Hard Gold Plating for Edge Connectors: A Process Guide

An edge connector is the part of the board that will be inserted and withdrawn repeatedly, and it is the one place where a surface finish has to survive mechanical wear rather than only a soldering operation. Hard gold plating is the answer, and it is a different process from the thin gold used on a bonding pad, with its own underlayer, its own thickness requirement and its own set of failure modes. Hard gold is specified by thickness and by hardness, and both have to be measured rather than inferred from the colour of the deposit.

Why Edge Connectors Need Hard Gold

The contact fingers of an edge connector slide against a mating socket every time the board is inserted, and each insertion removes a small amount of material from both surfaces. A soft finish wears through quickly and exposes the layer beneath, which then oxidises and raises the contact resistance.

The requirement is therefore both a low and stable contact resistance and a mechanical durability measured in insertion cycles. The two are linked, because the resistance rises sharply once the finish is worn through to the underlying metal. A worn finger can still look gold in a photograph while its contact resistance has already risen by an order of magnitude.

The Nickel Underlayer

Hard gold is always deposited over a nickel barrier. The nickel provides the hardness that stops the gold from being displaced, it slows the diffusion of copper into the gold layer, and it gives the gold something to bond to that will not migrate. Copper diffuses readily through gold at modest temperatures, and without a barrier the surface would become a copper gold alloy with poor contact properties.

The nickel layer also has to be thick enough to act as a barrier and thin enough not to introduce stress. A nickel layer that is too thin allows copper to diffuse to the surface and causes a resistance rise, while one that is too thick can crack and take the gold with it. Our plating thickness guide covers how both layers are measured. Measuring the nickel layer after the gold has been applied is difficult, so the thickness is normally confirmed on a coupon or on a masked witness area.

Gold Purity and Alloy Additions

Pure gold is soft, so the plating bath contains a small amount of a hardening element, usually cobalt or nickel, that is codeposited with the gold. The result is a deposit with a hardness several times that of pure gold and a much better wear resistance.

The addition level affects the colour and the contact resistance as well as the hardness. A deposit with too much hardener is brittle and can crack during insertion, while one with too little wears quickly, so the bath is analysed and the deposit is checked rather than assumed. The hardener content is measured in the bath and the result is confirmed by testing the hardness or the wear resistance of a sample.

Thickness Requirements

The thickness requirement depends on the number of insertion cycles the product is expected to see. A board that is inserted once during assembly needs far less gold than one that is plugged and unplugged daily, and the specification should state the cycle count as well as the thickness. A drawing that quotes a thickness without a cycle count leaves the supplier guessing which class of product is being ordered.

Thickness is usually quoted as a minimum over the finger area rather than an average, because the lowest point is what fails first. Measurement is normally by X-ray fluorescence on the finger, supported by a microsection for qualification and for failure analysis. A thickness that is measured at one point on one finger says very little about the rest of the board, so the sampling plan matters.

Selective Plating Methods

Gold is expensive, so it is applied only where it is needed. The common methods are to plate through a resist window, to use a brush or a jet that deposits metal on a defined stripe, or to plate the whole panel and then etch the unwanted gold away.

Each method has a different edge definition. A resist window gives a sharp edge but requires the resist to survive the plating bath, while a brush or jet is faster and produces a slightly softer edge. The choice affects the layout, because the finger area has to accommodate the tolerance of the method. The layout should also include a witness area where the thickness can be measured without destroying a product finger.

Masking and Resist Design

The resist that defines the window has to be thick enough to prevent plating under it and adherent enough to survive an alkaline or acidic bath. A resist that lifts at the edge allows the plating solution to creep and produces a gold finger that is longer than the drawing. Where the gold must stop at a defined line, the resist edge definition is what decides whether the finished part is acceptable.

The finger length and the spacing between fingers are also process dependent. Fingers that are too close together make the resist between them unstable, and fingers that are too long increase the current density variation along the strip, which shows up as a thickness gradient. A plating thief or a dummy finger at the end of the strip evens out the current and reduces the gradient along the connector.

Adhesion and Porosity

Porosity is the enemy of a gold contact, because each pore is a path for the base metal to corrode through to the surface. Pores come from a contaminated bath, from hydrogen evolution during plating and from a nickel layer with an uneven surface. Filtration, agitation and a controlled current density all reduce porosity, and the bath should be carbon treated on a schedule.

Adhesion has to survive both insertion and any subsequent thermal cycle. A gold layer plated onto an oxidised nickel surface may look perfect and lift during connector insertion, which is why the time between the nickel and the gold steps is controlled. A short transfer time between the two tanks, with an effective rinse, keeps the nickel surface active enough for a strong bond.

Inspection and Test

Inspection includes the thickness measurement, a visual check for the position and length of the gold area, and a check for plating defects at the finger edge. A porosity test is used on qualification samples and after any bath change.

Where the connector is critical, an insertion cycle test on a sample confirms that the finish meets the durability requirement rather than only the thickness requirement. Our quality documentation describes how these results are classified at gopcb.

Process Control Points

The controls are the nickel bath chemistry and thickness, the gold bath analysis and hardener content, the current density, the resist adhesion, the rinse quality between the two plating steps and the thickness result on the finished finger. Each of those is recorded, and the records make it possible to explain a thickness trend rather than argue about it.

Because the gold is expensive, the bath is also a target for theft and for contamination, so it is normally kept in a controlled area with restricted access. Our fabrication notes guide lists the data that should accompany an order that includes plated edge fingers. Specifying the gold thickness, the hardener type and the insertion cycle requirement at the quotation stage avoids a redesign later.

FAQ

Why is hard gold used instead of pure gold? Pure gold is too soft and wears through quickly on a sliding contact. A small amount of cobalt or nickel in the deposit raises the hardness and dramatically improves the number of insertion cycles.

How thick should the gold be on an edge connector? It depends on the insertion cycles the product will see, and the specification should state both. The thickness is quoted as a minimum anywhere on the finger, not as an average.

What causes a gold finger to corrode? Porosity in the deposit lets the nickel or copper beneath corrode through to the surface. Bath contamination, hydrogen evolution and a rough nickel layer are the usual causes.