Fretting Corrosion in Connectors: Plating Wear and Contact Resistance

Fretting corrosion is the damage that occurs when two metal surfaces in contact are rubbed together by a small, repeated motion. In a connector on a printed circuit board the motion is rarely more than a few micrometres and comes from vibration, thermal expansion or the flexing of the board in service. The result is a build-up of oxide debris in the contact area, a rise in contact resistance and, in the worst case, an intermittent open circuit that appears only under vibration.

What Fretting Corrosion Is

Two mechanisms act together. The mechanical action breaks through the surface layers and exposes fresh metal, and the chemical action oxidises that metal as soon as it is exposed. Where the contact metal forms a hard, insulating oxide, as tin does, the debris accumulates in the contact zone and cannot be displaced by the contact force.

Gold behaves differently because its oxide is unstable at these temperatures, so a gold contact does not accumulate oxide debris in the same way. The failure mode for gold is wear through to the underlying nickel or copper, after which the contact behaves like a base metal contact with a much thinner plating.

Where It Appears on a Board

The classic location is a board-to-board or wire-to-board connector that sees vibration, such as those in automotive, industrial and instrument products. Card edge connectors, press-fit pins and spring contacts are all exposed to the same mechanism, and the smaller the contact force, the more readily the debris stays in place.

The second location is where the board itself flexes. A connector mounted near a mounting screw, or a heavy component whose mass generates motion at the solder joint, produces relative movement at the contact. Once the contact pair has been plated and assembled, the only remaining control is the mechanical design that limits the motion.

Connector Plating and Plating Wear

Plating wear is the mechanical half of the problem, and it is driven by the contact force, the radius of the mating surfaces and the number of small movements. A hard plating on a compliant substrate distributes the wear over a larger area and delays the exposure of the underlying metal.

Nickel under gold is the usual construction, with the gold thickness chosen by the number of mating cycles expected. Thin gold of 0.1 to 0.2 micrometres is porous and wears through quickly, while 0.8 micrometres or more survives thousands of cycles. Where the finish is tin, the plating is much thicker but softer, and the oxide that forms is the problem rather than the wear.

Contact Resistance Growth and Measurement

The signature of fretting is a contact resistance that starts low, rises in steps and becomes unstable. The measurement has to be made at a low current that does not break down the oxide film, typically below 100 mA, and it has to be repeated under the motion that causes the damage rather than on a static pair.

Dry circuit measurement is the appropriate technique, because a higher test current can arc through the oxide and give a good reading on a contact that would fail in service. The comparison of interest is the resistance at the start of a test and after the test cycles, together with the variation between cycles.

Mating Cycles and Micro-Motion

Mating cycles are a normal part of the specification, and connectors are rated for a number that reflects the plating and the contact design. A few hundred cycles may be acceptable for a service connector, and a few dozen for a production connector that is mated once. What matters is whether the intended cycles are fewer than the number at which the plating is worn through.

Micro-motion is the more damaging event, because it occurs without anyone noticing. A connector that is mated once and then vibrates for years experiences far more relative movement at the contact than the mating cycle rating suggests. Terminals that are supported against the housing, and housings that are constrained against the board, reduce the amplitude of that movement.

Finish Selection: Tin, Gold and Palladium

Tin is cheap and solderable but forms a hard oxide, so a tin-to-tin contact under micro-motion fails earlier than a gold-to-gold pair. Gold is the most tolerant, and its limitation is wear rather than chemistry. Palladium and palladium-nickel sit between the two, with better wear resistance than gold at a lower cost in some designs.

Mixed pairs are the case worth thinking about. Tin mated with gold transfers tin onto the gold surface through the mechanical action, and the transferred tin then oxidises, so a pair that was expected to behave like gold behaves like tin. Where a tin-plated part must mate with a gold-plated one, the contact force and the number of cycles should both be reviewed.

Assembly and Board-Level Contributors

The assembly process contributes to the problem in two ways. Residue left on the contact surface after soldering or cleaning increases the contact resistance from the start and provides sites where debris collects, which is why cleanliness verification on connector contacts is worth doing on high-reliability products.

The board itself is the second contributor. Flexing under a connector moves the contact pair, and the amount of flexing depends on the layer stack, the copper balance and the mounting arrangement. Where a product is known to see vibration, the board level vibration test is the place to demonstrate that the connector survives it.

Test Methods and Qualification

The standard laboratory method subjects a mated pair to relative motion at a controlled amplitude and frequency with the contacts carrying a dry-circuit current, and it records resistance through the test. Thermal cycling is often combined with the mechanical motion, because the two act together in service and the combination is more damaging than either alone.

Where a connector has a history of the problem, the relevant qualification is a mixed flow, combining thermal cycling with vibration and measuring contact resistance at intervals. A test that only cycles temperature will not reveal a mechanism that needs mechanical motion, and a test that only vibrates will miss the thermal contribution.

Corrective Actions and Records

Once a product is in production the available actions are limited to the contact force, the plating, the lubricant and the mechanical restraint. A connector lubricant that displaces oxygen from the contact zone can extend the life of a tin contact substantially, and it is often the cheapest available improvement.

The records worth keeping are the plating specification, the contact normal force, the mating cycle count in service and any lubricant applied. When a contact resistance failure appears in the field, those four values decide whether the cause was a plating that wore through or a contact that lost its force.

FAQ

Why does a gold-plated connector still fail in the field? Usually because the gold has worn through to the nickel underplate, or because the contact force has dropped. A gold contact removes the oxide problem but is not immune to wear or to loss of force.

Does a lubricant really help? It can, by excluding oxygen from the contact zone and by reducing friction that redistributes debris. It is not a substitute for adequate contact force or for a plating that suits the number of cycles expected.

Should contact resistance be measured with a multimeter? Not if the aim is to look for fretting damage. The measurement should use a dry circuit current so that the oxide film is not broken down during the test, which a normal multimeter may do.