Selective Surface Finish: ENIG and OSP on the Same Board

Surface finish is usually specified as a single choice for the whole board, but some products genuinely need two. A design may require a flat, gold plated surface on a fine pitch connector and a cheap organic finish everywhere else, or a solderable finish for the assembly area combined with a hard finish on an edge connector. Selective surface finish processes make that possible by applying different treatments to different regions of the same panel.

Why Use Two Finishes on One Board

The electrical and mechanical requirements differ by region. A press fit or edge connector needs a hard, wear resistant surface that survives repeated insertion, which normally means hard gold. Wire bonding needs a very flat, very clean surface. The main assembly area simply needs good solderability for one reflow pass, and the cheapest finish that provides it is usually enough.

Applying the most expensive finish across the entire board wastes money and, in some cases, hurts performance. Gold on a large solder area can embrittle joints if the thickness is not controlled, and an organic coating is unsuitable for a connector that will be handled. Selective processing lets each region receive what it actually needs rather than forcing a single compromise.

How Selective Finishing Is Done

The common method applies a plating resist that covers the regions to be protected while the exposed regions are plated. The sequence matters: finishes that require immersion in a plating bath are applied first, then the resist is stripped and an organic finish is applied to the remaining areas. Getting the order wrong means the second bath attacks the first finish.

Masking can also be done with tapes or dedicated fixtures, which is practical for small panels and prototypes but awkward in volume production. Whatever the method, the quality of the mask edge determines the quality of the finished board, because a poorly defined boundary leaves a band of uncertain finish that neither process controls properly.

Not every combination is practical. Hard gold over nickel is electroplated, which needs a plating bus and therefore demands care when the region is not otherwise connected to a net. Immersion finishes such as ENIG, immersion tin and immersion silver cannot be mixed freely, because each requires a different chemistry and a different surface condition on the copper. The organic coating is the most permissive partner, since it can be applied last to almost any exposed copper that has not yet been treated. A sensible design keeps the number of finishes to two and defines the boundary clearly.

ENIG: Gold Thickness and Black Pad Risk

Electroless nickel immersion gold gives a flat surface that suits fine pitch and wire bonding. The nickel layer provides the barrier and the mechanical strength, while the thin gold layer prevents oxidation and keeps the surface solderable. Thickness control is everything: too little gold and the nickel oxidises through it, too much and the joint becomes brittle because the gold dissolves into the solder.

The classic defect is black pad, in which the nickel surface corrodes during the immersion gold step and produces a joint that looks acceptable and fails mechanically. It is linked to the nickel bath chemistry and to over aggressive gold plating, and it is one reason to keep gold thickness at the low end of the specification rather than treating more gold as better. Our notes on immersion tin surface finish make the same point about process control mattering more than nominal thickness.

OSP: Coverage, Storage and Handling

Organic solderability preservative is a thin organic film that protects copper from oxidation during storage and is consumed by the flux during soldering. It is inexpensive, gives a flat surface and leaves no metallic residue, which is why it dominates high volume consumer boards. Its weakness is that the protection is finite and sensitive to handling.

Every touch leaves a mark on an OSP surface, and repeated reflow cycles degrade it. Boards finished with OSP should be kept in their sealed packaging until assembly, handled with gloves, and assembled promptly. Where a second reflow is unavoidable, confirm that the coating still performs, because a finish that has been consumed will not protect the copper again. The comparison of finish options in our PCB material requirements notes covers how the choice interacts with laminate and copper weight.

Contamination Between Process Steps

Mixing finishes in one panel creates opportunities for cross contamination. Plating solution trapped under a mask edge can bleed onto a region that was already finished, and rinse water carrying nickel ions can leave residues on areas that will receive an organic coating. These effects are subtle and often appear only as a marginal solderability problem in production.

Rinsing, drying and mask design all need attention when two finishes share a panel. Keep the boundary away from fine features, provide enough clearance for the mask to seal reliably, and confirm with the fabricator which sequence they intend to use before the design is released. A short conversation here avoids a long investigation later.

Documentation and Inspection

The fabrication drawing must state which finish applies to which region, with a clear boundary that a human can follow. Referring to a Gerber layer is usually the most reliable method, because it removes ambiguity about pad numbers and reference designators. Adding a note about the intended thickness range for each finish prevents the fabricator from assuming a default.

Inspection then has to verify two finishes rather than one. Gold thickness is measured by X-ray fluorescence at several points, and the organic coating is checked by a solderability test or by a colour and wetting check. Our overview of PCB quality control describes how those checks fit into a receiving inspection plan, and why the record matters when a batch later shows a soldering problem.

Cost and Process Trade-offs

Selective finishing adds steps: extra masking, an extra plating sequence, extra rinsing and an extra inspection. On a low volume board with a genuine connector requirement, that cost is small compared with a redesign. On a high volume consumer board with no real need for two finishes, it is pure overhead and should be avoided.

The decision should be driven by function. Ask whether any region genuinely needs wear resistance, wire bondability or a very flat surface. If the answer is no, one finish is enough, and a well controlled single finish will always be more predictable than a selective process that exists only to satisfy a habit.

FAQ

Can ENIG and OSP be applied to the same panel? Yes, and it is a common combination. The ENIG regions are plated first, the plating resist is stripped, and the organic coating is applied to the remaining copper. The two processes must be sequenced so that the later bath does not attack the earlier finish.

How thick should the gold be in a selective finish? For solderable pads, a thin immersion gold layer of around 0.05 to 0.1 micrometres over nickel is typical, which is enough to protect the nickel without embrittling the joint. Hard gold for a connector is much thicker and is applied by electroplating over a nickel barrier, which is a different process entirely.

Does selective finishing affect impedance or signal performance? The finish is a very thin surface layer, so it has little effect on impedance. What can affect performance is the nickel layer under gold, because nickel is magnetic and adds loss at high frequency. Where loss matters, the thickness of that barrier deserves a check against the frequency of interest.