Hmm, that's something to keep in mind.

The flexing can really be unnerving, and the crappy plastic mobo standoffs in many old cases don't help the matter. I'm pretty sure the reseating of VLB cards broke some solder joints on one of my motherboards.

I can hear the memory counting, and the system boots up but the screen remains black. This card has very few surface mount devices, so I'm not sure what's going on.

Check the jumpers and try to set them differently (if you don't have the manual for the card). One of my Tseng VLB showed the same symptoms and after setting jumpers differently it works 100% now 😀

> Cardex Challenger

Try the jumper suggestion first.

- The mainboard gives a beep code without any card, but not with this card inserted?
- Does it generate sync signals?
- You can try without the memory upgrade.
- Also try to clean the gold fingers with a rubber (soft side).
- Check that no pins on the solder side are bent and thus touch anything conductive nearby.

With the exception of checking the sync signal, I have tried all other suggestions long ago. I don't have a scope, but the monitor doesn't detect anything and just shuts off. Maybe a bad gendac?

There was another question, aiming at the video BIOS 😀
It may be corrupted or somebody plugged in a non-matching chip.

One more thing would be to replace the crystal, but then there's not much left to try...

I tried the BIOS chip as well. I cleaned the socket and legs. I also dumped it to verify the contents. It seems to be the one that came with the card. Maybe if somebody else with the same card comes along I can ask them for a copy of their BIOS just to be safe.

I might try swapping the crystal, just because I can.

How did you figure out it was the resistor packs?

Continuity tests between them and the other joints to where the traces lead?

I have a couple of unresponsive VLB cards which I may spend come time on and test each of the individual SMD components for bad joints.

Many years ago when I actually bothered to sell my gear when I upgraded I sold an AMD 486/80 system. I was going to sell it to a friend of a friend so I have to take it over my friends house to set it up. When I got it all set up and turned it on the screen showed junk and I could not figure out why. Time was running out before the buyer was going to show up so I took the video card out (Tseng Et4000WP 2MB I think it was) and looked it over. I spotted a bad solder joint on one of the ram module pin so I bent it (RAM chip pin) to make a good contact, everything worked. Sold it, buyer never had any issues with it, everyone was happy.

So even with lead solder sometimes you can have a bad solder joint but nothing like the problems we have with new solder. And yes, long VLB cards do flex a bit.

Anonymous Coward wrote:

I tried the BIOS chip as well. I cleaned the socket and legs. I also dumped it to verify the contents. It seems to be the one that came with the card. Maybe if somebody else with the same card comes along I can ask them for a copy of their BIOS just to be safe.

I might try swapping the crystal, just because I can.

I don't think I have this particular card, but as I don't have an EPROM programmer (am possibly gonna buy one at some point) I wouldn't be able to send you the file anyway.

edit:
Had a look on my list of ISA graphics cards. Apparently I do have an ISA Cardex graphics card, has a ET4000AX 'GPU', but I can't read which chip you have.
I have no idea if this BIOS file could be of any help to you. Is it possible to extract the content of the VGA BIOS without having an EPROM programmer btw?
I never done this, but I'd need to know first if this file could actually be of any use to you

386_junkie wrote:

How did you figure out it was the resistor packs?

Continuity tests between them and the other joints to where the traces lead?

I have a couple of unresponsive VLB cards which I may spend come time on and test each of the individual SMD components for bad joints.

I applied pressure to every component on the PCB until I found the loose IC. I used a c-clamp to make things easier.

@Tetrium

I doubt your ISA ET4000AX BIOS would work very well on my card, if at all. I am not entirely sure, but it may be necessary to match the BIOS up with the card's RAMDAC/GENDAC as well.

You can dump ROMs without a programmer. You can either send commands using DEBUG, or you can use a program called "pcjrcart" which automates the process; all you have to do is specify the starting and ending addresses of the ROM location.

Anonymous Coward wrote:

I applied pressure to every component on the PCB until I found the loose IC. I used a c-clamp to make things easier. […]
Show full quote

386_junkie wrote:

How did you figure out it was the resistor packs?

Continuity tests between them and the other joints to where the traces lead?

I have a couple of unresponsive VLB cards which I may spend come time on and test each of the individual SMD components for bad joints.

I applied pressure to every component on the PCB until I found the loose IC. I used a c-clamp to make things easier.

@Tetrium

I doubt your ISA ET4000AX BIOS would work very well on my card, if at all. I am not entirely sure, but it may be necessary to match the BIOS up with the card's RAMDAC/GENDAC as well.

You can dump ROMs without a programmer. You can either send commands using DEBUG, or you can use a program called "pcjrcart" which automates the process; all you have to do is specify the starting and ending addresses of the ROM location.

Nice, I didn't know it could be done using debug. Used to use it for "formatting" harddrives quickly 😊