I recently had similar issues with the time machine build in my signature. My ATA133 card seemingly randomly isn't detected. I haven't had time to sort it out yet. Come to think of it, I should probably take all the cards out and check the "teeth" in the PCI and ISA slots. I know that sometimes they can get stuck open in such a way that inserting a card doesn't make a reliable connection. I wonder if blasting the ports with Deoxit D5 would be a good or bad idea.

Also, don't totally disregard the possibility of a resource conflict. I've had issues with other boards that I swore were hardware but in the end the problems were fixed by changing resource settings in the BIOS. I don't know how this works exactly, but is there a chance that a conflicting device won't even be detected since it has no resources?

Contacts get old, worn and dirty. It's a known problem. It also affects memory modules in my experience. An old toothbrush and isopropanol alcohol can be used to clean them. Inside the PCI slots too. Also use an air duster with a tube nozzle to blow any dust out of the slots.

Even with all this, contacts are designed for only a certain number of insertions, and will start to fail eventually. There's nothing that can be done about this.

Errius wrote:

Contacts get old, worn and dirty. It's a known problem. It also affects memory modules in my experience. An old toothbrush and isopropanol alcohol can be used to clean them. Inside the PCI slots too. Also use an air duster with a tube nozzle to blow any dust out of the slots.

Even with all this, contacts are designed for only a certain number of insertions, and will start to fail eventually. There's nothing that can be done about this.

What about replacing them?

Thanks guys, I used some non conducting contact cleaner on the pci cards and brushed them with a toothbrush. Then wiped off the excess cleaner and waited for 30 minutes. They immediately worked (also the gold on the connector is shiny now) not sure if problem is solved but turned the PC off and wiggled the cards a bit and still working fine.

Also gonna take a look at the slots if the problem resurfaces. Ports being stuck open sounds interesting, some slots require a bit more force to insert the cards.

Errius wrote:

Contacts get old, worn and dirty. It's a known problem. It also affects memory modules in my experience. An old toothbrush and isopropanol alcohol can be used to clean them. Inside the PCI slots too. Also use an air duster with a tube nozzle to blow any dust out of the slots.

Even with all this, contacts are designed for only a certain number of insertions, and will start to fail eventually. There's nothing that can be done about this.

True. like the Slot1 CPUs are designed by Intel for about 50 insertions. After this number, the thin gold coating disappear and the contact isn't as good. I wonder how much insertions PCI/ISA cards are designed for. I guess they are really more tolerant to bad contacts than CPUs though

.legaCy wrote:

Errius wrote:

Contacts get old, worn and dirty. It's a known problem. It also affects memory modules in my experience. An old toothbrush and isopropanol alcohol can be used to clean them. Inside the PCI slots too. Also use an air duster with a tube nozzle to blow any dust out of the slots.

Even with all this, contacts are designed for only a certain number of insertions, and will start to fail eventually. There's nothing that can be done about this.

What about replacing them?

Then you have to desolder the whole slot and replace it entirely. Usually they are made in such a way that you can't take them apart easily when they're mounted to something.

I've actually never had problems like that. But, before inserting any card or module in my retro hardware, I use some lubricating contact spray. It's basically contact cleaner that contains a lubricating oil. I spray some on a cotton pad and wipe down the contacts. This leaves a very thin lubricating film which prevents the contacts from scratching/wearing and also from oxidation. It also makes the cards go in noticeably easier (less force required).

Perhaps it's the reason why my Audigy 2 isn't detected by device manager?

Kreshna Aryaguna Nurzaman wrote:

Perhaps it's the reason why my Audigy 2 isn't detected by device manager?

Probably, I've had this problem many times, cleaning it worked for me. Try cleaning it.

yawetaG wrote:

.legaCy wrote:

Errius wrote:

Contacts get old, worn and dirty. It's a known problem. It also affects memory modules in my experience. An old toothbrush and isopropanol alcohol can be used to clean them. Inside the PCI slots too. Also use an air duster with a tube nozzle to blow any dust out of the slots.

Even with all this, contacts are designed for only a certain number of insertions, and will start to fail eventually. There's nothing that can be done about this.

What about replacing them?

Then you have to desolder the whole slot and replace it entirely. Usually they are made in such a way that you can't take them apart easily when they're mounted to something.

Seems pretty doable with a dessoldering station.

Server hardware is built to tougher standards and will likely last longer. Something to bear in mind.

Worn out slots? Is this really a thing? I mean, I've heard of it on consoles, where the cartridge slot would be worn out after thousands of insertions (heh), but on PCs? I have a bunch of C64s, and their cartridge slots have doubtlessly seen many, many more insertions (chuckle) than any PC slot would have, and none of them have any issues.

okenido wrote:

True. like the Slot1 CPUs are designed by Intel for about 50 insertions. After this number, the thin gold coating disappear and the contact isn't as good. I wonder how much insertions PCI/ISA cards are designed for. I guess they are really more tolerant to bad contacts than CPUs though

50 insertions for slot 1? Do you have any links where this information can be found? It seems quite low, of course most users will never insert a CPU more than 2-3 times, but I am quite sure that the slot 1 can take *much* more than 50 insertion before starting to fail.

In my opinion it's more likely for the internal slot to fail, because the internal connectors will lose their "grip", also they might start to sag after a while, especially if the card inserted in them is heavy. Then the top part will not make contact with the card. But this shouldn't happen for slot 1, since they are usually placed vertically, not horizontally, like the slots.

Even if you can replace PCI slots, what can you do about worn contacts on expansion cards?

And not just cards. I have an old 5.25" floppy drive with a very worn edge connector. How to fix this?

Errius wrote:

Even if you can replace PCI slots, what can you do about worn contacts on expansion cards?

And not just cards. I have an old 5.25" floppy drive with a very worn edge connector. How to fix this?

A terrible way to do would be follow the end of the trace, scrap the solder mask and solder wires and then another edge connector on the floppy drive. about worn connectors on the card itself idk if there is a easy way to "replate" it again.

RaverX wrote:

okenido wrote:

True. like the Slot1 CPUs are designed by Intel for about 50 insertions. After this number, the thin gold coating disappear and the contact isn't as good. I wonder how much insertions PCI/ISA cards are designed for. I guess they are really more tolerant to bad contacts than CPUs though

50 insertions for slot 1? Do you have any links where this information can be found?

Sure it's on Intel datasheet page 21 :

http://www.dexsilicium.com/Intel_Pentium2.pdf

4. The mechanical integrity of the latch arms is specified to last a maximum of 50 cycles. 5. The electrical and mechanical inte […]
Show full quote

4. The mechanical integrity of the latch arms is specified to last a maximum of 50 cycles.
5. The electrical and mechanical integrity of the processor edge fingers are specified to last for 50 insertion/extraction
cycles.
6. While insertion/extraction cycling above 50 insertions will cause an increase in the contact resistance (above 0.1Ω) and
a degradation in the material integrity of the edge finger gold plating, it is possible to have processor functionality above
the specified limit. The actual number of insertions before processor failure will vary based upon system configuration
and environmental conditions

0.1 ohms is not much but I guess it could have an impact if you have a fine-tuned overclocking/undervolting.

Thanks, it's a nice document. I think Intel wants to be "safe", as Intel always does, they probably listed "the worst" scenario. I don't like to destroy hardware, especially old hardware, and Pentium II is my favorite CPU, so I won't test a CPU by inserting/pulling out in/from the slot 1, mostly because I'm sure it will last way above Intel specs 😀