Edge Plating And Hard Gold Wear Resistance
A gold contact surface is not simply a layer of gold. It is an electroplated deposit from a bath that contains cobalt or nickel as a hardening addition, and the resulting layer is harder and more wear resistant than pure gold. The distinction matters because the failure mode of a connector surface is wear rather than corrosion, and a soft gold layer rubs away after a few insertion cycles and leaves the nickel underneath exposed.
This article explains what hard gold is, how the surface wears, how the thickness is chosen, and how edge plating differs from plating a connector finger.
What Hard Gold Is
Pure gold is soft and ductile, and it deposits as a relatively soft layer that is easy to solder and easy to wear through. A hard gold deposit contains a small percentage of a hardening metal, typically cobalt or nickel, which is codeposited from the bath. The hardness rises from around 40 to 60 on the Vickers scale for pure gold to 130 to 200 for a hard deposit, and the wear resistance improves in proportion.
The trade is a loss of ductility and a slightly higher contact resistance, and a bath that is not controlled can produce a deposit with an internal stress that cracks when it is bent. The gold is also porous when it is thin, and the porosity is what allows the underlying nickel to corrode. For a contact surface the porosity has to be low, which means the deposit has to be thick enough for the application and the substrate has to be clean.
How A Contact Surface Wears
Two mechanisms act at once. Adhesive wear occurs when the two surfaces are pressed together and asperities weld and tear, which transfers material from one surface to the other and leaves a roughened area. Abrasive wear occurs when a hard particle, often debris or a fragment of a coating, is dragged across the surface and cuts a groove. A connector that is inserted repeatedly sees both, at the point where the contact first touches and at the point where it finally rests.
The environment changes the balance. A surface that is exposed to sulphur or to chlorine compounds forms a film that is removed by the wiping action of insertion and then reforms, and the repeated removal consumes gold. A surface that is cleaned with the wrong solvent leaves a residue that holds moisture against the metal. The design responses are the same: more gold, a harder gold, and a contact geometry that keeps the wear away from the final resting point.
Thickness, Cycles And Substrate
The thickness is chosen from the number of insertion cycles the product will see. A board that is inserted once in the factory needs only a thin layer, often a fraction of a micron, while a card that is inserted daily for years needs several microns. Typical figures are 0.5 microns for a few insertions, 1.3 microns for a card edge used intermittently and 2.5 microns or more for a connector that is mated frequently.
Under the gold is a nickel barrier, typically 2.5 to 5 microns, and it does two jobs. It stops the copper from diffusing into the gold, which would raise the resistance and reduce the corrosion resistance, and it provides a hard base that supports the thin gold layer. A gold layer over a poorly plated nickel will fail at the same thickness that would be adequate over a good one, because the substrate deforms under the contact load and cracks the gold.
Edge Plating And Finger Plating
Plating a connector finger is done as part of the surface finish or as a separate selective plating operation, and the layers are deposited on a flat surface that is defined by the mask. The process is well understood and the thickness can be measured directly with an X-ray gauge on the finger.
Edge plating is different because the plated surface is the cut edge of the board. The edge is routed to a depth before plating and the last cut is made afterwards, so the plating has to adhere to a laminate wall that has been cut and to a copper layer that has been exposed by the routing. The adhesion at the edge is the weak point, and the plating thickness on a vertical wall is measured on a coupon rather than on the board. The mechanical arrangement around the edge is described under board outline and mounting design.
Specification
The drawing should state the gold thickness, the hardness or the type of the bath, the nickel thickness and the area that is to be plated. It should also state the number of insertion cycles the figure is intended to support, since a thickness without that context cannot be judged and a supplier will meet the letter of the requirement with a finish that fails in the application.
The contact area itself belongs to the mechanical design. The gold should cover the whole range of travel of the contact, including the point at which the board first touches, so the fingers are longer than the rest position requires. The pad dimensions that the contact rests on are described under PCB pad design standards, and the fabrication sequence that produces the plated surface is described under PCB design and fabrication.
Verification
The thickness is measured with an X-ray fluorescence gauge on the finger or on a coupon, and the measurement is made at several points because the current density varies across the panel. The porosity is assessed by a test that exposes the surface to a corrosive medium and counts the pores, and it is used where the application demands a low porosity finish.
The functional verification is a wear test. A sample connector is mated and unmated for the required number of cycles and the contact resistance is measured at intervals. The resistance rises slowly as the gold wears and sharply once the nickel is exposed, so the shape of the curve over the cycles is the evidence that the thickness and the hardness were adequate. The test is performed when the finish or the supplier changes rather than on every lot.
Contact Force And Geometry
The wear rate depends on the contact force as much as on the gold. A higher force increases the real area of contact and lowers the resistance, and it also accelerates the wear and the fretting that follows. The connector designer balances the two, and the board designer has to respect the result: a board thickness at the thick end of the tolerance raises the contact force, and one at the thin end lowers it and may produce an unstable contact.
Fretting is the failure that follows a low contact force in a vibrating environment. The two surfaces move by a few microns against each other, the gold wears locally and the debris oxidises, and the contact resistance rises until the circuit fails intermittently. The remedy is a higher contact force or a lubricant at the contact, neither of which is available after the board has been designed, which is why the thickness tolerance and the surface finish have to be held.
FAQ
Is hard gold as good as soft gold for soldering? No. The hardening addition makes the deposit harder to solder than pure gold, and a surface that has to be soldered rather than mated is normally finished with a thin soft gold or with another finish entirely.
Can the gold be too thick? It can. A very thick deposit is more expensive, it can be more stressed and it holds more of the bath chemistry in its grain boundaries, which can lead to corrosion in a humid environment.
How is the wear point controlled? By making the finger long enough that the first contact and the final rest position are both inside the plated area and separated from each other, so that the region where the contact sits when the product is in service is not the region that is wiped during insertion.
