To me this sounds like a missing component, break in solder joint(s) or in the pcb traces.

There may be a signal or signals that need to be pulled to ground for example in order to card to function correctly. Now if that line is broken, or if the possible pull down resistor is broken or missing that line will float in rather unpredictable voltage levels and poking it with fingers is usually enough to make it flip states.

I have seen this kind of fault several times before.

Just bumping this... I just tried this card again and it seems to be working more reliably (no idea why), but I still get graphical corruption that slowly creeps in when all the memory sockets are populated. If I touch that area on the back of the card (it's hard to tell but I think it is pin-11 of either of the SOJ-40 RAM sockets) the corruption changes. I can gently poke this pin or the area around it a few times and it will eventually clear right up. I checked the pinout for SOJ-40 memory and it looks like this pin is not connected, and this reflects what I found when testing with my DMM (nothing was connected to the pin). It also seems to react when I merely get something CLOSE to this pin (even a piece of plastic), so that just makes it more odd. The card works fine if I keep prodding this area (on either RAM socket) when the corruption starts creeping in.

I installed the NV1 Windows 95 drivers from Vogons Drivers (they say they are for the Edge 3D) and the card actually works quite well in Windows 98SE. It has limited color support (16bit max) but it functions fairly smoothly. On top of that, the digital audio and built in wavetable synth both sound decent considering they're just a secondary feature of what is one of the very first 3D accelerator cards. I just pointed device manager to the "nv" folder in the nv1 driver package and it was able to install both devices this way. I may have some more tinkering to do though because when I try to run the 3D Shapes demo (part of the NV1 SDK) the rendering area of the demo just gets broken (frozen image in that area) when trying to load either of the two demo objects. Could be related to the hardware issue, or maybe I just need to keep playing with the drivers.

Anyway, this is an interesting card. I'll be happy if I can at least see it accelerate something in 3D. 😀

Kizmo wrote on 2019-03-03, 10:57:

To me this sounds like a missing component, break in solder joint(s) or in the pcb traces.

There may be a signal or signals that need to be pulled to ground for example in order to card to function correctly. Now if that line is broken, or if the possible pull down resistor is broken or missing that line will float in rather unpredictable voltage levels and poking it with fingers is usually enough to make it flip states.

I have seen this kind of fault several times before.

Just wanted to update this after two years.

I've examined the card very closely and I haven't noticed any other bad solder joints. I also replaced all the caps on the card (was feeling ambitious) and it seems to be less prone to doing the weird artifacting thing, but it DOES still do it. Very strange. To make it go away I just swipe my finger across the back of the memory socket solder joints. I can turn the artifacts on or off this way actually... using my finger or a piece of plastic, without even touching anything some times. Once I make the stuff go away, it will work for a long time and may or may not even come back until I take the card out and put it back in.

None of it seems to hurt the stability of the system. I don't get any freezing or crashing in Windows 98SE.

However, I still can't get the 3dshapes demos to run. They still just give a blank screen.

Any further suggestions would be much appreciated, for either the artifacting or the 3dshapes demo not working. Maybe the demo doesn't work in 98SE or it doesn't work with the newest Edge3D driver?

id trace out those memory joints and check for voltage that isnt supposed to be there. perhaps why it gos away is because you are grounding discharging it. either that or theres a loose ground and its not connected, or the ground is grounding to voltage somwhere its not supposed to. or maybe something stupid like the vram socket is actually bad, but i cant believe you wouldnt of checked or soldered a new one if thats the case

it kinda sounds like a grounding issue try running a alligator clip to the black of a molex plug and the other to the ground plane of the card

see if you have any ICs that look like they drain to the ground plane maybe they are bad? you could also put a volt meter between black on the molex and the ground plane and see if you have any potential.

I had a similar problem with my stereo. the case would build up charge and glitch after a while so I ran a wire from the case to the wall plate screw 🤣

I may be totally off base but hell, you can safely test it and tell me I'm wrong 🤣

Did you check for shorts on ceramic caps (or did you replace these too)? They sometimes short themselves out, it happened on one of my slotkets.

I've just got a 3240 and it works fine but has a non-stop interference pattern across the screen. The angle of the interference lines varies depending on resolution.

I tried 2 different LCD monitors, put PCI card into a different slot, wiggled cable about, etc, but still get these: https://imgur.com/a/28nIV8o

These glitches are not present with my ATI Rage 128 Pro card (or, indeed, any of the other cards I've put in the PC), so it's not the VGA cable.

I'm concerned that:

1. it might be something hardware side wearing out (as it's a 27 year old card now!) and I'm no soldering expert (especially as many of the capacitors on the card are *tiny*)

2. it's likely to show up if/when I use VGA capture hardware

Anything else I can try? I could try touching it (as per other story above) as long as that's not likely to short anything out or electrocute me 😂

VirtuaIceMan wrote on 2022-04-06, 08:55:

I've just got a 3240 and it works fine but has a non-stop interference pattern across the screen. The angle of the interference […]
Show full quote

I've just got a 3240 and it works fine but has a non-stop interference pattern across the screen. The angle of the interference lines varies depending on resolution.

I tried 2 different LCD monitors, put PCI card into a different slot, wiggled cable about, etc, but still get these: https://imgur.com/a/28nIV8o

These glitches are not present with my ATI Rage 128 Pro card (or, indeed, any of the other cards I've put in the PC), so it's not the VGA cable.

I'm concerned that:

1. it might be something hardware side wearing out (as it's a 27 year old card now!) and I'm no soldering expert (especially as many of the capacitors on the card are *tiny*)

2. it's likely to show up if/when I use VGA capture hardware

Here's a pic of my actual card (whilst it was still on eBay for sale), it looks pretty tidy: https://i.ebayimg.com/images/g/c5AAAOSwBvBhmM4h/s-l1600.jpg

Anything else I can try? I could try touching it (as per other story above) as long as that's not likely to short anything out or electrocute me 😂

Check the black things behind the vga port.

One person noticed that theirs were cracked when one stood up in two pieces while reflowing the solder in the area with a heat gun.

Do you mean the 3 small black boxes right next to the back of the blue VGA port?

VirtuaIceMan wrote on 2022-04-06, 14:43:

Do you mean the 3 small black boxes right next to the back of the blue VGA port?

Yes

Okay I've got pics of the card (the one on eBay wasn't the one I received so I think was a placeholder). I've also taken pics of the 3 VGA boxes (as I've called them). They're a bit wonky lay out (as are a few other bits on the card), see pics here: https://imgur.com/a/elrwhtN (right-click each and open in new tab to see full detail).

Any ideas of what I can try next? I have almost zero soldering skills and these boxes aren't standard caps...

VirtuaIceMan wrote on 2022-04-06, 19:18:

Okay I've got pics of the card (the one on eBay wasn't the one I received so I think was a placeholder). I've also taken pics of the 3 VGA boxes (as I've called them). They're a bit wonky lay out (as are a few other bits on the card), see pics here: https://imgur.com/a/elrwhtN (right-click each and open in new tab to see full detail).

Any ideas of what I can try next? I have almost zero soldering skills and these boxes aren't standard caps...

I think they are filters of some sort. What kind of reading are you getting across them? I think they should all be the same,

Idk if those are the problem, it’s just a hunch. You can also try a recap as well.

I don't have any technology or skills to do a reading, or a recap 🙁

VirtuaIceMan wrote on 2022-04-06, 23:27:

I don't have any technology or skills to do a reading, or a recap 🙁

Time to get a volt meter :p get a good one, like a fluke.

Thanks for the pointer but they look a bit pricey and I've spent enough on cards 😒

I could borrow a multimeter though, but do I test when it's plugged in or on a bench? And even if I find the culprit, if it's something as small as those black boxes then I wouldn't know what to replace it with and of very small, easy to damage with soldering iron.

Is there anywhere online where people do these sort of fixes or can be asked for help?

VirtuaIceMan wrote on 2022-04-07, 07:47:

Thanks for the pointer but they look a bit pricey and I've spent enough on cards 😒

I could borrow a multimeter though, but do I test when it's plugged in or on a bench? And even if I find the culprit, if it's something as small as those black boxes then I wouldn't know what to replace it with and of very small, easy to damage with soldering iron.

Is there anywhere online where people do these sort of fixes or can be asked for help?

Well, it’s a bit of a complicated first foray into electronics repair. Maybe your best bet is to sell it on ebay and note the issue on the description. And buy a different one.

Or we could try and help, but you may end up destroying the card.

Anyway, yeah. Borrow a volt meter and set it on ohms. And read with it sitting on a non conductive bench

I picked it up for half the price of other ones so I think I'll use it for captures then maybe look at fiddling with it, in case something goes wrong! Thanks for all the advice though 😀

Ozzuneoj wrote on 2019-03-01, 15:20:

[..] but when I squeezed the chips\sockets with my fingers I noticed that the corruption changed and went away a little. Eventually I realized that it wasn't the pressure that was doing it, but it was me actually touching the solder points on the back of the memory sockets that was making all of the corruption go away.

"hand capacity"?

Edit: https://en.wikipedia.org/wiki/Body_capacitance

I remember it from toying with radio experimenters kits. 😀

VirtuaIceMan wrote on 2022-04-07, 07:47:

Is there anywhere online where people do these sort of fixes or can be asked for help?

local hackerspaces are best bet