PCB Gold Fingers: Design, Plating, and Cost
An edge connector is the part of a board that carries everything: power, ground, and every high-speed signal that leaves the card. On memory modules, graphics cards, and expansion cards, that interface is a row of gold-plated contacts along the board edge, and its quality decides how many insertion cycles the product survives and whether the contact resistance stays stable.
What PCB gold fingers are
Gold fingers are the gold-plated contact pads formed along the edge of a printed circuit board. They allow the board to be inserted into a matching slot so that the contacts mate directly with the socket, which removes the need for a connector on the card itself.
The gold layer is not cosmetic. It provides corrosion resistance, stable contact resistance, and wear resistance, and it is the reason a card can be inserted and removed many times without the contact surface degrading.
Hard gold and soft gold
Two types of gold finish are used, and confusing them is a common design error. Hard gold is electroplated with alloying elements that raise its hardness, which makes it suitable for repeated insertion. Soft gold is a chemically deposited, high-purity layer used for wire bonding, and it wears quickly, so it is not appropriate for an edge connector.
The distinction is measurable rather than stylistic. A soft gold surface may pass an inspection and still fail after a few hundred cycles, because the property that matters at an edge connector is mechanical durability rather than purity.
An edge connector is a mechanical interface that happens to be electrical. The specification should be written for the number of cycles the product will see.
Where they are used
Memory modules are the most familiar application, where the contact row mates with a memory slot. Graphics cards and other expansion cards use the same approach through a peripheral component interconnect express slot. Solid-state storage, audio cards, and network cards follow the same pattern, and internally, modules are often joined or upgraded through edge connections rather than cables.
That last point explains why the interface survives. A direct edge connection avoids a mating connector pair, which reduces insertion loss, removes a mechanical failure point, and lowers cost, provided the board can be inserted into its slot without damaging the contacts.
How the plating is produced
The process starts at the board edge, where the surface is prepared and cleaned to remove oxide and contamination. A clean surface is the basis of adhesion for everything that follows, and it is the step most often compromised when a plating bath or a cleaning line drifts.
Nickel plating follows, typically three to six micrometres thick. The nickel layer is the barrier that prevents gold from diffusing into copper, which would otherwise destroy the contact over time. The gold layer is then electroplated on top of the nickel, usually between 0.76 and 1.3 micrometres, balancing conductivity and wear resistance against the cost of gold.
After plating, the edge is bevelled, usually at 30 or 45 degrees. The bevel guides the card into the slot without scraping the contacts or the socket, and it removes the sharp edge that would otherwise chip. Inspection completes the process, covering appearance, electrical continuity, and coating adhesion.
Design rules worth fixing early
Finger width and pitch are set by the slot they will mate with. A pitch of about 1.0 mm is common, with finer pitches used in compact systems, and the width has to carry the current the interface requires while still fitting the available edge length.
Length deserves the same attention. A contact must be long enough for stable mating and short enough not to waste gold, which is the most expensive material on the board. The bevel angle is a compromise between mechanical strength and ease of insertion.
Plating thickness is the parameter that decides service life. Hard gold between 30 and 50 microinches supports thousands of insertion cycles, and the choice within that range should follow the expected number of insertions rather than the desire for margin. Because the interface is a mechanical one, it should also be reviewed against the relevant design and acceptability standards, and the dimensional rules for pad geometry are collected in our guide to PCB pad design standards.
Gold is specified in microinches for a reason: the layer is thin enough that a small change in thickness is a large change in cost and in life.
What drives the cost
Gold dominates the calculation. Moving from a 30 microinch layer to 50 microinches can raise the plating cost by 20 to 30 percent, and the difference buys insertion cycles that many products will never use.
Volume comes next, because the plating process has fixed costs that are spread across the order. Sample and small-batch builds therefore carry a much higher unit price than production quantities of the same card.
Board size and connector length set the plated area, and since the cost of gold scales with area, a long edge connector on a large card carries more of the total cost in the plating than in the laminate. As a result, the plating can add a substantial per-board premium, which is worth modelling before the mechanical outline is frozen. The wider plating considerations, including bath chemistry and its control, are discussed in our guide to electroplating additives for PCBs.
Failure modes and how to prevent them
Delamination arises from inadequate surface preparation or poor adhesion, and it is prevented by controlling both. Uneven nickel or gold thickness produces the same outcome, so plating uniformity is a process discipline rather than a cosmetic preference.
Wear is the failure mode that the gold specification addresses directly; a hard gold layer on a properly plated nickel base resists it far better than a soft gold surface. Contamination is subtler, because gold resists corrosion but its surface can still collect material that interferes with contact, which is why antistatic packaging and dry storage matter for a board that may sit in inventory.
Mechanical damage closes the list. A poorly formed bevel, or careless handling during assembly and shipping, can chip the edge of the contacts, so the bevel process and the packaging are part of the product specification. The mechanical decisions that govern how a card sits in its enclosure are covered under board outline and mounting design.
The interface as a system
The plated contacts, the slot they mate with, the retention force of the socket, and the way the card is inserted all belong to the same mechanical design. A gold thickness chosen without reference to the socket, or a bevel angle chosen without reference to the insertion force, produces a specification that passes inspection and fails in the field.
FAQ
How thick should the gold layer be? Between 30 and 50 microinches of hard gold is the industry range for edge connectors. Choose from the number of insertion cycles the product will see.
Can an immersion gold finish be used for the contacts? No. That finish is soft gold, designed for solderability and wire bonding, and it wears quickly under repeated insertion. Hard gold is required at the edge.
Why do memory modules and graphics cards use this construction? Because edge contacts provide a direct, short, and electrically stable path into the slot. Removing the mating connector reduces loss, cost, and mechanical complexity at the same time.
