Gold Finger: Design Rules and Process Limits

Gold fingers are the part of a board a customer touches, plugs and unplugs, and they are expected to survive thousands of mating cycles without losing contact resistance. They are also one of the hardest features to plate well, because the geometry is thin, the tolerance is tight and the finish has to be both hard and clean. This article explains what drives gold finger quality and how to specify a finish that will pass the first insertion test and the thousandth.

What a Gold Finger Finish Has to Do

The finish on an edge connector serves two purposes: it must conduct reliably at low contact resistance, and it must resist wear as the contact wipes across the surface during mating. Nickel provides the hardness and the diffusion barrier, while gold provides the low and stable contact resistance.

Both metals must be present in the right thickness and with the right structure. A finish that passes a visual check can still fail a wear test when either layer sits outside its specification. A thin or porous gold layer lets the contact penetrate to the nickel beneath it, and nickel oxide then raises resistance in a way that no amount of cleaning will fix.

Nickel Underplate and Diffusion Control

Nickel forms the mechanical backbone of the finish, and it is usually deposited over plated copper that has already been verified for thickness. Typical specifications call for two to five micrometres of nickel, plated from a sulphamate or similar bath that produces a ductile, low stress deposit. The layer also stops copper from diffusing into the gold, which would otherwise raise resistance over time.

Nickel quality is judged by appearance and adhesion rather than by thickness alone. A dull, stressed nickel layer can crack under the bending that connector edges often see, and the crack then appears as a blister or a lifted finger after assembly.

Plating adhesion is verified with a bend or tape test on a coupon from the same rack, because the failure usually appears at the finger root rather than on the flat surface.

Gold Thickness and Wear Life

Gold thickness is chosen from the number of mating cycles expected. A few insertions may need only a flash, while products specified for hundreds or thousands of cycles typically call for hard gold in the range of 0.5 to 1.5 micrometres over the nickel, with heavier deposits where the wipe length is short and the contact pressure is high.

Thickness alone does not guarantee wear life. Hard gold contains a small amount of cobalt or nickel that raises hardness and reduces the adhesive wear that removes material during sliding. Soft pure gold gives lower initial resistance but wears through much faster.

Hard Gold Versus Soft Gold

Hard gold is the standard choice for edge connectors that will be mated repeatedly. Soft gold, typically a purer deposit, suits wire bonding and some low force contact systems where hardness would work against the application. Service temperature also matters, since heat accelerates the diffusion that the nickel layer exists to block. Confusing the two leads to premature wear or to a bond that does not form.

Because the two finishes look similar, the specification has to state the type, the alloying addition and the thickness rather than simply saying gold. A supplier quoting gold without qualification should be asked what deposit is actually being applied. Where a finish is specified by reference to a standard, request the exact class and thickness range rather than accepting the standard name alone.

Chamfer, Bevel and Finger Geometry

The board edge is usually chamfered so the connector guides the card into place without catching. The chamfer angle and the depth of the bevel determine how smoothly insertion occurs, and they must leave enough gold on the finger to contact the spring once the card is seated.

Finger width, spacing and length come from the connector datasheet. Deviations that look small on the drawing can put a finger partly outside the contact area, which reduces the current path and increases local heating. Check finger geometry against the connector datasheet before releasing the artwork.

Plating Resist and Masking

Gold must not reach the rest of the board. The areas to be plated are exposed through resist or plating tape, and everything else is masked. Poor masking produces stray gold on adjacent copper, which cannot be removed afterwards without damaging the surface and often forces a panel to be scrapped.

The transition between the finger and the board, sometimes called the gold finger tail, should be defined on the drawing. Controlling where plating stops prevents the ragged edge that later causes solderability problems on the adjoining pad. Review the plating outline with the fabricator, because panel tooling often sets how clean that transition can be.

Common Plating Defects and Their Causes

The defects that appear most often are nodular growth, porosity, discoloured deposits and edge build-up where the finger meets the resist. Nodules usually trace to bath contamination or to current density at the finger tips, while discolouration suggests the bath chemistry or the current waveform has drifted.

Because fingers are plated in a narrow window, small changes in rack position or contact resistance produce visible results. Daily hull cell checks and regular visual inspection of first-off panels catch those changes before they become a shipment. Keep a reference photograph of an acceptable finger so the comparison is objective.

Inspection and Test Methods

Inspection combines visual examination at magnification, thickness measurement by X-ray fluorescence, and adhesion checks on a coupon. Record the measurement location on each finger, because thickness varies between the tip and the root. Contact resistance is normally verified on the finished assembly rather than on the bare board, but a micro-ohm measurement on a sample can catch a plating problem earlier.

Wear testing, where a sample is cycled against a representative contact, gives the most meaningful measure of life. It is expensive, so it is usually run for qualification and after any change in the plating process rather than on production lots. Retain the cycles-to-failure number so later process changes can be compared against a baseline.

Specifying the Finish on the Drawing

The drawing should state the nickel thickness, the gold type and thickness, the chamfer dimensions, the plated length on each finger and the tolerance on finger position. It should also identify the reference standard used for acceptance, so that inspection has an unambiguous basis.

gopcb reviews edge connector requirements at the quotation stage, because finger geometry interacts with panel layout, routing and plating racks. Settling those details before the artwork is released avoids a costly re-spin when the first insertion test fails. A short review at that stage costs far less than replating an entire lot.

FAQ

How much gold thickness do I need on a gold finger? It depends on mating cycles and contact pressure. Common production specifications use 0.5 to 1.5 micrometres of hard gold over nickel, with heavier deposits for frequent insertion.

Why does contact resistance rise even though the gold looks intact? Porosity or a thin deposit can let the contact reach the nickel beneath. Nickel oxide then builds up on the exposed area and raises resistance, which is why thickness and porosity both matter.

Can gold fingers be reworked? Only by stripping and replating the entire edge, and even then the result is usually inferior. Preventing stray plating and plating defects is far cheaper than repairing them.