|
To Fail, or Not to Fail, That is the Question.
By
Dr. Robert S. Mroczkowski, Bishop & Associates Inc.
It is
probably not necessary to say that the opinions Max Peel and I express
in our articles for ConnectorSupplier.com are our personal
opinions, but opinions based on our many years of experience in dealing
with the design, materials, testing, and failure analysis of connectors.
I chose to open this article with that statement because the following
opinions are more “theoretical,” and less “experiential” than is the
general case. The topic this time is the determination of a “failure”
criterion. (There will be lots of quotation marks around “problematic”
topics in this article.) I ended my last article on acceleration
factors, with the following:
“When
reliability assessment is the objective, the quality of the rationale
(for defining acceleration factors) takes on additional importance and
another factor comes into play. The failure criterion, generally
referenced to a change in contact resistance, must be known. Discussion
of this issue is controversial and will be discussed in my next
article.”
The
allowed change in contact resistance is arguably the most important
criterion in assessing the performance and reliability of a connector
system. This article is intended to provide some insight into deriving
that criterion for a connector application.
There are many ways a connector may fail, some “catastrophic” and some
“systematic.” Catastrophic failures are a result of “defects,” and
systematic failures a consequence of design/materials choices. Defects
include poor solder joints, bent pins, and “inadequate” plating, among
others. In other words, defects, in this context, are a result of
manufacturing or assembly processes. Failures in this category are often
experienced immediately, or rather abruptly, after short field life.
Such failures, though important, are outside the scope of this
discussion.
The focus here is on systematic, that is, design/material-related
failures. Examples include inadequate contact force in the initial
design; loss in contact force due to stress relaxation, as a result of
improper spring material selection; loss in contact force due to housing
creep, another improper material selection; inadequate plating thickness
for the intended durability life or operating temperature; and variants
or combinations of these and other design/material choices.
Catastrophic failures may manifest themselves in initial high
resistances or abrupt dramatic increases in contact resistance.
Systematic failures, in contrast, generally occur over time and are
characterized by an increase in contact resistance, perhaps at an
increasing rate, after a period of stable contact resistance
performance; or through intermittent high-resistance events. So it is
reasonable to expect that contact resistance measurements, and an
“acceptable” change in contact resistance criterion, may provide a way
to “assess” connector performance over time.
It’s a “reasonable” expectation, but one not necessarily easily
realized. Two issues arise immediately: How should contact resistance be
measured, and what is an “acceptable” change in contact resistance?
Consider each in turn.
Contact Resistance Measurement
Previous articles in this series, “Measure Twice, Test Once” and
“Contact Resistance: Key to Success,” discussed details of contact
resistance measurement practices. Connector qualification testing
protocols generally require low-level circuit resistance (LLCR)
measurements where the open circuit measurement voltage is held to 20
millivolts. This voltage level, sometimes referred to as “dry circuit,”
ensures that the applied voltage is insufficient to disrupt any films,
due to various degradation processes that may exist on the contact
surfaces. Measurements are made before and after a prescribed series of
conditioning and exposure steps, as described in previous articles, “The
How and Why of Connector Testing” and “Devil or Angel: Connector
Testing.” This protocol, however, does not detect resistance
intermittents that were identified above as an indicator of contact
resistance degradation. A degrading contact interface will typically
show an increase in contact resistance with time. But, and more
importantly for our purposes, it will also show increasing frequency of
intermittents and an increase in the magnitude of the change in
resistance at each event as a function of time. Conventional contact
resistance measurements will only detect the overall gradual increase in
contact resistance. To detect the intermittent contact resistance
pattern requires a more sophisticated measurement protocol. Continuous
measurement at a reduced sampling time may be required to capture the
intermittent nature of the changing contact resistance.
Acceptable Change in Contact Resistance
The
increase in contact resistance that can be tolerated by an electrical or
electronic system is, of course, application dependent. In a low current
DC, or low-frequency application, the general gradual increase in
contact resistance cited may be the appropriate criterion. In these
applications, a resistance change of several, or perhaps several tens or
hundreds of milliohms in contact resistance, may be acceptable. The
“upper limit” of acceptable change in resistance may well be determined
by the fact that the general shape of the contact resistance with time
curve begins linearly and transitions to an exponential rate as the
contact interface degrades. This pattern is reasonable when the asperity
model of the contact interface (see Connector Degradation Mechanisms) is
considered. Recall that the contact interface resistance is proportional
to the inverse of the contact area, as contact area decreases the
contact resistance increases. Initially the contact area decreases
slowly in proportion to the initial contact area, but as the contact
area decreases the rate of the proportional decrease in area increases.
So, one limitation to an acceptable change in contact resistance would
be to remain below the “knee” in the contact resistance versus time
curve. This consideration may limit the acceptable change in contact
resistance to the tens of milliohms range.
If, however, the application is a power application, a lower criterion
may be applicable to avoid the effects of Joule, I2R,
heating. It is important to note that a power application is not
necessarily a high-current application if the physical size of the
contact is small. Small contacts will have both a higher bulk resistance
and, more importantly, a higher contact interface resistance because the
contact area will generally be smaller. So for a power contact, a change
in contact resistance criterion may range from a fraction of a milliohm,
to several milliohms, depending on the application current.
High-frequency digital data applications are a different story. In this
case it is the intermittents that may dominate the acceptable change in
resistance criterion. As mentioned, the magnitude of intermittent
resistance fluctuations increases as the contact interface degrades.
And, arguably more importantly, the duration of the intermittent events
tends to increase. The combination of these two effects will limit the
acceptable resistance level to ensure that the “knee” in the contact
resistance versus time curve, as it affects the intermittent event
characteristics, is avoided.
The mechanisms of contact interface degradation leading to increased
contact resistance are reasonably well known in principle, but complex
in practical interpretation. The purpose of this article is simply to
highlight some of the major factors that must be considered in
attempting to derive an acceptable change in contact resistance for a
given application arena, a critical parameter in assessing the
performance and reliability of a connector through a testing protocol.
 |
|
Dr. Robert S. Mroczkowski
Director Technology, Bishop and Associates Inc.
In 1998, Dr. Mroczkowski founded connNtext associates, a
firm providing consulting services in connector
applications to the electronics industry. Dr.
Mroczkowski has more than 30 years experience in various
aspects of the electronics industry. He joined AMP Inc.
in 1971. While at AMP, his responsibilities included
consulting on connector design, materials, and
reliability concerns within AMP, and providing an
interface to AMP customers on the same issues. In 1990
he joined the AMP Advanced Development Laboratories,
where he was responsible for the development of
microstrip cable connectors and a new microcoaxial
connector for medical ultrasound diagnostic equipment.
Dr. Mroczkowski retired in 1998 as an AMP principal. He
is the author of the McGraw Hill Electronics Connector
Handbook, has contributed chapters on connectors and
interconnections to a number of packaging handbooks, and
written more than 20 technical papers. He holds seven
patents. In 1997, Dr. Mroczkowski received the Lifetime
Achievement Award of the International Institute of
Connector and Interconnection Technology.
He holds a bachelor’s, master’s, and doctorate of
science degrees in physical metallurgy from the
Massachusetts Institute of Technology. |
|
|
