Return it.

If you don't want or can't return it, I would leave it like this. Check for continuity between the pads and the sinkholes/vias/components. The traces and pads don't look that bad...
Check for shorts first, use the GND and all the voltage pins on the card. After that, does the main IC get warm? Any beeps from the system you have the card in?
If you happen to have an oscilloscope, you can check that crystal as well.

Solder is the worst thing you can do to this, it guarantees the death/permanent unreliability of every VLB slot you then plug this card into and from experience, the oxide layer on the copper or nickel is far more forgiving than that of solder.

There are ways to replate the pins but they're not very DIY friendly from what I know.

I would be leaving the card as is, or return it if possible.

Alright, I'll leave it as is for now. Unfortunately I cannot test it right now as I'm away from home but I think the card will work. And if it works I'll be even more annoyed because of the pins. Too bad, some people just don't know what they're doing.

I cannot return it as the postage was like double of the card's actual cost. It was a gamble that may or may not pay out.

Tiido wrote on 2024-10-25, 14:15:

Solder is the worst thing you can do to this, it guarantees the death/permanent unreliability of every VLB slot you then plug this card into and from experience, the oxide layer on the copper or nickel is far more forgiving than that of solder.

There are ways to replate the pins but they're not very DIY friendly from what I know.

I would be leaving the card as is, or return it if possible.

Electroplating should be an option, as long as you've got some container, some nickel strips, some wires and a bench power supply with constant current / voltage, right?
How is nickel plating compared to gold, surely that's better than just the copper layer.
I've seen how bad HASL solder plating gets when in contact with gold contacts, the HASL blackens in the contact areas pretty quickly even if it mostly works.

The card is usable as-is for testing and should work for a while safely with the gold layer removed. It's probably a loose pin on the QFP, or a clock crystal. Check the jumpers too, I can't find the post but I had a similar TGUI9440 card and it was only not working because a memory bank selection jumper was missing.

Let's say I want to do some electroplating procedure, then I guess I should remove the last remnants of the original coating, right? How, with some polishing tool?

There are abrasives made purposely for PCB cleaning, like this one: CIF AR21.
I generally use this, plus fiber glass pencil, and for very strong situation even a rubber tip for micro drilling machines (to be used very carefully).

Electro plating is possible, but the kits costs a lot, and DIY solutions are not easy. You need to do a lot of tests before get the correct setup. While the kits give you the correct everything, time, current, solutions ecc.
I think you can even solder the pins before electroplating them, so you will be sure to have hall the dirt and oxidation removed.

Thermalwrong wrote on 2024-10-26, 01:47:

Electroplating should be an option, as long as you've got some container, some nickel strips, some wires and a bench power supply with constant current / voltage, right?
How is nickel plating compared to gold, surely that's better than just the copper layer.

I don't know all the details myself, but nickel is the first stop under the gold plating in most cases. Gold doesn't like copper very much but it attaches to nickel very well, and nickel goes well on copper. Cheap(er) stuff is bare nickel layer so that alone definitely works.

You do need galvanic connection to every pin, it isn't a matter of just submerging things with voltage applied. I'm not sure what is the best way to get every one of the pins connected, perhaps a comb like brush.

Nickel first, gold later. Elctroplating is done in steps. There was some tutorial on YT. I can't find them now.
But an adhesive copper strip is used to connect all the pins, then one end of the power supply is connected to the strip. The other end to a sort of pencil with the end filled with the solution. So it can be applied to the pins by rubbing it to them.

No. You would dip the the card edge into a resist layer, like urethane wax, then use a wooden tool to remove the areas on the fingers.

THEN, you would use a passivating solution to remove any oils and oxides from the fingers.
THEN, you would use a homebrewed clip, to connect the fingers to the cathode of your elecroplating rig, and connect a carbon or glass annode, then submerge just the fingers (with the resist on), into a stannic chloride bath, and plate a thin layer.
Rinse in distilled water, re-passivate, and THEN put into an auric chloride electroplating bath.
The existing gold will be fine, and will accept the gold no trouble. The passivated tin layer will help it stick to the copper.

Keep in mind, that even doing everything possible to protect the card from the dangerous acidic mix that are the stannic and auric chlorides will NOT guarantee you won't ruin the card.

When you are done with the plating process, remove the urethane wax resist with a warm water treatment and a melamine sponge, with just a tad of detergent.

The process is a little complicated but doable. However my main concern is cost - the card is not rare or valuable enough to justify electroplating with gold. So I thought maybe nickel plating would be enough. Found on wiki there are multiple different nickel solutions, one of them is conductive, hard and shiny, which would probably ideal for card fingers that need to be inserted into a slot. The result would not be as good as an "out-of-factory" product but would probably be good enough.

While a bit overkill, this science paper details some lab techs screwing around with cleaning and passivation solutions for copper.

https://zzz.physics.umn.edu/!lowrad/_media/me … passivation.pdf

They basically hit upon using a mixture of citric acid and hydrogen peroxide to clean and passivate the copper prior to performing other surface treatments on equipment designed to test for radioisotopes.

The same process could be used to less-aggressively clean your card edge here, and ensure good adhesion of the subsequent plating bath treatments.

Yeah, cleaning of the pins is another problem. And what about chemical nickel bath? Seems rather cheap and available but I'm not sure if the result is good for this specific application.

Did some research online and there were kits for re-plating damaged fingers on PCBs, as demonstrated in an Youtube video. Looked at their website and they don't offer the kits anymore but I guess you can get one of those somewhere else. I'm not sure what the price of those kits was.

The process seems to be quite easy.

EDIT: Another YT video.

Pace once have a kit for that. But I'm not sure they sell it anymore.

CircuitMedic have a kit, and I think they still sell it.
But, as I said they are very expensive: https://www.circuitmedic.com

The kit in the second video is interesting. The site of the company who sell it seems too much home made.
Maybe I will check if is sold elsewhere.

In a pinch, you could try spraying with Deoxit. It supposed to leave a coating designed to prevent oxidation of the metal.

Probably not a long term solution, but could do in the mean time.

The card is not worth investing money in, it will hardly ever regain its original look, because parts of the soldering mask around the edge connectors were also sanded away, I'd just polish them if I were in your place. Tinning is another option, but the end result won't be pretty.

It may not be worth it for this specific card (even though it's a very nice VLB card) but it may be interesting to find a cheap and good enough method for hobbyist. For example I found several tools on Aliexpress that are similar to the one found in the YT video. And the price is not that bad.

And there is a plating pen for applying nickel/gold solution.

Once found this solution/procedure could be used cheaply for other PCBs, not just PC hardware.

Understandable, in such case, the best price-quality ratio for plating bare copper would be achieved by using nickel salts solution (nickel sulfate/chloride) which can be bought or made at home. For the electric current source, you can use any old AT/ATX PSU, 3.3-5V rails should do just fine. Now for the plating method you have two options here - either buss all of the pads together and submerge the edge connector into the nickel solution or you could plate them separately with a swab tip. Swab can be cheaply made from a Q-tip or a depleted marker pen.

Tiido wrote on 2024-10-25, 14:15:

Solder is the worst thing you can do to this, it guarantees the death/permanent unreliability of every VLB slot you then plug this card into

Does this mean the HASL for pcb finish is a hard no-go if I'm about to experiment with making ISA cards, and ENIG is the only way to be considered?