Quick: If you haven’t had your chance to answer the question, go back and read it first.
Done? Okay.
#1. This is an OSP (organic surface preservative) finished PCB. That in itself isn’t a problem. However, take notice of the two different colors of pads. The darker pads are oxidized or otherwise contaminated — not a well-preserved surface. And that means that this board isn’t going to work.
#2. Oops. I already answered #2 in with my answer to #1. It’s got some oxidized/contaminated pads. Perhaps the PCBs weren’t stored properly, were kept too long or the OSP finish wasn’t well applied at the fab house.
#3. Planet 10.
#4. Real Soon.
And, no it can’t be used as is. The solder most likely just wont take on those bad pads.
A couple of other notes: The mask registration isn’t bad. Not perfect, but not bad. Big pitch BGAs like this do tend to work best with NSMD (non solder mask defined) pads like this. That permits the BGA ball to sag a bit and grasp the side edges of the copper pad for improved adhesion. (Note that for extra fine pitch BGAs, like 0.4 mm pitch, you need soldermask defined pads) And, note that there is a good web of soldermask between the pad and the via. This will keep the solder from being sucked down the via.
Duane Benson
I’ve been ionized, but I’m okay now
“Note that for extra fine pitch BGAs, like 0.4 mm pitch, you need soldermask defined pads”
No, not if the .4mm pitch WLCSP was designed correctly. You’re thinking ‘traditional’ .4mm pitch. Think ‘staggered’ .4mm pitch, and you get away from SMD pads. FYI – most chip manufacturers WILL NOT go to ‘traditional’ .4mm devices because of this aspect. SMD pads are way too costly because of the registration requirements.
“And that means that this board isn’t going to work.”
If the substrate is a $1 part, I’d tend to agree with you… scrap that board and move on. However, what happens when it’s a few hundred dollars or even a few thousand dollars worth of backplane or maybe it’s a valuable one-off prototype that is weeks overdue? Will I let a few oxidized pads stand in the way of my build? Heck no.
In this scenario, I’d have to give it some TLC, but it’s definitely a viable board. In order to build with this, you’d have to perform a careful removal of the defective OSP finish, and then deoxidize those pads under a scope. You’re looking to get down to bright, shiny copper (or whatever your base solderable layer happens to be).
That base will oxidize once exposed, so I’d immediately follow up the cleaning with a quick swab of No-Clean gel flux on each of the exposed pads. That will protect them for a short time while you print, populate, and flow the assembly.
Follow up with X-ray for confirmation of good wetting at those locations.
Hi Mitch;
You’re right and “Isn’t going to work as is” would certainly be more correct. A lot of defects can be fixed or worked around if the cost structure and ultimate reliability allow for it. In the world we live in, raw PCB costs in the thousands of dollars is not unheard of and for that, plus the down-time, you can justify a lot of rework.
Duane
Pingback: Lots of Things Can Be Fixed at Circuits Assembly blog