The risk is not just shorting any pins but interruptions between traces and the soldering lands, even when you proceed very carefully. Pulling the slot a little too hard can tear a trace off the landing.
I always take pictures before soldering the new slot so I can check for continuity of all the connections in case of any problems.

You should also check for solder blobs very carefully. They sometimes slip under some component and you don´t notice because you are focusing on the soldering.

majestyk wrote on 2023-07-09, 06:58:

The risk is not just shorting any pins but interruptions between traces and the soldering lands, even when you proceed very carefully. Pulling the slot a little too hard can tear a trace off the landing.
I always take pictures before soldering the new slot so I can check for continuity of all the connections in case of any problems.

You should also check for solder blobs very carefully. They sometimes slip under some component and you don´t notice because you are focusing on the soldering.

the RAM slot came out with no resistance and no delaminated vias, but i think i was a bit slow when doing the desoldering and the temp was at 450C, which may have been too high so i could've cooked it.

It's almost impossible to determine the quality of solder joints visually. They might look perfect, but be anything but.

In my experience, it's almost always bad solder joints when something like this fails, especially so when dealing with thick multi-layer laminate and large components. I'd try reflowing the joints, using a quality no-wash flux, and of course redo the tests you did before testing the computer again.

If you get --- on the POST card, the issue isn't localized to RAM. Assuming this is not a trivial problem like forgetting to re-insert the processor after swapping the RAM slot (or not inserting it firmly enough), the issue is most likely on the frontside bus. The address lines to the RAM are driven by the chipset or (on old boards) external RAM address multiplexers. Even if there were shorts, the board should show some POST codes. The same is true for /RAS, /CAS, /WE. On the other hand, the data lines on the RAM slot is connected directly to the FSB on a lot of 4868 boards.

Thus, I suggest to check continuity from the data pins D0-D31 to the RAM slot, and especially check them for shorts to GND or +5V.

mkarcher wrote on 2023-07-09, 07:42:

If you get --- on the POST card, the issue isn't localized to RAM. Assuming this is not a trivial problem like forgetting to re-insert the processor after swapping the RAM slot (or not inserting it firmly enough), the issue is most likely on the frontside bus. The address lines to the RAM are driven by the chipset or (on old boards) external RAM address multiplexers. Even if there were shorts, the board should show some POST codes. The same is true for /RAS, /CAS, /WE. On the other hand, the data lines on the RAM slot is connected directly to the FSB on a lot of 4868 boards.

Thus, I suggest to check continuity from the data pins D0-D31 to the RAM slot, and especially check them for shorts to GND or +5V.

i inserted the DX-33 and the RAM in the good slot before i tested it, it seems to be completely dead, i guess i may have overheated some IC that was near the area

you dont need ram for post codes

Pics of the area?