It might just be some sort of feed through capacitor. Can't be too high a value if it's on the video output of the card, the bulky size is probably for extra voltage rating (so that any grounding issues with the monitor don't kill the cap). You might get away with just putting a reasonably thick piece of wire there, ignoring the center terminal. Or use a U-shaped wire and put a standard cap inside the U, soldering it to the center terminal pad on the PCB and the wire on the other end. Not pretty but it'll work pretty well, the cross-section of the U wouldn't be big enough to pick up (or radiate) a lot of EM interference.

A long time ago I was specially remove filters from some low-cost cards to get a sharp picture in the 1024 screen resolution.

filters removal worked great with CRT monitors as those dont really have a horizontal resolution and circuity filters on its own. LCDs will pick up harmonics and might show picture degradation.

Deunan wrote on 2024-06-24, 13:39:

It might just be some sort of feed through capacitor. Can't be too high a value if it's on the video output of the card, the bulky size is probably for extra voltage rating (so that any grounding issues with the monitor don't kill the cap). You might get away with just putting a reasonably thick piece of wire there, ignoring the center terminal. Or use a U-shaped wire and put a standard cap inside the U, soldering it to the center terminal pad on the PCB and the wire on the other end. Not pretty but it'll work pretty well, the cross-section of the U wouldn't be big enough to pick up (or radiate) a lot of EM interference.

Hm interesting idea , ill consider it , i was just wondering if there was some off the shelf emi filter that might fit the bill , the voodoo 4500 pci seems to use a similar setup near the vga port so there has to be some smt filter that would do the job , problem is i don't know the values of the voodoo filters or if they would match my buffalo virge card.

quwy wrote on 2024-06-24, 16:54:

A long time ago I was specially remove filters from some low-cost cards to get a sharp picture in the 1024 screen resolution.

This doesn't seem like a low cost card as it was made in japan in the late 90s so id rather not rip the filters out as I'm sure there's a reason for them being there.

AppleSauce wrote on 2024-06-25, 02:50:

i was just wondering if there was some off the shelf emi filter that might fit the bill

Sure there are: https://mouser.com/c/passive-components/capac … ugh-capacitors/

In fact I think your PCB can use either SMD or THT parts, there is an empy space nearby that's connected to the SMD pads. The main problem is the value, I would guess it at most 30pF (and probably much less) because any higher capacitance will only degrade the video signal at higher resolutions / refresh rates. Keep in mind this card was most likely designed with CRTs in mind, possibly what we now consider lower-end monitors with rather limited horizontal frequency range. As it was already mentioned this might not really work all that well with modern flat screen monitors, especially if the resolution is not native and the input chips in the monitor have to find/guess and match the pixel frequency and phase.

Deunan wrote on 2024-06-25, 11:23:

AppleSauce wrote on 2024-06-25, 02:50:

i was just wondering if there was some off the shelf emi filter that might fit the bill

Sure there are: https://mouser.com/c/passive-components/capac … ugh-capacitors/

In fact I think your PCB can use either SMD or THT parts, there is an empy space nearby that's connected to the SMD pads. The main problem is the value, I would guess it at most 30pF (and probably much less) because any higher capacitance will only degrade the video signal at higher resolutions / refresh rates. Keep in mind this card was most likely designed with CRTs in mind, possibly what we now consider lower-end monitors with rather limited horizontal frequency range. As it was already mentioned this might not really work all that well with modern flat screen monitors, especially if the resolution is not native and the input chips in the monitor have to find/guess and match the pixel frequency and phase.

Thank you , it seems like feed through capacitors would probably be the type of missing component ,
I'm guessing I can just measure the amount of picoFarads with an appropriate meter if I remove one of the other feed through caps?
Plus I wouldn't worry about the screen , the card will be connected to a G500 Trinitron CRT.

Hokay so I got the duplicate card in recently , but the system wouldn't boot with the card , the pcb looked pretty warped some maybe that was why , I pinched one of the filters from the dead card and used it on the working one , and the system booted with the non dead card , installing the drivers was a bit of a pain since it was all in Japanese and I nearly nuked my windows 3.1 install due to missing font issues.

The card can go up to 1920x1440 in win 3.1 , or I can also go lower at 1280x1024 but run it at 16.7 million colors.
Weirdly the utilities for Win95 are worse than the decent 3.1 ones , and had a dingy little drop down menu , and I had to go into a .ini to manually set the refresh rate down to 75 from 87 because at 87 hertz the card kept running in interlaced mode.

The card can natively run croc in virge mode at 1024 , 768 which is nice , but not super useful cause its super slow,
also for some reason every resolution but 320x200 works well , which has artifacting , I'm not sure if its a ram issue or not ,

I also tried epic pinball in DOS 6.22 and the very upper right corner edge was missing a few pixels , but only in the table select menu? , which is odd.
I guess I can always steal the ram from the dead card to fix things , if that's the case , unless its something to do with the card bios not liking certain modes?

Anyway ill have to do more testing.