Ask Dr. Bob

Devil or Angel? Connector Testing
By Max Peel, Senior Fellow, Contech Research

Connector testing is generally viewed as a nuisance type of activity. So let’s try to set the record straight. The basic purpose of testing is to demonstrate the capability of a connector to meet the specific design characteristics called for and/or to determine the susceptibility, or lack there of, to time dependent failure mechanisms. In the outset of the previous article, it indicated a basic generic test program with five basic groups and one supplemental. This can translate into five basic type test programs as follows:

• Developmental Testing

• Verification Testing

• Qualification Testing

• Performance Testing

• Reliability Testing

• Development Testing

This involves first testing the article or prototype product. It generally targets specific design issues to assure that catastrophic conditions do not exist and that the product had performed in accordance to the design goals set. It normally consists of a series of individual test or test environments. It is performed during the product design phase, which may or may not result in design changes. The tests are designed to evaluate the following, but are not restricted to: 

1. Electrical Resistance

2. Compatibility of the Contact System to:
   - Mechanical Forces
   - Normal Forces
   - Wear Characteristics

3. Compatibility of Plastic Housings to:
   - Solvent Resistance
   - Resistance to Solder Heat
   - Moisture Absorption Factors
   - Mechanical Integrity

4. Impact of the environments being considered

In many instances basic environmental tests are performed with short durations. As indicated, this type of testing is to ensure that the design is progressing towards its projected objectives. There also are situations where long durations are used with periodic variable measurements, performed to determine critical electrical stability. It should be a part of the design phase, but is apparently diminishing in its use.

Verification Testing
This simply refers to making sure that a product meets its advertised characteristics. It does not include environments. It is performed by users and manufacturers to simply confirm that the variable requirements are in accordance with advertised values. This type of testing is almost non-existent.

Qualification Testing
This is one of the most popular test programs. It is a number of test groups following the generic sequences, as shown by Dr. Bob in the first article of the series. The programs may be established by manufacturers of connectors, proprietary user programs, or by special interest groups, such as the military, USB, or US CAR.

Although these plans may look the same, the severity levels, environmental considerations, etc. may be significantly different while concentrating on a specific area of interest.

Test durations are short, and very rarely go beyond moderate levels of test. There are two types of testing—success type and application-specific type. The success testing is designed to assure the connector will pass, it will have mild- to low-level environmental short duration and broad requirements. Application specific emphasizes those characteristics critical to a specific application, which tend to address the specific need of an industry or a specific user concern. The requirements tend to be tighter, with environments at a higher level reflecting slightly above operational conditions, and severity/test durations are extended. Both types cover a wide range. Success testing is gradually disappearing with the application of specific programs increasing in use.

Performance Testing
Performance testing is characterized by long-term environments with significant severity levels. They tend to concentrate on resistance stability with “tight” requirements. The environments are more long term in nature in order to allow “mother nature” to do it’s thing in determining if the connector is susceptible to time dependent failure mechanisms, such as corrosion, fretting corrosion, etc.

The GR-1217-Core and EIA 540BOAE-2000 are typical of this type of testing. Said specifications are generally developed by user organizations, either by themselves or by special interest groups, and usually sample sizes are moderate, with moderate to high-variable sampling required (i.e. 200-600 data points/group).

Reliability Testing
This type of testing is intended to establish meantime between failures. It involves a high number of samples, large number of variable measurements, long durations (up to 4,000 hours) with test environments used relative to specific long-term failure mechanisms. It is lengthy to perform (half to two and a half years) and costly (>$100,000). The basic problem is when testing has been completed, the MTBF that is established is for the specific lot of samples tested and may not account for the full tolerance range of the products involved. Due to the rapid obsolescence rate and new product introductions, this type of test has been replaced by performance test programs.

Having briefly outlined the various programs that have been used, there has been a shift in program emphasis. Developmental testing has significantly decreased in usage. The time to market has significantly decreased to a point that minimal testing is performed, with everyone hoping “it’s enough.”

Verification testing is pretty much eliminated.

Reliability testing is not used due to economics, the time involved, and the uncertainty of the results.

Thus the most common programs are the basic qualification test programs (application specific) or performance testing, with performance testing becoming the more prominent type program.

This type of test is being imposed by various users and by special interest groups. The basic intent is to have the product demonstrate its capability to perform and it’s immunity to long-term failure mechanisms.  This has become a necessity since material systems have been in a state of flux due to the increased use of gold flash (or gold colored) contacts, the lack of information from the manufacturers, among other things—especially with offshore manufacturing becoming more and more popular (where process controls are paper only). There is an old adage that I’m concerned many people have forgotten. If one buys a “cheap” product, one will get exactly that, a cheap product.

Future articles will address testing more in depth, the myths, things to do or not to do, things which are being done to beat the system, as well as the tests that are available to determine the life of a connector.

Testing is not a nuisance, but a necessary tool in the development of a product, to find a product’s weakness and susceptibility to failure mechanisms. In other words, testing uncovers what the product can do and what it cannot do.

Think of it this way: You can sit in a nice comfortable chair, smoking a $10 cigar (Maduro’s preferred), sipping on a glass of $60 Scotch while contemplating a fistful of dollars. If you decide to forego testing, the dollars can pay for repairs or returns. If you decide to test in advance of selling your product, those dollars can go to the bank. Your choice.

Send your connector testing questions to AskDrBob@ConnectorSupplier.com.

Max Peel is a Senior Fellow at Contech Research, an independent test and research lab located in Attleboro, Massachusetts, U.S.A. For more information, visit www.contechresearch.com.