Contact Design Considerations For Long-Term Reliability
In the latest installment in our Connector Basics series, Ed Bock of APEX Electrical Interconnection Consultants looks at the ability of a contact system to operate in spite of external contamination and contact design considerations for long-term reliability.
For those critical electrical contact applications where the consequences of failure are severe and long-term reliability is paramount, reliance is customarily placed on noble metal finishes. Gold is by far the noble-metal finish of choice; at the proper thickness and with an appropriate underplate, gold will form little or no tarnish film. A key issue with respect to the ability of a contact system to operate successfully in spite of external contamination factors is the presence/entrapment of non-conducting material at the contact interface.
Generally, the contamination will be either a chemical corrosion film or non-conducting particulates trapped at the interface. With the contact finish of choice – gold – the corrosion film would be an oxide or perhaps a tarnish film that involves chlorine or sulfur by-products and a reaction with base/underplate material that somehow migrated to the surface (perhaps as a result of metallic diffusion through the gold plating or through pores). The gold plate should be of a high quality (pore-free) and of such a thickness that base/underplate metal diffusion to the surface will not occur. For tarnish films that may have developed, the most effective countermeasure to ensure acceptable contact behavior is to require that contact forces be in a range that would aid in the mechanical disruption of these films. Experience has shown that successful long term performance of a pore-free gold deposit with a thickness on the order of 30 micro-inches over 50 micro-inches of nickel is obtained with a normal force on the order of 1N. Clearly a gold “flashed” contact system should not be selected for any application that requires long-term reliability, and would be discouraged for applications that involve anything other than a relatively benign environment.
A second type of contamination to consider is particulates including dirt/dust, various fibrous material, and mold flash. The relative size of these particulates may have a wide range, varying from being nearly invisible to the naked eye to fractions of an inch. Holm (R. Holm, Electric Contacts, Springer-Verlag, New York, 1967) discusses the possible effect of dust, essentially invisible to the naked eye, preventing electric contact. Examples are given in which a single dust particle approximately two microns in size was able to lead to an open circuit with relatively low values of normal force. As expected, as normal force increased, the trend was improved performance; with normal forces on the order of 150 grams, the number of open circuits for a single contact with dust in the one-to-10 micron range is extremely low. However, for large-sized contamination (mold flash, for example), the possibility does exist that opens will be produced even with extremely high values of normal force.
As a countermeasure for environments in which particulates are of great concern, the idea of utilizing a twin-contact configuration has been recommended. In that case, electrical continuity will be established unless both contact locations have particulates trapped at their interface; in essence, what you have is a condition in which contacts are in parallel. We will use the term “bifurcation” to describe a contact in which the contact beam is split longitudinally in order to provide this desirable condition of a “twin” contact.
A bifurcated contact will increase reliability in the presence of particulate contamination. The degree of increased reliability will depend upon the amount of mechanical independence of the contact spots. With complete independence, the failure rate will vary as the square of the single contact rate. If it is known that the failure rate of the single contact is one in a thousand, then the failure rate for the twin version would be expected to be on the order of one in one million. (1,000^2). However, it has been shown that generally the true improvement is less than predicted. Experience shows improvement to be on the order of a decrease of 100 to 1,000 times in the number of failures.
The reason that improvement will be less than theoretically predicted for the bifurcation case is that the individual contact points are generally not mechanically independent of each other. A particle lodged between one pair could tend to reduce the effectiveness of the other. The following figures illustrate a range of independence for some examples of bifurcation and a comparison of a contact offered in both a bifurcated and non-bifurcated version.
The normal force on each of the twin points for a bifurcated contact will ordinarily be about half that of the non-bifurcated configuration. While the reliability for particulate contamination will be enhanced, the ability to disrupt films could be compromised. It is recommended that the normal force associated with each individual contact spot be within the range suggested for the particular finish. If the recommended force is, as an example, 150 grams, then best practice would recommend that each of the twin contacts apply a force of 150gm.
While this discussion, and the limited number of illustrated examples, is more directed toward card-edge contacts, the same reasoning is applicable to other contact applications and geometries utilizing a multi-beam approach.
These recommendations may be somewhat restrictive, but keep in mind that the goal in the present case is to recommend parameters that will yield a far-reaching increase in reliability. Other factors to be considered would be wipe and back-wipe during mating, and specific contact geometries; shielding effects introduced by housing considerations would also be an important part of the overall connector design strategy.
Ed Bock has more than 44 years experience in the connector industry (33 with AMP) in the areas of contact physics, electrical contact phenomenon, fretting corrosion, contact lubrication, and wear studies (tribology). His experience includes the areas of electric contact behavior and contact performance at both signal and power levels while subjected to various environmental conditions. He has studied the performance of various contact finishes and their respective failure modes and mechanisms and recommended countermeasures for improved reliability. He is recognized as a prime researcher in the area of fretting corrosion, the principal failure mechanism for non-noble contact finishes (tin and tin-lead).