Can this list be at any help?

http://www.voodooalert.de/board/index.php?pag … 9046&highlight=

Can anyone confirm that the list is the real deal?

EDIT. Found some more information...

http://extreme.pcgameshardware.de/tagebuecher … 5-5500-agp.html

It is the picture with an sheet of paper, it contains a parts list.

Thanks dude, it might. According those pages the missing SMD caps are 100nF/0,1µF 16V 0603 X7R 5 or 10% - now I know what I need to order. I should also replace the 100uf electrolytic SMD cap on the front with one with same values but correct size (the one I put on is physically smaller then the original)

UPDATE:

I've replace all the SMD caps and resistors with one of appropriate values, but I'm still getting the same issue. The card works fine if I select "Anti-Aliasing - Single Chip" but if I select "Performance Mode" the machine displays some horizontal lines and hangs.

My current test setup is a Gigabyte KT333 board + Athlon XP 2400+ -> is it possible the card is not liking the motherboard?

Both chips are detected fine, total memory is reported to be 64MB and so on...

Good job.... Awesomme.
Those V5 cards are a piece of history. I must admit that I like hearing that. Every V5 revived are good news.
Though I personally don't want one. It is awesomme that a piece of history is working yet again.

About these problems you encounter. Have you tested it in a different motherboard?
Perhaps you need to look into other driver combinations?
If not, then at least the card can work on some level. 😀

Tried another board - an apollo pro 133 + p3 - same issue... I did notice it runs fine on default openGL in quake 2 - it gets a silly 157 fps @ 1024x768 with raziel's driver set - but with the latest reference drivers I get screen corruption + hang in default OGL too. I'm starting to think it may be a driver issue of sorts. Either that, or one of the larger aluminum caps is not doing it's job and the board is not getting enough power.

I'm looking at the 100uf 16v cap - I only had the small one pictured above - although it has the same rating as the original sanyo cap, there might be other differences...

kanecvr wrote:

I'm looking at the 100uf 16v cap - I only had the small one pictured above - although it has the same rating as the original sanyo cap, there might be other differences...

Are you sure that (and the original) are 100uF? Sometimes those surface mount caps encode the capacitance differently. 100 can sometimes mean 10 * 10^0 = 10uF.

shamino wrote:

kanecvr wrote:

I'm looking at the 100uf 16v cap - I only had the small one pictured above - although it has the same rating as the original sanyo cap, there might be other differences...

Are you sure that (and the original) are 100uF? Sometimes those surface mount caps encode the capacitance differently. 100 can sometimes mean 10 * 10^0 = 10uF.

Tantalums might say 226 for 22uF and 107 for 100uF, but 10 means 10 uF and 100 means 100 uF in case of electrolytics.

Other V5 5500 pictures on the net appear to have OS-CON type polymer caps, so I suggest replacing with similar type instead of normal electrolytics. Could be the reason why it does not work under high load.

shamino wrote:

kanecvr wrote:

I'm looking at the 100uf 16v cap - I only had the small one pictured above - although it has the same rating as the original sanyo cap, there might be other differences...

Are you sure that (and the original) are 100uF? Sometimes those surface mount caps encode the capacitance differently. 100 can sometimes mean 10 * 10^0 = 10uF.

Yes I'm sure.... I looked it up and the card originally uses Sanyo 100uf 16v and Sanyo 470uf 16v... oh and they're electrolytic caps, not tantalum...

In any case I found out what's wrong with the thing - it's not the VSA100 chips, nor the caps - its actually the SDRAM chips connected to the second VSA100.... I found the edge-most pins on two of the rear ram modules connected to the "faulty" vsa100 were bent and shorting 🙁 I streightned them out with a razor and made sure they're connected to their respective pads but before I could run 3dmark the motherboard on my test PC failed (the thing hangs regardless of what video card or memory I put in it).

I'll retest the v5 tomorrow but I doubt it will work - shorts on sdram module pins usually kill the memory chip 🙁

The DRAM chips could conceivably be replaced if you can find replacements, so all is not lost. Hopefully you get that mobo repaired without too much difficulty.

gdjacobs wrote:

The DRAM chips could conceivably be replaced if you can find replacements, so all is not lost. Hopefully you get that mobo repaired without too much difficulty.

I found the exact chips on a Geforce 2 MX - I'm going to try to transplant them to the V5.

But check if the current chip still works before.

h-a-l-9000 wrote:

But check if the current chip still works before.

It doesn't. I get red linear artifacts on half the screen. 3dmark99 runs for a few seconds (with some artifacting) and then the computer hangs.

The missing smds are not the problem with this card. From the information you provided the VSA100 master chip is working fine reason why the card boots-up and works in single chip mode without issues. The problem could be that due to overheating the slave VSA100 chip is damaged and because it's only active in 3D in SLI/performance mode it doesn't show problems in 2D mode. The fix is to replace the slave VSA100 chip with a new working one but BGA rework isn't easy or cheap. The card usually works with bad memory although it shows artifacts it won't crash but a bad VSA100 chip would do that. Of course you can replace the memory modules for the slave VSA100 to check for that possibility also.
I had as similar issue with a V5 AGP and I ended up replacing both VSA100 chips to make the card functional again.

Good luck,
G1nX

I had a partially broken Geforce 4 Ti 4200 AGP4x and I fixed it by "baking the GPU".
The VSA100 are BGA chips as well. I think there is just a contact problem. Good luck!

thanks. I know the ram is bad because I found shorted pins. Physically shorted. After clearing the shorts, the card works in performance mode as long as I limit texture memory to 2MB / chip. No crashing, no artifacts... I checked that SLi is working by switching between single chip and performance mode and comparing results (quite a bit faster in GL_Quake using performance mode).

I still haven't found a donor card with the appropriate hyinx chips...

Who says that only the RAM can be bad? 😀 The short will affect both the RAM and the VSA chip as those are connected to each other.

And it could also be that the physical damage happened after the card became unusable for its previous owner.

kanecvr wrote:

I still haven't found a donor card with the appropriate hyinx chips...

Try to look for GF2 MX or TNT2 Pro - both must use 32MB / 128bit in 4 TSOP 86pin chips. Rated speed 6ns (166 MHz) or better. I'm sure there are still plenty of suitable donors pretty much everywhere. Also it doesn't have to be Hynix, other brands will work as well... just look a bit different.

I gave you an ebay link with propper mems. I don't know if it's ok to put link here. Hope it can help you as it helped me 😀

Just wanted to give on update on this -

IT"S ALIVE!!! I had to replace all memory chips on the 2nd VSA chip, and the card works fine now 😁 Fans need replacing tough.

Thanks for all the help!