 seem we have a blured screen why?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 09.12.13 at 14:09:55

Much more image quality improvements this time. Even in dos mode letters are not shifted. WoW!!!! :D :D :D

Congratulations Dominik!!!

Take one more inside BIOS with dongle.

Thanks,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 09.12.13 at 14:29:24

This really looks very good :) I couldn't replicate this improvement on my testsetup.

1280x1024... that's a resolution when shielding gets very important. Try using 75hz @ 640x480 and above. The image should start blurring here too.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by goriath on 09.12.13 at 15:47:10

Loeschzwerg wrote on 09.12.13 at 14:29:24:

Would be nice to know if the original dongle acts the same at that resolution.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 09.12.13 at 16:11:05

goriath wrote on 09.12.13 at 15:47:10:

Loeschzwerg wrote on 09.12.13 at 14:29:24:

Would be nice to know if the original dongle acts the same at that resolution.

It's better if Nightbird can send us all the values of her original rampage dongle.
If you want to contact her, you can wait until the next year to have an answer. This girl is a ghost.
Perhaps you can forget your idea.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by goriath on 09.12.13 at 23:17:23

She has been pretty friendly and helpful every time I needed help.

BTW, I was just rhetorical.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 09.12.13 at 23:52:59

goriath wrote on 09.12.13 at 23:17:23:

She has been pretty friendly and helpful every time I needed help.

BTW, I was just rhetorical.

Ok so ask her some helps by email or PM in VA board (creating a new topic there) about dongle like for exemple all the components values or take some pictures in differents resolutions and we will progress fastly with the rampage dongle.
is it ok?

Thanks for your help Goriath.
I will wait patiently about her answer sent to you.

nb. the creator of the dongle revA2 do a hard work and Nightbird look in all rampage dongle topics without help him. What is the problem with her? No share of informations to help the 3dfx community? if i can advise you, don't sell anything rare to her because for the futur you will have some problems to ask help.
Look the pictures Nightbird sent to gdonovan about her dongle, very bad picture!

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 10.12.13 at 09:25:59

come back to this topic,
I can't set others refreshs like 75 hz.
I didn't installed the lcd drivers. I just have plug and play monitor. perhaps it's the cause.
If i want to change the hertz how can i do?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 10.12.13 at 12:08:27

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 10.12.13 at 12:54:06

800x600 still is ok, 1024x768 it starts to blurr.

Looking at the drivers I can see that different hz (up to 160hz at 640x480) settings should be possible.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 10.12.13 at 12:54:34

I am really surprised about the new dongle update since the image quality is much better now. Tomorrow I will have here SMD resistor and I will do some test but simple view the improvement is amzing.

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 10.12.13 at 13:32:07

The rampage B001 is an amazing card.

When I have played Quake 1 (with dongle revA2), I had a good feeling... It's so excited to play game with a rampage card on an old PC.

Thanks Loeschzwerg the colors are nice in game (not blured).
Thanks Osckhar for the repair of my jewel. I hope you will repair my Rampy A0 too.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Geri on 11.12.13 at 00:16:40

please note that titaniumgl will only work when at least directx9 is installed. titaniumgl does not work with directx8 or 7.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 11.12.13 at 00:38:07

Michel,

Test TitaniumGL with Quake3. You will have a surprise. 8-)

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 11.12.13 at 09:29:02

osckhar wrote on 11.12.13 at 00:38:07:

Michel,

Test TitaniumGL with Quake3. You will have a surprise. 8-)

- Oscar.

I already tested it puting the opengl32.dll file in the same directory as quake 3 arena autoexec file without luck.
How can it work?
nb. I have directX 8 installed on win98se.
perhaps i need to install directX 9, a,b or c version?to run the game.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 11.12.13 at 09:36:03

Geri wrote on 11.12.13 at 00:16:40:

please note that titaniumgl will only work when at least directx9 is installed. titaniumgl does not work with directx8 or 7.

is it possible to make compatible titaniumGL with directX 8?rebuild it and post the link here to download it.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 11.12.13 at 09:36:16

Install last DX9.C version and put the OpenGL32.dll in the Quake directory exe.

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 11.12.13 at 09:43:10

osckhar wrote on 11.12.13 at 09:36:16:

Install last DX9.C version and put the OpenGL32.dll in the Quake directory exe.

- Oscar.

Is it better to use directX 8 or directX 9.0C with the rampage card?
For all the games.
I download directX 9.0c dated on 22 July 2004.
or it's better to take the version of 8 december 2006?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 11.12.13 at 09:51:35

For using the Geri's wrapper you need at least DX.9 or upper.

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 11.12.13 at 11:06:18

Quake 3 run but after there was a BSOD when i try to play like gdonovan.
nb. is it safe if remove scandisk after a windows bug? because i don't want to wait everytime lol.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 11.12.13 at 12:07:48

Yes, it is normal.

Now Geri needs to work on in his wrapper but possibly we will have a playable Quake3 game.

What I think is the 3dfxogl.dll comes with original Ramapge set driver is not to able to move any OpenGL demo. I need to test it much better. What I think is with this opengl32.dll library Rampage cant run Quake/2 or 3 if is not done via SW wrapper.

The library makes this reference. All the time doiing reference in debugg mode and there are a mix info, for Voodoo3, Rampage... so on.

SOFTWARE\3Dfx Interactive\Voodoo3\ICD
GL_CURRENT_INDEX GL_CURRENT_NORMAL GL_CURRENT_TEXTURE_COORDS GL_CURRENT_RASTER_COLOR GL_CURRENT_RASTER_INDEX GL_CURRENT_RASTER_TEXTURE_COORDS GL_CURRENT_RASTER_POSITION GL_CURRENT_RASTER_POSITION_VALID GL_CURRENT_RASTER_DISTANCE GL_POINT_SMOOTH GL_POINT_SIZE GL_POINT_SIZE_RANGE GL_POINT_SIZE_GRANULARITY GL_LINE_SMOOTH GL_LINE_WIDTH GL_LINE_WIDTH_RANGE GL_LINE_WIDTH_GRANULARITY GL_LINE_STIPPLE GL_LINE_STIPPLE_PATTERN GL_LINE_STIPPLE_REPEAT GL_LIST_MODE GL_MAX_LIST_NESTING GL_LIST_BASE GL_LIST_INDEX GL_POLYGON_MODE GL_POLYGON_SMOOTH GL_POLYGON_STIPPLE GL_EDGE_FLAG GL_CULL_FACE GL_CULL_FACE_MODE GL_FRONT_FACE GL_LIGHTING GL_LIGHT_MODEL_LOCAL_VIEWER GL_LIGHT_MODEL_TWO_SIDE GL_LIGHT_MODEL_AMBIENT GL_SHADE_MODEL GL_COLOR_MATERIAL_FACE GL_COLOR_MATERIAL_PARAMETER GL_COLOR_MATERIAL GL_FOG GL_FOG_INDEX GL_FOG_DENSITY GL_FOG_START GL_FOG_END GL_FOG_MODE GL_FOG_COLOR GL_DEPTH_RANGE GL_DEPTH_TEST GL_DEPTH_WRITEMASK GL_DEPTH_CLEAR_VALUE GL_DEPTH_FUNC GL_ACCUM_CLEAR_VALUE GL_STENCIL_TEST GL_STENCIL_CLEAR_VALUE GL_STENCIL_FUNC GL_STENCIL_VALUE_MASK GL_STENCIL_FAIL GL_STENCIL_PASS_DEPTH_FAIL GL_STENCIL_PASS_DEPTH_PASS GL_STENCIL_REF GL_STENCIL_WRITEMASK GL_MATRIX_MODE GL_NORMALIZE GL_VIEWPORT GL_MODELVIEW_STACK_DEPTH GL_PROJECTION_STACK_DEPTH GL_TEXTURE_STACK_DEPTH GL_MODELVIEW_MATRIX GL_PROJECTION_MATRIX GL_TEXTURE_MATRIX GL_ATTRIB_STACK_DEPTH GL_CLIENT_ATTRIB_STACK_DEPTH GL_ALPHA_TEST GL_ALPHA_TEST_FUNC G

3Dfx Interactive Inc. Rampage 1.1.0 0.0.1 GL_ARB_texture_cube_map GL_EXT_abgr GL_EXT_blend_func_separate GL_EXT_blend_minmax GL_EXT_blend_subtract GL_EXT_bgra GL_EXT_compiled_vertex_array GL_EXT_cull_vertex GL_EXT_packed_pixels GL_EXT_paletted_texture GL_EXT_point_parameters GL_EXT_stencil_wrap GL_EXT_texture3D GL_EXT_texture_cube_map GL_EXT_texture_env_add GL_EXT_texture_env_combine GL_EXT_texture_filter_anisotropic GL_EXT_vertex_array GL_SGIS_texture_edge_clamp GL_WIN_swap_hint WGL_ARB_extensions_string WGL_EXT_extensions_string GL_3DFX_multisample GL_3DFX_tbuffer __glSSTCreateContext

Seems SW engineers already were working on Sage too:

/WIN9X/debug/sst2rfvertex.obj
bjs/WIN9X/debug/sst2texture.obj
D:\3DFX\DEVEL\opengl\lib\GL\drivers\sst2\objs\WIN9X\debug\gevlib.obj
D:\3DFX\DEVEL\opengl\lib\GL\drivers\sst2\objs\WIN9X\debug\sageheap.obj
D:\3DFX\DEVEL\opengl\lib\GL\drivers\sst2\objs\WIN9X\debug\sageucode.objJ
D:\3DFX\DEVEL\opengl\lib\GL\drivers\sst2\objs\WIN9X\debug\sst2rfvertex.obj
ID:\3DFX\DEVEL\opengl\lib\GL\drivers\sst2\objs\WIN9X\debug\sst2texture.obj
ED:\3DFX\DEVEL\opengl\lib\GL\drivers\sst2\objs\WIN9X\debug\g_point.obj
HD:\3DFX\DEVEL\opengl\lib\GL\drivers\sst2\objs\WIN9X\debug\sagevertex.obj
ID:\3DFX\DEVEL\opengl\lib\GL\drivers\sst2\objs\WIN9X\debug\sstrfvertex.obj
FD:\3DFX\DEVEL\opengl\lib\GL\drivers\sst2\objs\WIN9X\debug\sstpixel.obj
GD:\3DFX\DEVEL\opengl\lib\GL\drivers\sst2\objs\WIN9X\debug\sstbuffer.objJ
KD:\3DFX\DEVEL\opengl\lib\GL\drivers\sst2\objs\WIN9X\debug\sst2x86vertex.obj
ND:\3DFX\DEVEL\opengl\lib\GL\drivers\sst2\objs\WIN9X\debug\sagedrawelements.obj
GD:\3DFX\DEVEL\opengl\lib\GL\drivers\sst2\objs\WIN9X\debug\sagestate.obj
ED:\3DFX\DEVEL\opengl\lib\GL\drivers\sst2\objs\WIN9X\debug\sstfifo.obj
ED:\3DFX\DEVEL\opengl\lib\GL\drivers\sst2\objs\WIN9X\debug\sstinit.obj
FD:\3DFX\DEVEL\opengl\lib\GL\drivers\sst2\objs\WIN9X\debug\ssttexdl.obj

By the way, I was playing witn info driver and I added some features on 3Dfx Tools. Sure Ramapge cant any of such features but even so I think is a nice appealing to the eye.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 11.12.13 at 13:03:46

I've installed my fujitsu lcd drivers and now I can choose frequency.

I have added a new test patern.

Here are some pictures:

up to 800*600, 75 Hz applied (very blured screen):

640*480 (75 Hz, much better):

Installing DirectX 9.0c:

UT 800*600 (D3D mode):

Here is a freeze of the game in D3D mode:

Videos:

Unreal Tournament (software mode with dongle):

http://youtu.be/3YWiBk25I0c (part 1)

http://youtu.be/KLPJm3nXZeo (part 2)

Quake 3 (TitaniumGL):

http://youtu.be/dtWX5Ztz1Q8

Half-Life (Direct3D mode: 800*600):

http://youtu.be/gDOTbp4bIdo

Trevor.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 11.12.13 at 17:13:57

@Geri and all others users:

I have used TitaniumGL (opengl32.dll) with Quake 3 + dongle + rampage + DirectX9.0c

Win98se Fatal Error : 0D 0147:035C0ABA

What is the problem?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Geri on 11.12.13 at 17:32:36

Win98se Fatal Error : 0D 0147:035C0ABA

What is the problem?

maybe solar erupts
*cough*
:D

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Geri on 11.12.13 at 19:16:34

Voodoo Rampage support has been added to TitaniumGL

http://www.falconfly.de/cgi-bin/yabb2/YaBB.pl?num=1268949568/90#93

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 11.12.13 at 22:56:25

osckhar wrote on 11.12.13 at 12:07:48:

Now Geri needs to work on in his wrapper but possibly we will have a playable Quake3 game.

The direct3d driver is unstable- Till it is fixed the wrapper will be just as unstable.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 12.12.13 at 09:43:20

Hi Gary,

Agree, but what if Geri can do is check a lot of Demos with different OpenGL features checking stability of Rampage. Once done the codify the new code using or deleting features according stability showed in different tests making the card the most stable possible.

Less is nothing!

Me needed find someone interested to help with D3D core.

Regards,
Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 12.12.13 at 13:12:58

I have make some tests about temperatures of the cooler.

- The temperature of the socket (front of the card) without turn ON pc is 30° in my room because I have this temp. here.

- When I am in windows 98 desktop (after the boot) the socket is at 32°.

- When I run a game in software mode during 5 or 10 min the temperature go from 32° to 38° max (without go up than 38°).

- The sdrams can reach 39°-40° but it's the max.

nb. when I have a freeze I think it's because one pin or more inside the socket don't touch the chip.

Regards,
Trevor.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 12.12.13 at 14:07:54

If you want to make an accurate test temp, you need to buy TC (Thermo Couple). Put it between chip and test socket heatsink.

This is the best manner to monitor real temperature.

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Geri on 12.12.13 at 17:12:16

trevormacro wrote on 12.12.13 at 13:12:58:

what software do you used to stresstest, and how many hours do you had it run?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 12.12.13 at 17:40:11

Geri wrote on 12.12.13 at 17:12:16:

trevormacro wrote on 12.12.13 at 13:12:58:

what software do you used to stresstest, and how many hours do you had it run?

I just put my Laser IR gun pointed on the socket cooler in front of the card during a short time (15 min max) running some games and escape.
I have noticed the temperature goes down when I reboot the pc. The temp. doesn't go up than 38°/39° for the aluminium cooler.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Geri on 12.12.13 at 17:47:49

so you runned various games, and exitted, and the gpu reached 39c this way. in this case with continous work we simply can predict at least 80°c working temperature even at winter.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by m14radu on 12.12.13 at 18:35:30

Nice work Geri !

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Geri on 12.12.13 at 19:43:36

m14radu wrote on 12.12.13 at 18:35:30:

Nice work Geri !

thx ^^

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Tim on 13.12.13 at 16:18:24

Agreed, great work.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 16.12.13 at 10:28:58

Finally I can run Quake 3 with my rampage now (in game or demo) incl. TitaniumGL.

I have taken some fatals errors pictures but where can i see a log file of these errors in windows 98se?

After two hours of rampage running I have some stranges things on the screen and perhaps it's artifact. anyone need to confirm me this.

I will upload here some videos and screenshots for quake 3.
Some fatal errors pictures too.

Trevor.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 16.12.13 at 10:59:47

If you are running Quake3 with TitaniumGL wrapper you will get bad texture color. If you are talking about it, yes, it is normal since TitaniumGL has problems handling textures.

- Oscar.

ps. I only hope you are not talking about your card has problems. :'(

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 16.12.13 at 10:59:58

Excellent- I have found the opengl files I was using, I will forward them to Oscar and trevormacro after work once I verifiy some information this afternoon.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 16.12.13 at 11:11:03

I am using GLDirect, not sure if you are talking about it. For now it is the wrapper with more stability even I have not BSOD and I can run all the timedemos.

Later I will upload pics, but I got 800x600 all full demo002, AVg FPS 12 and 640x480 full Avg FSP 21. Not sure where place such screenshots since there are a lot of open topics with Ramapge Shots. LoL

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 16.12.13 at 11:30:06

osckhar wrote on 16.12.13 at 11:11:03:

I am using GLDirect, not sure if you are talking about it. For now it is the wrapper with more stability even I have not BSOD and I can run all the timedemos.

That is what I was running and provided on the Rampage CD, I finally found a serial (provided by the software company as it is no longer supported)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 16.12.13 at 11:30:22

osckhar wrote on 16.12.13 at 10:59:47:

Hi,

Fatals Errors pictures:

Some others pictures:

(After 1 hour in action...)

(After 2 hours of UT, Half-Life and Quake 3 in final)

It's a picture of artifact I have is it normal?

37,5° inside the socket on a gold pin ( chip removed ).

Turn off the Rampy for 30 minutes and Turn ON artifact disappear.

nb. I don't know if it's artifact or just bad pin contact btw socket and chip.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 16.12.13 at 11:38:02

trevormacro wrote on 16.12.13 at 11:30:22:

(After 2 hours of UT, Half-Life and Quake 3 in final)

It's a picture of artifact I have is it normal?

1) That does not look good.

2) Why run the card that long? Unless you are active in working on a driver or stability issue then you are risking the board for no good reason.

In my opinion.

Untill we come up with a better D3D driver I don't see things progressing much further, it is one reason I stopped testing since it was clear to me the situation was not going to improve. I had tried three different OS supported by the driver, 4 or 5 different versions of direct x and several different motherboards in a search for better function.

That is why I posted all the results of my testing so other Rampage owners would not beat down the same path reinventing the wheel and putting their cards at risk.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 16.12.13 at 11:41:18

BSOD are normal since D3D driver core is unstable and it needs to be fixed.

About the problem in Quake3, I never got such error in the menu. It does not look as a problem for bad contact between chip and socket if you are using an OpenGL warapper there can be the problem.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 16.12.13 at 11:44:04

osckhar wrote on 16.12.13 at 11:41:18:

I can't say I have ever run across that in my testing, does look like something overheating.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 16.12.13 at 11:51:27

@gdonovan:

I just run some games and benchmarks to take some videos and upload them here today. It's because I put my rampy for 2 hours without risks because the temperature doesn't go UP than 38° on the socket. when I get the artifacts i measure the temp.
Thanks for your tips.

@osckhar:

I have used TitaniumGL drivers only with quake 3.

I will upload the video. You will see here:

Quake 3 after 2 hours:

http://youtu.be/YQLnPVwLI5A

My card work good not worry about that.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 16.12.13 at 11:55:32

Nice to hear it.

I was checking BSOD and any of them makes reference to VXD driver. Seems the BSODs you are getting makes reference to NOT stable set pc.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 16.12.13 at 12:01:24

osckhar wrote on 16.12.13 at 11:55:32:

Nice to hear it.

I was checking BSOD and any of them makes reference to VXD driver. Seems the BSODs you are getting makes reference to NOT stable set pc.

Did you check my quake 3 video artifacts what is that?
Ok and how can I set a stable pc?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 16.12.13 at 12:07:31

I know such problems from overclocking including a high temperature. The problem is clearly GPU realted and not memory.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 16.12.13 at 12:09:42

Loeschzwerg wrote on 16.12.13 at 12:07:31:

I know such problems from overclocking including a high temperature. The problem is clearly GPU realted and not memory.

Yes I had the same problem with my Voodoo5 5500 when i was playing a long time.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 16.12.13 at 12:18:57

Try using some thermal grease between GPU and socket clamp. Don't use to much of it, the thermal grease must not flow under the GPU!

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 16.12.13 at 12:25:03

Loeschzwerg wrote on 16.12.13 at 12:18:57:

Try using some thermal grease between GPU and socket clamp. Don't use to much of it, the thermal grease must not flow under the GPU!

I have never put thermal grease btw chip and top of the cooler. So the chip is never cooled by the aluminium heatsink. One day Oscar said me it's not necessary to put thermal grease.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 16.12.13 at 12:30:12

Well it shouldn't be necessary since the GPU is clocked very very low. But it could help to get a few °C so ist works more stable, at least under that high ambient temperature you have at home.

Just an idea ;)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 16.12.13 at 12:34:15

Loeschzwerg wrote on 16.12.13 at 12:30:12:

Ok thanks for these nice informations. and when i have some problem like that:

http://img4.hostingpics.net/pics/845625DSC07643.jpg

It means what?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 16.12.13 at 12:42:55

It means not good contact between rampage chip and test socket.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 16.12.13 at 14:34:33

Some videos:

3dfx Rampage #B001 (Quake 3 TitaniumGL):

http://youtu.be/HuGiGHx2TS0

http://youtu.be/FdS_Qc8R6Fw

3dmark01se:

http://youtu.be/bfSZMJzyNpA

Half-Life:

http://youtu.be/qnh16Ik2CAw

Unreal Tournament:

http://youtu.be/yl4OlvIjQsY

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 16.12.13 at 16:05:10

I have installed GLDirect and it's much better than TitaniumGL.
Quake 3 run very well.

I have a new conclusion:

when you play the games in software mode the rampage never crash.
when you play the games in OpenGL mode the rampage often crash because it use more the chip so more heat than the software mode.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 16.12.13 at 17:11:50

trevormacro wrote on 16.12.13 at 16:05:10:

when you play the games in software mode the rampage never crash.

No surprise there, it has Banshee/Voodoo 3 2D core.

Quote:

when you play the games in OpenGL mode the rampage often crash because it use more the chip so more heat than the software mode.

Yes and no.

Rampage crashes for a number of reasons- The D3D driver for one and the nature of the chip for another. It crashes just fine in D3D, these same crashes are happening with the wrapper since the wrapper is using the D3D driver.

As I pointed out before, the chip had several "hacks" done to it in the attempt to get it functioning. It uses no AGP functions (they were disabled) and I was told by Hank it does not access the ram on the board proper so they could focus on different functions of the core while debugging. This MAY be why the ram is not reported correctly by several programs.

The best we can hope for is a revised D3D driver that is more stable. A new BIOS is possible but I'll let someone else flash their card first!

Hank indicated some hacks were chip level, these can never be undone.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by EMPEROR on 16.12.13 at 17:32:47

I am curious if someone can try to run GPU-Z, just to see if there will be any info shown.
It's a free little tool, that can be downloaded from here:
http://www.techpowerup.com/gpuz/
I am not sure if it will work on 98, but it's worth a try...

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by paulpsomiadis on 16.12.13 at 18:59:31

GPU-Z will probably have a brain-fart as it's a GPU that was never officially released... ::)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 16.12.13 at 19:18:26

paulpsomiadis wrote on 16.12.13 at 18:59:31:

GPU-Z will probably have a brain-fart as it's a GPU that was never officially released... ::)

I think I have tried, crashes. I'll test again.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 16.12.13 at 19:43:47

paulpsomiadis wrote on 16.12.13 at 18:59:31:

GPU-Z will probably have a brain-fart as it's a GPU that was never officially released... ::)

Went all the way back to version 0.0.7, no WIN98 support

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 16.12.13 at 19:53:58

Sorry, best you are going to get. Ran several versions of CPU-Z and this was the most info available about AGP device.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 17.12.13 at 10:03:10

Quote:

The best we can hope for is a revised D3D driver that is more stable. A new BIOS is possible but I'll let someone else flash their card first!

Not possible to flash the card directly using a 3Dfx flash tool. Card has a bug what not allowed flash the bios and stop the process. For such reason in A1 revision board engineers added a bios socket for update the bios across an external flash program utitlity based PLCC format.

I have one here and it works.

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 17.12.13 at 12:26:38

Small sidenote regarding Atmel AT49BV512: Get an Intel Pro/100+ network card with a bios socket. Put in the PLCC you want to write and flash the rom using fboot.exe from intel proutil.exe package.

The cheapest flash writer I know :)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 17.12.13 at 12:31:53

About the dongle Rev A2:

A little blured screen in low resolution:

Take a look:

http://img4.hostingpics.net/pics/185112DSC07738.jpg

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 17.12.13 at 12:34:47

To have a good set up motherboard for the rampy in the bios what can i change:

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 17.12.13 at 14:35:51

Since Rampage cant handle memory texture. Set it as AGP1x

- Aperture 4/ or 8/ or 16 or 32Mb. The min. as possible.
- Memory Cache disabled.
- AGP4x 1x Mode.

Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 17.12.13 at 15:13:27

osckhar wrote on 17.12.13 at 14:35:51:

Since Rampage cant handle memory texture. Set it as AGP1x

- Aperture 4/ or 8/ or 16 or 32Mb. The min. as possible.
- Memory Cache disabled.
- AGP4x 1x Mode.

Oscar.

thanks for these tips.
what about AGP FastWrite override and Enable memory gap?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 17.12.13 at 15:16:08

All disabled.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 17.12.13 at 15:29:37

osckhar wrote on 17.12.13 at 14:35:51:

Since Rampage cant handle memory texture. Set it as AGP1x

- Aperture 4/ or 8/ or 16 or 32Mb. The min. as possible.
- Memory Cache disabled.
- AGP4x 1x Mode.

Oscar.

I have found the card most stable with the smallest aperture possible, my VA6 goes down to 4mb. It makes a huge difference in crashes, the smaller the better.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 18.12.13 at 19:05:37

here is the ddr i use for my tyan k7x pro is it good or not? perhaps it's my ddr not compatible?

DDR 400 MHz PC3200 184 pins:

http://img11.hostingpics.net/pics/24617820131218007.jpg

nb. perhaps it can crash my system like blue screen fatal error?

look the tyan website: http://www.tyan.com/archive/support/html/memory_s2469.html

nb2. or perhaps i need to put only one cpu amd 2800+? it can cause a fatal error or not?

nb3. about all the bios: http://www.tyan.com/archive/support/html/b_s2469.html

I have found this website for fatal error:

http://www.computerhope.com/issues/ch000223.htm

0D: General Protection Fault

anyone can help me please?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 18.12.13 at 23:03:35

Dont be nervous there.

First what you must know if BSODs refers VXD.dll library or NOT.

If BSODs makes reference all the time to VXD library, it is normal, driver is still in beta phase. If BSODs makes reference NOT VXD library, then it means you have some problem in your pc, incompatibility HW, memory, general drivers if you installed something, so on...

If your problem refers second one, then I recommend you to use another system or disabled the second CPU AMD.

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 18.12.13 at 23:06:56

osckhar wrote on 18.12.13 at 23:03:35:

It's not a VXD fatal error.
Just 0D fatal error.
I have my MSI 848P Neo-V perhaps i need to test with this mb and it will be better.
Did you use some thermal grease on the chip or never?
If my DDR has more frequency than accepted by motherboard, do you think it's bad? (2,6V 400 MHz Non ECC Kingston I have compared to 2,5V max 266 MHz ECC of the MB ref.)

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 18.12.13 at 23:26:19

Quote:

It's not a VXD fatal error.

Good new. It means you have a problem there with your PC.

Quote:

I have my MSI 848P Neo-V perhaps i need to test with this mb and it will be better.

It is time you test it if your motherboard is compatible with 1.5/3.3V_AGP.

Quote:

Did you use some thermal grease on the chip or never?

Nope, I dont like the idea and I will never do it. Rampage chip runs very cold here.

Quote:

If my DDR has more frequency than accepted by motherboard, do you think it's bad? (2,6V 400 MHz Non ECC Kingston I have compared to 2,5V max 266 MHz ECC of the MB ref.)

Inverse would be a problem but if your memory can run high frequency. Memory is ready to work to low frequency.

My recommendation is you test another platform.

-Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 18.12.13 at 23:34:33

osckhar wrote on 18.12.13 at 23:26:19:

Quote:

It's not a VXD fatal error.

Good new. It means you have a problem there with your PC.

[quote]I have my MSI 848P Neo-V perhaps i need to test with this mb and it will be better.

It is time you test it if your motherboard is compatible with 1.5/3.3V_AGP.

Quote:

Did you use some thermal grease on the chip or never?

Nope, I dont like the idea and I will never do it. Rampage chip runs very cold here.

Quote:

If my DDR has more frequency than accepted by motherboard, do you think it's bad? (2,6V 400 MHz Non ECC Kingston I have compared to 2,5V max 266 MHz ECC of the MB ref.)

Inverse would be a problem but if your memory can run high frequency. Memory is ready to work to low frequency.

My recommendation is you test another platform.

-Oscar.[/quote]

My MSI MB official spec (taken here : http://www.msi.com/product/mb/848P-Neo-V.html#/?div=Detail)

One AGP slot supports 8x/4x at 0.8V (AGP 3.0) or 4x at 1.5V (3.3V not supported).

Is it good?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 18.12.13 at 23:39:02

Yupe, it is compatible.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 18.12.13 at 23:41:40

osckhar wrote on 18.12.13 at 23:39:02:

Yupe, it is compatible.

Well compatible and stable are two different things.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 18.12.13 at 23:43:48

gdonovan wrote on 18.12.13 at 23:41:40:

osckhar wrote on 18.12.13 at 23:39:02:

Yupe, it is compatible.

Well compatible and stable are two different things.

I will try and let you know if it's stable.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 18.12.13 at 23:48:28

gdonovan wrote on 18.12.13 at 23:41:40:

osckhar wrote on 18.12.13 at 23:39:02:

Yupe, it is compatible.

Well compatible and stable are two different things.

Totally agree.

Just test it and you will have other reference for comparing with your TYAN motherboard.

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 20.12.13 at 09:35:53

Installing Win98SE on my MSI mb:

http://img15.hostingpics.net/pics/81526120131219003.jpg
http://img15.hostingpics.net/pics/78437520131219004.jpg

DirectX 9:

http://img15.hostingpics.net/pics/55261420131219007.jpg

I set hardware acceleration to off in safe mode but when i am in normal windows mode and i put full hardware acc. and i reboot i get this freeze screen and i can't go to windows desktop why?

http://img4.hostingpics.net/pics/25795320131219008.jpg

oscar do you have a solution for me?

when i have installed the drivers i have chosen agp 2000 is it good? the first one. others choices are 3000, 3500 etc.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 20.12.13 at 10:49:53

If you followed the driver set nstall then all is correct. Possibly your motherboard is NOT compatible. I found using different motherboards once installed the driver pc lock up in windows 98 symbol before to load the desktop.

Questions:

Do you use more than one memory stick?

Did you disabled all the option reference to AGP in bios?

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 20.12.13 at 11:06:53

osckhar wrote on 20.12.13 at 10:49:53:

See my thread on motherboard testing, a good deal of motherboards do NOT work well with Rampage. If you need one that does, I have a spare MSI unit that works well

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 20.12.13 at 11:52:32

I have installed the drivers set easily in normal windows mode.
I have installed DirectX 9.0C easily.
If I put OFF hardware acceleration like this picture (http://www.deskshare.com/Resources/articles/images/window_troubleshoot.gif) my Windows 98 can boot very fastly without any freezes even if i choose more resolution or more bits.

BUT if I put at the max hardware acceleration and I reboot I have a freeze.

nb. In my Bios setting I only have Aperture 4/ or 8/ or 16, 32 or 64Mb and I can choose what I want but now it's on 64Mb.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 20.12.13 at 11:53:37

gdonovan wrote on 20.12.13 at 11:06:53:

osckhar wrote on 20.12.13 at 10:49:53:

See my thread on motherboard testing, a good deal of motherboards do NOT work well with Rampage. If you need one that does, I have a spare MSI unit that works well

Spare MSI unit what model? including processor and RAM?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 20.12.13 at 11:54:30

Use 4Mb for aperture.

And uninstall the Nvdia drivers and NOT install any motherboard driver.

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 20.12.13 at 11:59:13

osckhar wrote on 20.12.13 at 11:54:30:

Use 4Mb for aperture.

And uninstall the Nvdia drivers and NOT install any motherboard driver.

- Oscar.

I will try for aperture.
But it's not my desktop on the picture. just an exemple.

Motherboard Drivers:

I have installed before the rampy drivers this program:
infinst_enu.exe (Intel Chipset software Utility 6.2.1.1001 for my MSI mb). perhaps it can cause a conflict?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 20.12.13 at 12:01:20

trevormacro wrote on 20.12.13 at 11:59:13:

osckhar wrote on 20.12.13 at 11:54:30:

Use 4Mb for aperture.

And uninstall the Nvdia drivers and NOT install any motherboard driver.

- Oscar.

NOT INSTALL CHIPSET driver. It causes several crashes when Rampage is installed.

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 20.12.13 at 12:03:11

osckhar wrote on 20.12.13 at 12:01:20:

trevormacro wrote on 20.12.13 at 11:59:13:

osckhar wrote on 20.12.13 at 11:54:30:

Use 4Mb for aperture.

And uninstall the Nvdia drivers and NOT install any motherboard driver.

- Oscar.

NOT INSTALL CHIPSET driver. It causes several crashes when Rampage is installed.

- Oscar.

I don't know so it's perhaps the problem. I see now.
This chipset driver after the reboot detect a lot of hardware on the mb like PCI or USB etc...
I can uninstall it from control panel? possible?
Without chipset drivers it will be possible to run any games/benchmarks?
Thanks.
I will let you know.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by osckhar on 20.12.13 at 12:08:38

Yupe, possibly.

I recommend you- make a fresh Win98Se install and pleas, NOT INSTALL CHIPSET driver. ;)

- Oscar.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 20.12.13 at 12:13:02

osckhar wrote on 20.12.13 at 12:08:38:

Yupe, possibly.

I recommend you- make a fresh Win98Se install and pleas, NOT INSTALL CHIPSET driver. ;)

- Oscar.

Promise my friend. Sorry.
And after the rampy drivers installed I can run all games/benchmarks without chipset drivers installed sure?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 20.12.13 at 12:14:04

trevormacro wrote on 20.12.13 at 11:53:37:

Spare MSI unit what model? including processor and RAM?

Yes, had a Slot-A 500 or 550 if remember with ram. MSI 6191

http://www.falconfly.de/cgi-bin/yabb2/YaBB.pl?num=1373721600/0#0

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 20.12.13 at 13:36:37

gdonovan wrote on 20.12.13 at 12:14:04:

trevormacro wrote on 20.12.13 at 11:53:37:

Spare MSI unit what model? including processor and RAM?

Yes, had a Slot-A 500 or 550 if remember with ram. MSI 6191

http://www.falconfly.de/cgi-bin/yabb2/YaBB.pl?num=1373721600/0#0

Let me test my MSI board before and i will send you an email if it doesn't work with my rampy.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 21.12.13 at 13:34:34

Good New:

The Rampy 001 is compatible with the MSI MB. yes :)

I have tested it with Quake 3.

So good!

EDIT:

insalling quake 3 demo in full harrdware acceleration on my MSI mb.

http://img15.hostingpics.net/pics/27997020131221004.jpg

in quake 3 game i don't have any fatal error related to the mb. just some freeze or vxd blue screen :) seems it s correct.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 21.12.13 at 22:59:52

trevormacro wrote on 21.12.13 at 13:34:34:

Good New:

The Rampy 001 is compatible with the MSI MB. yes :)

in quake 3 game i don't have any fatal error related to the mb. just some freeze or vxd blue screen :) seems it s correct.

Try some games which are known to work well to see how long it will run.

System Shock II, Thief the Dark Project and Tomb Raider II Gold

Did you download the DX test pack?

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by trevormacro on 22.12.13 at 06:26:48

gdonovan wrote on 21.12.13 at 22:59:52:

trevormacro wrote on 21.12.13 at 13:34:34:

Good New:

The Rampy 001 is compatible with the MSI MB. yes :)

in quake 3 game i don't have any fatal error related to the mb. just some freeze or vxd blue screen :) seems it s correct.

Try some games which are known to work well to see how long it will run.

System Shock II, Thief the Dark Project and Tomb Raider II Gold

Did you download the DX test pack?

Remember in my country the temp room is at 30. much more heat than in USA or Spain in winter.
I will try and let you know.
I don't download yet the DX test pack. and you did you tested it?
I need to have the same settings as you to compare how long the games can run without freeze?

nb. when i had installed the rampy drivers i take the voodoo X 2000 the first one for win98 is it good? cause there are others versions like 3000 or 3500 etc...

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 22.12.13 at 11:00:29

trevormacro wrote on 22.12.13 at 06:26:48:

I don't download yet the DX test pack. and you did you tested it?
I need to have the same settings as you to compare how long the games can run without freeze?

nb. when i had installed the rampy drivers i take the voodoo X 2000 the first one for win98 is it good? cause there are others versions like 3000 or 3500 etc...

The DX test pack has many small applications for testing different aspects of Direct X

I will send you an inf file to use, I have modified it so there is only one choice and other smaller tweaks too.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by gdonovan on 20.01.14 at 11:56:51

Loeschzwerg wrote on 17.12.13 at 12:26:38:

Could you be so kind to e-mail me the fboot.exe? I have a computer with a corrupted BIOS chip I'm trying to repair using this method.

Title: Re: 3dfx Specter 1000 AGP TEST SOCKET Rev.A1 4700
Post by Loeschzwerg on 20.01.14 at 16:23:37

You've got mail ;)