Jim 
Phil, this is an anonymous question. We don't even have the initials of the perpetrator of this question It relates to the surface finishes on finished circuit boards.

The question is: "Is there a general specification for an acceptable level of exposed copper for circuit conductors on circuit boards where all the copper surfaces are gold plated?" And a further question is: "What problems could we see if there are very small non-contact areas if some conductors have exposed copper?"

Phil
Well let's see, the answer to the first question, general specification: YES - generally speaking there shouldn't be any exposed copper. As far as potential problems you could see, well obviously that copper can corrode, discolor.

The thing here is, we would politely say, "This a process indicator." It's also a supplier indicator. What kind of crappy boards are you buying? What's with this exposed copper? Hey, it's supposed to be gold plated, why the hell is the copper showing through here? You know what's going on here? This is a process out of control.

Jim
But let's apply some logic to this, Brother Phil. What is special about gold?

Phil
It's expensive..

Jim
Right, so in the fab process that calls for gold coating, obviously any supplier doesn't want to put anymore gold on than they absolutely have to.

Now what happens in real life? Some board fabricators are reasonable about that, they try to keep it low, but well within the specs. Others try to take it right to the limit of the spec.

Phil
And cross over the line.

Jim
So one thing might be here as a process indicator, you might want to measure the average thickness of the gold on the other conductors. Is it not only exposed on little edges and so forth, but is it really thin on the other places so you can't see it, but whereas it might not give you 100% protection? The question I would have is: Are you getting good soldering on all of your joints that are supposedly fully protected by this gold coating?

Phil
And I'd particularly concerned also about the inside barrels of the through-holes and vias. If you're seeing it on the surface, one can only wonder what's beneath. If it's exposed copper, wow, the guys even skipping on the nickel for crying out loud. What is this world coming to, Jim?

Jim
Well remember, we don't know that. There are flash gold finishes that go right over copper, Phil, so we don't know that specifically here. We typically think of a ENIG finish where we get a nickel barrier, and I agree with Phil, but we don't have data here to make that claim.

If it is a ENIG finish and you got exposed copper, then Phil's right; not only did they not get the gold on, but they didn't get the nickel on, which brings us to another thing that you have to ask is: Was it not a problem with the gold being too thin but the fact that the copper wasn't cleaned adequately underneath, and that's death with any finish. Just remember, I don't care what your surface finish is - OSP, immersion silver, HASL or anything, you're starting with bare copper and you've got to get that copper clean.

All of the surface coating methodologies start by actually etching a little bit of the copper off to get any deep seated corrosion. You get down to really virgin copper. So whatever finish you put on gold or nickel or whatever, you can get a good complete coverage.

Phil
So coming or going, it's a process out of control. This case it could be poor cleaning at one end and/or skimping on the gold at the back end. Nevertheless, it might be time to look for another vendor or get this one in line.

Jim
Yes, unfortunately where so much of the PCB fab is gone offshore, the supply link is longer and problems have increased.