Gold Finger: Design Rules and Process Limits

The gold fingers on a plug in card are a row of plated contacts along the board edge, and the chamfer at the leading corner is the feature that guides the board into its socket. The chamfer and the contacts are close together, and the geometry that makes insertion easy is also the geometry that can damage the contact area if it is specified carelessly.

This article covers what the bevel does, how its angle and depth are specified, how it is kept clear of the contact, and how the contacts themselves are plated.

What The Bevel Does

The bevel is a cut or a milled chamfer on the leading edge of the board, at the corner where the board first enters the socket. It reduces the effective thickness of the leading edge and produces a ramp that lifts the socket contacts as the board advances. Without it, the square edge of the board meets the contacts at a right angle and either the contacts or the board edge is damaged, depending on which is harder.

The bevel also centres the board. Two chamfers, one on each face, produce a wedge that guides the board into the slot even when it is offered at a slight angle, which is what makes a card cage usable by an operator rather than by a machine. The depth of the chamfer is measured along the board face and the angle is measured from the board surface, and both are quoted on the drawing because a machined feature at the edge of a laminated board is difficult to inspect after the fact.

How The Bevel Is Specified

A typical specification gives an angle of between twenty and forty five degrees and a depth that produces a leading edge a fraction of the board thickness. The angle controls how gradually the contacts are lifted: a shallow angle lifts them gently and over a longer travel, and a steep angle lifts them quickly with a sharper edge. Both can work, but the shallow angle is easier on the socket contacts and requires more material to be removed.

The depth has to be enough that the leading edge enters the socket without interference and not so much that the chamfer reaches into the contact area or reduces the material around the outermost contact. The two are related: a shallow angle needs a greater depth to reach the same edge thickness, so a design that needs a small depth must accept a steeper angle. The board edge itself is defined with the rest of the outline requirements under board outline and mounting design.

Keeping Clear Of The Contact Area

The chamfer must finish before the first contact begins. Where it overlaps, the contact at the end of the row sits on a sloped or reduced surface, so it does not make full contact with the socket and it wears faster. In a design where the fingers run to the very end of the board edge, the chamfer has to be made shallower, or the first finger has to be pulled back, and both options are cheaper to arrange at layout than after the panel is made.

A second constraint is the position of the fingers relative to the board edge on the other axis. The contacts are usually set back from the two long edges by a small margin so that the socket body does not touch them, and the chamfer has to respect that margin as well. Where the fingers are close to a corner, the chamfer can cut into the plating of the outermost finger, and the copper that is exposed at the cut edge is a corrosion site. The routing space and the edge requirements that bound these decisions are covered under design guidelines for manufacturability.

Plating Of The Contacts

The contact surface is normally a hard gold over a nickel barrier, and the reason for each layer is worth keeping in mind when the bevel is specified. The nickel provides a hard, corrosion resistant base that stops the copper from diffusing into the gold; the gold provides a low resistance, non oxidising contact surface. Hard gold, deposited with a small amount of cobalt or nickel in the bath, is used rather than pure soft gold because it resists the abrasive wear of repeated insertion.

The thickness of the gold sets the number of insertion cycles the contact survives, and the thickness of the nickel sets the barrier quality. Plating defects in this area are more serious than elsewhere on the board, because the contact is a separable interface that is expected to be mated and unmated many times. The failure mechanisms of the deposit itself are described under copper plating defects prevention, and the same care applies to the edge plating that carries the plating current to the fingers during deposition.

Insertion Force And Socket Wear

Insertion force is the sum of the forces from all the socket contacts as they ride over the board edge and settle onto the fingers. The chamfer reduces the initial force, the finger thickness and the socket spring set the steady state force, and the number of contacts multiplies both. A board that is at the top of its thickness tolerance and has a shallow chamfer can require enough force that the operator damages the board or the socket while inserting it.

Where the insertion force is critical, as it is for a high pin count connector, the thickness of the board at the finger area is specified with a tighter tolerance than the rest of the board, and the chamfer depth is set from that tolerance rather than from the nominal thickness. Fingers that are plated over several insertion cycles also wear, and the wear appears as gold removed from the contact and nickel exposed, which raises the contact resistance. The number of cycles the design allows is stated in the product specification and it depends directly on the gold thickness.

Inspection And Common Faults

Inspection of the bevel looks for the angle, the depth and the surface finish of the cut. A routed chamfer leaves tool marks that are acceptable if they do not intrude on the contacts; a cut that is too deep shows as a wide bright band on the board edge; a cut that is too shallow shows as a square corner with only a small flat. The gold finger area is inspected for edge nicks, for plating voids and for any copper exposed by the chamfer.

The most common fault in this area is a chamfer that has cut into the outermost finger, which is both a cosmetic and a corrosion problem, and it usually results from a layout where the finger was placed too close to the corner. The second is a bevel that is specified only as a note without a dimension, which leaves the fabricator to choose an angle. Stating the angle and the depth as dimensions on the drawing removes the ambiguity and makes the feature measurable, which is what allows a supplier to hold it consistently.

FAQ

Is the chamfer cut before or after plating? It is normally cut before the final plating so that the cut edge is not left with exposed copper, and the sequence is set by the fabricator. Where the bevel is cut after plating, the drawing should require the exposed copper to be protected or re-plated.

Does the chamfer have to be on both faces? For a card that can be inserted either way up, yes. Where the board can only be inserted one way, a single chamfer on the leading face is sufficient and is often used to key the board so that it cannot be inserted incorrectly.

How many insertion cycles should the fingers survive? It is set by the product specification and by the socket, and it is achieved by the gold thickness together with the contact force. A specification of a few dozen cycles is common for a serviceable card, while a fixed connection with no expected mating is treated differently.