The thing that wouldn't leave...
Sometimes black pad seems like that. It should be pretty well understood by now, but it still pops up (or should I say "pops off") now and then. I just got a call from an engineer asking for some help in talking about a black pad problem with his board fab house.
He's having issues with ENIG boards from one specific board house and gull-wing parts. We most commonly associate black pad with BGAs and they do seem to be more susceptible, but any part can suffer from black pad. In this case, the fab house suggested that the specified gold layer was too think and the house said that the thick gold layer was causing the black pad.
I double checked with one of my engineers and he responded that a thicker gold layer may hide the black pad problem for longer then a thin gold layer, but it won't cause it. The industry accepted understanding is that it is a problem with the nickle layer during fab that causes it. As Phil Zarrow states (Circuits Assembly April 2008, pg 20), IPC Class II calls for a minimum of 1.97 um of gold, but sometimes that may not be quite enough because of normal surface variations. His failure analysis guru friend suggests 3, minimum and nominal 4 micro meter thickness.
There may be some issues with too thick a gold layer, but Phil and everyone else I've read from or spoken to is pretty clear that it's the nickle layer processing or too thin a gold layer that leads to black pad.
Pop goes the SSOP