Funny how completely unrelated hardware sometimes shows identical issues. Last night was corroded contact night. I had a pile of I/O controllers and an unresponsive motherboard to test. Nothing worked first time and every single one had at least one contact issue.
The board (the last of my Morse 486 ISA boards, and only one not just to work or instantly blow a tantalum cap) turned out to have issues with the LIF CPU socket. Jiggling around the CPU got it to start booting, a bit more movement & pressure got it doing so reliably. First (lightly) corroded contacts.
Then the most interesting I/O controller - a Longshine LGS-6633 rev b1. ISA IDE+FDD+SCSI controller. First up it didn't get detected at all - corroded ISA contacts - then it didn't detect the FDD - corroded 34p contacts. Eventually I managed to get it booting from floppy. Then moved on to IDE. Same story, absolutely no detection at first, then detection *something* was there but not able to use it, even after entering correct CHS in BIOS. Oh well, goes beck onto the 'to do' pile.
Afterwards three more normal, simple ISA I/O controllers with either ISA, 34p and/or 40p connectors needing TLC. They all did work fine eventually, but this was more effort required than in ages. The weird thing is they came from three different sources and each of them was in a box with other hardware with no corrosion issues. Just pure coincidence I suppose.
Fortunately the last bit of testing went smoother: Soundblaster 16 CT1750 and AWE64 CT4520 did what they were supposed to as soon as I hooked them up (and ran CTCM.EXE in the case of the CT4520). All in all a satisfying evening, and one to remind me to check even better for dodgy connectors in future.