 |
For Connector Specifier’s 20th anniversary, we asked industry veterans with many years of involvement in the market to review the past two decades of connector technology.
By Max Peel
A 20th anniversary is an opportunity to look back at the connector industry in the early 80s and compare the past to the present, with respect to the changes that have taken place within the connector industry. Such a retrospective helps to answer a question that politicians like to ask: “Are we better off now than we were twenty years ago?” As always in these types of comparisons, we have the good, the bad, and the ugly.
The good
Connector concepts and designs have significantly changed. In the early 80s, connectors were on 0.100” centers with two or three rows and up to 200 contact positions. Anything less than 0.100” centers was considered state of the art with most designed using the English grid patterns.
 Max Peel (in 1976, left, and 2001, right) is still known for his ever-present cigar-“not the same one,” says Peel. |
Today, connector designs are routinely on 0.050” centers with metric grid patterns on 0.5-mm centers and up to eight rows. Connector densities have increased up to 1500 positions with the possibility of going up to 5000 positions in the future. This has been achieved with the development of some important new features, such as the evolution of surface-mount-termination techniques, the development of the “small” compliant-pin design concepts, allowing the use of low normal-force systems (as low as 25 grams) with multiple beam designs, and significant improvement in manufacturing techniques.
Twenty years ago, four basic terminations existed—crimp, wave soldering, “large compliant pins,” or IDC. The vast majority of contacts used single- or dual-beam concepts with a 50/50 split between screw machine contacts and stamped-and-formed contacts (today it’s more like a 20/80 split, with stamped and formed being predominate).
During the past twenty years, the explosion of new applications has significantly impacted the development of new concepts. In the 80s, the basic integrated circuit dominated. We have moved through the evolution of large-format PGAs to BGAs, to LGAs on smaller centers and significantly higher pin counts.
The generic term “electronic industry” is disappearing. We now have a segmented series of sub-electronic areas, particularly in telecom (central office, wireless, relay stations, and so on), computer (laptop, PC, super computer, and specialized), and automotive (under the hood and passenger compartments). Electronic technology is increasing in importance in other areas also, such as in process control, banking/financial, and medical. This total evolution has diminished the use of general-purpose connectors while developing application-specific designs and concepts. Whereas in the past, connector designs had to meet a standard set of requirements and environments, today each of the electronic segments has their own significant requirements and environments unique to their applications.
As a result of the application explosion involving more aggressive environments, engineers are in the process of developing new evaluation techniques and test environments. In the past, short environmental durations were common, as were benign (sine) vibration and small sample size for variable monitoring. Very few harsh environment and/or thermal conditions were considered.
Today, tests have expanded to longer duration exposures. Increased use of dry thermal cycling is now the norm, along with continuous resistance monitoring, random vibration, “glitch” detection, and use of mixed flowing gas. All of this is integrated with the development of high-density, low-force systems in conjunction with severity levels that differ depending upon the application. At least 100 to 200 variable data points are required. In other words, evaluation techniques have become more exacting and sophisticated.
Improved testing is a very positive movement, resulting in the development of new connector families and should continue as new markets evolve, but along with the good come the bad and the ugly.
The bad
The activity of globalization (a term I dislike) has resulted in significantly increased product manufacturing being moved offshore. Twenty years ago, expansion in these areas was to service local markets. Now it’s to obtain “cheap labor” and reship back to the U.S. for sale. This achieves lower component price, but the quality leaves a lot to be desired. In essence, one gets what one pays for.
Twenty years ago, the emphasis was to design and manufacture high-quality products at the lowest possible price. Today it’s all about price.
As a result of moving to offshore manufacturing, the downsizing of companies, early retirements, and the current economic squeeze, knowledge of basic connector technology issues has continued to significantly diminish. Experienced and knowledgeable engineers (the infamous “gray beards”) are gone with new inexperienced engineers entering the industry. Thus with the lack of understanding basic contact physics, failure mechanisms, and such, the same problems that were solved in the 60s and 70s have returned. The transmission of knowledge of the old to the new does not exist, resulting in “reinventing the wheel.”
Twenty years ago, the market introduction of new products was approximately 18 months. Today, it’s less than 12 months (targeted to nine months). This allows for little or no developmental or performance testing of new concepts. When performed, testing or “proving out the design” is looked upon as a detriment to meeting schedules. Many of the younger engineers are more interested in meeting a schedule as opposed to the quality of performance of the product in question.
And the ugly
During the 70s and up to the early 80s, a great deal of technological and applied research was performed and to a large degree was shared industry wide. Bell Laboratories (now Lucent Technologies’ Bell Labs Innovations) created this emphasis; so did the research departments of key connector manufacturers (mostly now non-existent or greatly diminished). Most of the work performed during that period of time is just as valid today as it was then. Since the early 80s, an iron curtain has descended on this activity. Very little technical (with some rare exceptions) information has been provided or shared. The information shared today is more product-oriented, and tends to say, “Look how great our product is; we cannot share data, but trust us, it works.” The technological information generated in the 70s and 80s is, for the most part, lost, even though it was published. “If it’s not on the Internet, then it simply doesn’t exist,” is the thought process of today’s engineer.
Some examples are tin-whiskering papers from Bell Labs and the Tin Research Institute published in the early 60s, many papers on corrosion issues published in the 70s and early 80s, including fretting corrosion issues, and wear studies done in the late 70s and the 80s.
The decrease in gold-plating thickness (flash) now being used and showing up in tin plating (as low as 30 µ”) bears careful watching. My well-known opinion of gold flash is that it’s a generic finish. (“Peel’s Law” according to Peel is, “Anybody who sells or uses gold flash on a connector gets all the trouble they deserve.”) Thinned-out tin plating is a significant change for the worse. Recent examples indicate discontinuous coverage with no published data to indicate the type of performance one can expect. Twenty years ago, the standard was 30 µ”of gold over nickel and 100 µ” of tin (alloys) over nickel. Data was readily available to confirm performance.
In many recent studies, the thin gold and tin are being plated directly over copper alloy. The nickel underplate is starting to be eliminated. As a result, I may have to even add a second “Peel’s Law”: “Anybody who doesn’t use nickel underplate on a connector gets all the trouble they deserve.”
On balance, the continuing development of new connectors is good. Some of the technological changes as well as the apparent lack of knowledge thereof is bad and ugly. The increasing knowledge of the user and their increasing awareness of the quality and changes occurring are good. However, the one way the connector industry can be hurt is to “shoot themselves in the foot” with low-quality offshore manufacturing, inexperienced personnel, lack of proper testing, and withholding of information.
Neutral
There is one neutral evolution that started to evolve in the late 90s and is continuing at this time. Many user agencies have created special interest groups or associations that have developed product and qualification standards. This is driving product development for the specific application involved (USB is one such example). Many user companies have developed their own performance standards, which address the issues they are concerned about. In other words, users are driving connector development, not the connector manufacturer. Twenty years ago, it was the opposite. Depending upon where one sits, it could be good, bad, or ugly.
MAX PEEL is a long-time editorial advisor to Connector Specifier and senior fellow at Contech Research, 67 Mechanic St., Attleboro, MA 02703. Tel: (508) 226-4800; email: maxconres@aol.com.
