IEEE TRANSACTIONS ON COMPONENTS. HYBRIDS, AND MANUFACTURING TECHNOLOGY, VOL. 1 I , NO. 4. DECEMBER 1988
`
`393
`
`A Limited Study of the Effects of Contact Normal
`Force, Contact Geometry, and Wipe Distance on
`the Contact Resistance of Gold-plated Contacts
`
`IRVIN H. BROCKMAN, CHARLES S. SIEBER, AND ROBERT S. MROCZKOWSKI
`
`results of a limited study of the interactions between
`Abstract-The
`contact normal force, contact geometry, and wipe distance on contact
`resistance are reported. Two values of normal force, three contact
`geometries, and three surface conditions were studied. The normal forces
`used were 50 and 120 g. The three contact geometries were a cylinder, a
`hemisphere, and an elliptical dimple. In all cases these geometries were
`mated to a flat which was either clean, lubricated, or lubricated and
`dusted to a 50 percent coverage condition. Both the contacts and the flats
`were gold over nickel plated. Wiping action was provided by a stepper-
`motor-driven table over a 20 mil distance under the full normal force.
`It is shown that wipe distances of less than 10 mils produce stable
`contact resistance values at a normal force of 120 g for the hemisphere-
`flat geometry, even on the dust-covered surfaces. The ellipse-flat
`geometry is less effective, although effective wiping action is observed at
`distances less than 20 mils. The cylinder-flat geometry shows marginal
`wiping effectiveness.
`For normal forces of 50 g, only the hemisphere-flat geometry shows
`effective wiping action on dust-covered surfaces. For this geometry, the
`results of wiping action were similar to those at 120 g. The elliptical
`geometry was marginally effective, and the cylinder ineffective in wipe at
`50 g.
`
`INTRODUCTION
`NE OF THE key requirements of an electronic connector
`
`0 is to establish and maintain an acceptable, usually low,
`
`and stable value of connection resistance. The resistance of a
`connection is made up of three contributions: termination
`resistance(s), bulk resistance(s), and contact resistance. Ter-
`mination resistance refers to the resistance associated with the
`permanent termination(s) made to external circuitry. The bulk
`resistance is the resistance due to the materials of the contact
`spring(s). The contact resistance is the resistance across the
`interface between the contact springs in each half of the
`connector. Our concern in this paper is with contact resist-
`ance, and, in particular, with the effects of contact wipe on
`establishing a low and stable value of contact resistance.
`Contact wipe refers to the relative motion of the two contact
`spring surfaces over one another as the connector is mated.
`Little work has been reported on contact wipe and this paper
`reports the results of a limited study of wiping action on clean,
`lubricated, and lubricated and severely dust-contaminated
`surfaces. Details of the samples will be provided below.
`A low and stable contact resistance will be obtained when a
`
`Manuscript received March 15, 1988; revised August 8, 1988. This paper
`was presented at the 38th Electronic Components Conference, Los Angeles,
`CA, May 9-11, 1988.
`The authors are with AMP Incorporated, Harrisburg, PA 17105.
`IEEE Log Number 8824138.
`
`metallic interface is established between the two halves of the
`connector. One of the major functions of a contact finish is to
`facilitate the creation of such a metallic interface. In the case
`of precious metal contact finishes, such as gold, the nobility of
`the surface provides metallic contact as long as surface
`contaminants are absent or displaced. Contact wipe in this case
`is intended to displace surface Contaminants. In non-noble
`finishes, such as tin, wiping action is necessary to displace the
`natural oxides which occur on such surfaces in addition to
`displacement of contaminants.
`The effectiveness of wiping action in displacing oxides and
`surface contaminants is dependent on at least four factors:
`1) the contact normal force, that is, the force perpendicular
`to the interface between the contacting surfaces;
`2) the geometry of the contacting interfaces;
`3) the extent and nature of the surface contamination;
`4) the distance over which the wipe occurs.
`This paper reports on the results of a limited study of the
`interactions of these factors on gold-plated contact surfaces in
`an attempt to answer the following question:
`
`How much wipe distance is necessary to ensure displace-
`ment of surface oxides or contaminants so that a low and
`stable value of contact resistance can be realized?
`
`The answer to the question depends, of course, on the first
`three factors mentioned previously. This study covered three
`contact geometries (hemisphere-flat, ellipse-flat, and cylin-
`der-flat), two values of normal force (50 and 120 g), and three
`surface conditions (a clean surface, a lubricated surface, and a
`lubricated surface with dust applied to a surface coverage of
`approximately 50 percent). The high level of dust contamina-
`tion was used to provide a severe condition for wiping action
`to overcome and is not intended to represent application
`conditions.
`table. The
`Wiping distance was controlled by an x-y
`effectiveness of wiping action was determined by the value and
`stability of contact resistance as measured by a four-wire
`method. Plots of contact resistance versus wiping distance
`were obtained and a plateau in contact resistance was taken as
`an indicator of adequate wiping distance.
`It was found, as expected, that wiping effectiveness
`increases with increasing normal force and with geometries
`which were increasingly capable of penetrating and displacing
`the dust. Even under the severe contamination conditions of
`
`0148-641 1/88/1200-0393$01.00 0 1988 IEEE
`
`Feinmetall Exhibit 2027
`FormFactor, Inc. v. Feinmetall, GmbH
`IPR2019-00082
`
`Page 1 of 8
`
`

`

`394
`
`IEEE TRANSACTIONS ON COMPONENTS. HYBRIDS, AND MANUFACTURING TECHNOLOGY, VOL. 11, NO. 4, DECEMBER 1988
`
`50-percent dust coverage, penetration under application of the
`load and effective displacement during wiping action with
`wipe distances as small as a few mils, was observed for
`hemisphere-flat contacts at 120 g of normal force. The other
`geometries showed a greater variability in wiping effective-
`ness under this load. The ellipse-flat configuration showed
`reasonably effective wiping action, but performance varied.
`The cylinder-flat geometry showed instances of effective
`wipe, but, in general, wiping action was not very effective.
`At 50 g normal force the hemisphere-flat configuration
`maintained good wiping performance, although penetration of
`the dust on application of the load was not achieved. The other
`two geometries showed widely variable performance, with the
`ellipse-flat configuration showing marginally effective wipe,
`and the cylinder-flat being totally ineffective at the reduced
`normal force.
`
`
`
`EXPERIMENTAL PROCEDURES
`
`Equipment
`The control of instrument functions and the measurement,
`acquisition, analysis, and subsequent presentation of data were
`accomplished via an automated contact resistance probe
`(ACRP). A photograph of the ACRP is presented in Fig. 1.
`The ACRP was designed and constructed in the Contact
`Physics Department of the Technology Group of AMP
`Incorporated. It is similar to other automated probes described
`in the literature [ 11, [2] but contains unique features and differs
`in the mode of operation from these systems.
`The loading mechanism of the probe is hydraulic, using a
`float and two liquid reservoirs as shown in Fig. 2 . The motion
`of the float controls the motion of the loading arm and the rate
`of application of contact normal force. There is no wiping
`action on application of the load. Vibration-free loading is
`obtained through a system of baffles and isolation of the x-y
`table using shock-absorbent materials.
`Movement of the x-y table is accomplished via precision
`stepping-motor drives. Indexing of the table controls the
`location of the contact resistance measurement points. In this
`work, the x-y table was also used to provide the wiping
`motion after the load was applied.
`Contact resistance measurements were made using four-
`terminal dry-circuit measurement technique. The experimental
`arrangement of the test points is shown in Fig. 3. The
`maximum open-circuit voltage was 50 mV. The constant dc
`test current was limited to 50 mA or less, and the current was
`reserved for each data point to eliminate possible contact
`potential effects. The contact resistance reported is the average
`for the two current directions.
`Control of all the functions mentioned is provided by an HP
`9845 desktop computer using software developed specifically
`for the ACRP. Software was also developed to allow the
`computer to record, store, analyze, and plot the data for
`presentation in various formats, some of which are used in this
`paper.
`
`Samples
`Test Flats: The test flats used in this work were one inch
`squares of C51100 phosphor bronze plated with 1.3 pm nickel
`
`Fig. 1. A photograph of the automated contact resistance probe (ACRP)
`used in this study.
`
`1
`
`Fig. 2. Schematic diagram of the hydraulic loading system used in the
`ACRP.
`
`T
`0
`
`Fig. 3. Experimental arrangement of test points for four-wire contact
`resistance measurements in the study.
`
`followed by 1.3 pm cobalt hardened gold and a 0.25 pm top
`layer of soft gold.
`Three surface conditions of the flats were used:
`1) Clean: The flat was degreased with 1, 1, 1 trichloro-
`ethane followed by an acetone rinse.
`2) Lubricated: After the cleaning process the flat was
`lubricated with Exxon Turbo 2380 Synthetic Lubricant.
`3) Lubed and Dusted: After cleaning and lubricating, the
`flats were dusted to a 50 percent coverage condition by sifting
`a proprietary dust formulation through a 270 mesh screen.
`The presence of the lubricant served to retain the dust under
`these heavy dust conditions. This lubed and dusted surface is
`
`Page 2 of 8
`
`

`

`BROCKMAN et al.: CONTACT KESISTANCE OF GOLD-PLATED CONTACTS
`
`39s
`
`//
`
`I
`
`~
`
`TABLE I
`WIPE DISTANCE (AVERAGE) IN MILS FOR STABLE CONTACT RESISTANCF
`-~ __ _ ~ ~ ~ _ _ ~ - ~ _ _ _
`Surface Condition
`Lubricated
`
`Geometry
`
`Clean
`
`LubeiDust
`
`Ellipse
`
`Fig. 4. Schematic diagram of the contact geometries: cylinder. hemisphere,
`and ellipse used in the study. These geometries were mated to a flat in all
`cases
`
`Hemisphere
`120 g
`50 g
`Ellipse
`120 g
`50 g
`Cylinder
`120 g
`50 g
`
`none (HC120)
`none (HC50)
`
`none (HL120)
`none (HL50)
`
`none (EC120)
`none (ECS0)
`
`none (EL120)
`none (EL50)
`
`< 10 (HLD120)
`10+ (HLD50)
`
`> 15 (ELD120)
`>20 (ELD50)
`
`none (CC120)
`none (CCSO)
`
`none (CL120)
`none (CL50)
`
`>20 (CLDlZO)
`ineffective (CLDSO)
`
`Fig. 5. Experimental detail of the retention of the contact spring in the
`ACRP sample holder.
`
`believed to be a severe condition for wiping action to penetrate
`and displace.
`Contact Springs: The contact spring and contact geome-
`tries used in this work are schematically illustrated in Fig. 4.
`The contact geometries were stamped into the C51100
`phosphor bronze contact springs using typical manufacturing
`practices.
`The dimensions of the contact geometries were as follows:
`Cylinder: The cylinder contact geometry was a flat surface
`containing no embossment. The radius of curvature of the
`cylinder was 34 mils. The direction of motion during wiping
`action was perpendicular to the axis of the cylinder.
`Hemisphere: A hemispherical dimple of 25 mil (nominal)
`radius was embossed into the cylinder described above. The
`dimple protruded approximately 5 mils above the surface of
`the cylinder.
`Ellipse: The embossment in this case was roughly elliptical
`with a major axis of 40 mils and a minor axis of 18 mils. The
`embossment protruded approximately 2 mils above the cylin-
`der surface. The direction of motion of the ellipse during
`wiping was parallel to the major axis.
`After stamping, the contact springs were plated, under
`typical production plating practices, with a 1.3 pm nickel
`underplate followed by 1.3 pm of cobalt-hardened gold.
`
`Procedures
`The contact spring containing the contact geometry of
`interest was mounted in the sample holder as shown in Fig. 5.
`Proper mounting procedures permit perpendicular alignment
`of the contact geometry with the test flat to ensure that no
`wiping action occurs on application of the normal force.
`
`\ L
`
`HLD120-2
`
`+ i k e i k J
`VI '
`LORD (qn)
`WIPE DISTRNCE (mils)
`Fig. 6 .
`Representative curve, HLDIZO-2, of contact resistance versus load-
`wipe. Wiping action begins after the full contact load is applied. in this case
`at 120 g.
`
`The individual steps in obtaining contact resistance data as a
`function of load and wiping action were as follows:
`1) The contact load was applied at a rate of 2 g/s.
`2) Voltage and current measurements, used to calculate
`contact resistance, under the dry-circuit conditions described
`previously, began at 10 g and were taken at 5 g intervals up to
`the maximum load. In this work the maximum load was either
`50 or 120 g.
`3) After the maximum load was applied, wiping action
`through motion of the x-y table was initiated. Wipe pro-
`ceeded, in 1 mil intervals, up to a maximum of 20 mils.
`4) Voltage and current measurements were made at each 1
`mil increment of wipe.
`5 ) The contact load was removed.
`On completion of the load-wipe cycle the contact spring was
`removed from the holder and a new one inserted. The X-y
`table was indexed to a new location and the load-wipe cycle
`repeated. Nine such cycles were performed on each contact
`geometry-surface condition combination.
`
`RESULTS
`A summary of the data obtained from this study is presented
`in matrix form in Table I. Contact resistance versus normal
`force-wipe distance curves for the individual combinations are
`provided in the text. A representative curve, HLD120-2, is
`shown in Fig. 6 to illustrate the manner in which the variables
`are presented.
`
`Page 3 of 8
`
`

`

`396
`
`IEEE TRANSACTIONS ON COMPONENTS, HYBRIDS. AND MANUFACTURING TECHNOLOGY, VOL. 1 I , NO. 4, DECEMBER 1988
`
`-
`80
`-
`60
`-
`40
`
`.-.
`2 2 0 -
`g
`
`‘1
`
`EC120
`
`- WIPE A DISTRNCE i (mIIs) &
`i $ A
`k i + i
`LOAD (qm)
`Fig. 9. Contact resistance (distribution) versus normal force and wipe for
`ellipse-flat geometry on a clean surface at 120 g normal force, EC120.
`
`6
`
`cc50
`
`80
`~ -
`60
`-
`40
`
`-
`d 20-
`
`0
`
`I
`
`60
`-
`40
`
`-
`$ 2 0 -
`B
`
`‘1
`
`HL50
`; A & & & ’
`l & +
`L & k
`LWD ( q m )
`WIPE DISTRNCE (mils)
`Fig. 8. Coc!act resistance (distribution) versus normal force and wipe for
`hemisphere-flat geometry on a lubricated surface at 50 g normal force, HL50.
`
`
`
`The curves display both normal force and wipe distance
`along the abscissa. The first segment of the abscissa shows
`contact resistance variation as the normal force is increased.
`The application of normal force occurs under nonwiping
`conditions as commented previously. The second segment of
`the abscissa shows the variation in contact resistance as wiping
`action occurred.
`Since the curves for the clean and lubricated surfaces show
`little variation, only representative samples are included in
`Figs. 7-10. These curves show that wiping action is not
`necessary on clean and clean-lubricated surfaces. Simple
`application of normal force serves to establish the desired
`metallic interface for all geometries. It was noted, however,
`that slight increases in contact resistance occurred as wiping
`action began in some cases. The magnitude of the increases
`varied with the geometry, being higher for the ellipse (Fig. 9).
`The increase is attributed to a reduction in the contact area as
`sliding begins. The initial contact area is developed over time
`and some “creep” of the soft gold plating on the flat may
`occur. This time-dependent area is lost as the contact begins to
`wipe. The variation of the magnitude of the effect with
`geometry is attributed to alignment variations, especially in
`maintaining the major axis of the ellipse parallel to the surface
`of the flat.
`
`‘1
`
`EL50
`
`I
`
`HIPE DISTANCE ( m i l s )
`LORD ( q d
`Fig. IO. Contact resistance (distribution) versus normal force and wipe for
`ellipse-flat geometry on a lubricated surface at 50 g normal force, EL50.
`
`The data from the lubricated and 50 percent dust-covered
`surfaces are far more interesting. In these curves effects of
`both normal force and geometry are readily apparent.
`
`120 g Normal Force
`At this normal force, a value typical of many connector
`systems, all the geometries show some degree of wiping
`effectiveness. The data in Figs. 11-13 show the average
`contact resistance versus load-wipe data for the three contact
`geometries. The hemisphere-flat geometry, HLD120, shows
`effective wiping action beginning at 2 mils (Fig. 11). The
`ellipse-flat combination requires more than 15 mils (Fig. 12)
`to achieve a plateau in contact resistance. The cylinder-flat
`geometry, CLD120, does not provide, on average, effective
`wipe at the 20-mil limit of this study (Fig. 13). This average
`data do not, however, completely demonstrate the difference
`in wiping effectiveness of the three geometries.
`Figs. 14-16 show the individual contact resistance versus
`load-wipe curves for these geometries. The range of variation
`in the distance at which wiping becomes effective varies
`significantly with geometry.
`For the hemisphere-flat geometry, HLD120, Fig. 14,
`application of the normal force alone is sufficient to produce
`
`Page 4 of 8
`
`

`

`L
`
`I
`
`l 0 8 8 ~
`
`>
`'\
`
`.
`
`'1,
`
`t
`
`ELi)12c
`-
`-
`- m g z
`m m m m
`&-.-I-
`Z
`Z
` E
`d - - L - L L - L '
`m
`m
`?OR3 1 q m 1
`WIPE CIS-ANCE ~ m l
`l s ~
`Fig. 12
`Contact resistance (a\erage) versus normal force and wipe for
`ellipse-flat geometry on a lubricated and du\ted urface at 120 g normal
`torce. ELD120
`
`
`
`-_
`
`\ \
`'-1
`
`! 000 -i B L
`
`
`
`CL D120
`
`WIPE DISiRNCE (mils)
`LORD (sal
`Fig. 13. Contact resistance (average) \ersus normal force and uipc for
`cylinder-flat geometry o n a lubricated and duated surtace at I20 g norni;d
`force. CLDI2O.
`
`-
`
`-
`
`N
`
`
`
`penetration of the dust in some cases. The onset of wiping
`action produces sharp decreases in contact resistance almost
`immediately. Wiping in the range of 6 to I O mils produces the
`plateau in contact resistance indicative of stability.
`
`The results are different for the ellipse-flat case. ELD120.
`The data in Fig. IS show that normal force alone is not
`sfifiicient in this geomt.try under these severe contamination
`conditionx. The onset of wipe does not immediately lead to
`sharp decreases in contact resistance. In some cases. a wiping
`distance of 6 mils is necessary to produce 9 drop in contact
`resistance. Variation in contact resistance as wiping occur\ is
`also larger for this geometry.
`The data in Fig. 16 illustrate the poor wiping characteristics
`of the cylinder-flat geimetry . (11-D 120. under :hex w't're
`dust conditions. Wiping action begins to show effects at 5
`mils, but the effect is inconsistent e\en in the best case. At the
`other extrenic. wiping action of 20 mils, the longest in the
`program. was ineffectivc for some of the runs.
`The inconsistency ot'the M iping effecti\encl.;\ is attributed to
`the poor penetrating ability of the cylinder to flat mating which
`is due to the larger apparent contact area of the cylinder-flat
`geometry. As the surfaces conic in contact, the dust is trapped
`between the two relatively flat surfaces rather than being
`penetrated o r displaced. In addition to this cffecT. the la:-ger
`apparent i'ontact area also affects the Tone of susceptibility to
`the dust 131. Dust can be effective in physical separation of the
`
`Page 5 of 8
`
`

`

`398
`
`IEEE TRANSACTIONS ON COMPONENTS. HYBRIDS, AND MANUFACTURING TECHNOLOGY, VOL. 1 1 , NO. 4, DECEMBER 1988
`
`\
`
`:
`:
`"
`Z
`"
`WIPE DISTRNCE (mils)
`Fig. 16. Contact resistance (distribution) versus normal force and wipe for
`cylinder-flat geometry on a lubricated and dusted surface at 120 g normal
`force, CLD120.
`
`
`
`!
`:
`" E "
`WIPE DISTANCE ( a i l s )
`Fig. 18. Contact resistance (average) versus normal force and wipe for
`ellipse-flat geometry on a lubricated and dusted surface at 50 g normal
`force. ELD5O
`
`
`
`L+kcA$A"LbAm'
`WIPE DISTRNCE (rills)
`LORD (qn)
`Fig. 17.
`Contact resistance (average) versus n o d force and wipe for
`hemisphere-flat geometry on a lubricated and dusted surface at 50 g normal
`force, HLDSO.
`
`N
`
`
`
`-
`
`-
`
` 2
`"
`F
`°
`WIPE DISTANCE ( I 1 Is)
`Fig. 19. Contact resistance (average) versus normal force and wipe for
`cylinder-flat geometry on a lubricated and dusted surface at 50 g normal
`force, CLDSO.
`
`I
`
`
`
`I
`
`contact surfaces when it lies anywhere within the linear zone
`of contact established between the cylinder and the flat.
`The apparent contact areas of the ellipse and hemisphere to
`flat geometries are significantly smaller so the area of
`influence of the dust is concomitantly reduced. The ellipse-flat
`geometry, which has a larger apparent contact area, shows a
`greater variability in wiping effectiveness than does the
`hemisphere-flat, as would be expected.
`50 g Normal Force
`A normal force of 50 g was selected because it is expected
`that normal forces in this range will see increasing use as
`connectors decrease in size and increase in contact density. At
`50 g normal force, the results follow a similar pattern to those
`at 120 g, but the required wipe distance and the effectiveness
`of the wiping action in displacement of the dust are, in general,
`different from those at the higher normal force.
`Figs. 17-19 show the average contact resistance versus
`load-wipe data for the three contact geometries. The data for
`the hemisphere-flat configuration, HLDSO, Fig. 17, appear
`similar to the data at 120 g. One difference is that the
`application of normal force is not sufficient to penetrate the
`
`dust-covered surface. Wiping action of 5 mils, however, does
`begin to achieve displacement at 50 g with the hemispherical
`geometry.
`The data in Figs. 18 and 19 illustrate the relative ineffective-
`ness of wiping action in the ellipse-flat, ELDSO, and cylinder-
`flat, CLDSO, geometries. The ellipse shows limited wiping
`effectiveness under these severe dust conditions, and the
`cylinder shows virtually no effect of wipe.
`Fig. 20 shows the individual curves for the hemisphere-flat
`geometry at 50 g, HLDSO.
`Wiping action produces a sharp drop in contact resistance at
`less than 5 mils of wipe even at 50 g normal force. The
`penetrating geometry and the small area of susceptibility to
`dust are responsible for this good performance. Figs. 21 and
`22 show the individual data for the ellipse-flat, ELDSO, and
`cylinder-flat, CLDSO, configurations. The ellipse demon-
`strates variable wiping characteristics. At wipe distances of the
`order of 20 mils, the contact resistance has attained an
`apparently stable value. At lower wipe distances, however, the
`performance varies significantly. While some instances of
`effective wiping occur with the cylinder geometry, the general
`wiping behavior is not satisfactory.
`
`Page 6 of 8
`
`

`

`BROCKMAN et al.: CONTACT RESISTANCE OF GOLD-PLATED CONTACTS
`
`399
`
`LOAD
`
`WIPE
`AT€
`
`L
`
`-
`
`-
`
`
`
`
`
`1em
`-
`7
`L p
`8
`:
`-
`
`U )
`
`-
`
`
`
`U1 U ( Y -
`
`
`
`1 0 7
`
`~
`
`HLDSO
`b & i & .
`
`$
`
`z A ’
`2 “
`
`
`
`:
`
`:
`
`
`
`Fig. 23. Combined data for the three geometries on lubricated and dusted
`surfaces at 120 g normal force, HLD120, ELD120, and CLD120. Wiping
`action began at the 120 g marker.
`
`L
`
`P I - ; t
`
`I aae
`
`I
`
`w
`U z
`U c
`13
`v, w tr
`
`Fig. 21. Contact resistance (distribution) versus normal force and wipe for
`ellipse-flat geometry on a lubricated and dusted surface at 50 g normal
`force, HLDSO.
`
`I
`
`I
`
`Fig. 22. Contact resistance (distribution) versus normal force and wipe for
`cylinder-flat geometry on a lubricated and dusted surface at 50 g normal
`force, CLDSO.
`
`DISCUSSION
`The results of this study are summarized in Figs. 23 and 24.
`Fig. 23 shows the combined results at 120 g normal force, and
`Fig. 24, the results at 50 g. The differences in penetration,
`
`Fig. 24. Combined data for the three geometries on lubricated and dusted
`surfaces at 50 g normal force, HLDSO, ELDSO, and CLDSO. Wiping action
`began at the 50 g marker.
`
`indicated by deviations in the “flat” portion of the curve as the
`load is applied, are apparent. The variation in wiping
`effectiveness, however, is even more striking, as shown by the
`data beyond the point of maximum load.
`The results of this limited study cannot be directly related to
`performance in connectors, however. Differences exist be-
`tween the test conditions and typical application conditions.
`Among the differences are the following:
`1) The wiping action in the study occurred under full
`applied load while in a typical connector wiping occurs as the
`load increases. This is a minor difference since full load is
`generally achieved after a short mating distance.
`2) The dust in the study was loosely bonded to the surface,
`whereas in environments where corrosion products occur,
`those products will be bonded more tightly to the contact
`surface.
`3) The coverage of the surface by the dust was much higher
`than would be expected in operation.
`Factors 1 and 2 would enhance the effectiveness of wiping
`action while factor 3 would have an opposite effect.
`Despite these differences, these results demonstrate that
`wiping action can be effective in displacing surface dust
`contamination at wiping distances as small as a few mils
`depending on the geometry of the contact interface and the
`normal force.
`One other aspect of the use of penetrating geometries must
`be mentioned. Penetrating geometries can be very effective in
`establishing contact areas, but their effects on the durability of
`
`Page 7 of 8
`
`

`

`400
`
`IEEE TRANSACTIONS ON COMPONENTS, HYBRIDS, AND MANUFACTURING TECHNOLOGY, VOL. 1 1 , NO. 4, DECEMBER 1988
`
`the contact interface during subsequent matings must also be
`considered [4]. Separable connectors generally have a durabil-
`ity requirement in terms of the number of mating cycles which
`can be realized without affecting performance. This aspect of
`connector performance will be negatively affected by contact
`geometries which are “sharp” and this tradeoff must be
`considered in the overall connector design.
`
`CONCLUSION
`The results of this limited study demonstrate that wiping
`action can be effective in penetration and displacement of
`surface dust contamination at distances of a few mils. The
`effectiveness of wipe increases as normal force and the
`penetrating ability of the contact geometry increases. Another
`factor which influences the results is the size of the apparent
`contact area produced by the contact geometry. As the
`apparent contact area decreases the influence of dust particles
`decreases since the dust is less likely to occur within this area.
`
`These two effects work in tandem to improve the wiping
`performance of penetrating contact geometries. While these
`results cannot be directly related to connector performance due
`to differences between the test conditions and typical connec-
`tor applications, they do suggest that a wipe distance of the
`order of 10 mils is adequate to ensure effective wiping action
`in typical connector applications. The negative effects of
`penetrating geometries on durability performance, however,
`must also be considered in connector design.
`
`111
`‘*I
`
`131
`
`r41
`
`REFERENCES
`M. Antler, “Automated contact resistance probe,” Rev. Sci. Zn-
`strum., vol. 34, pp. 1317-1322, Dec. 1963.
`G. J. Russ, “A system for analyzing contact resistance,” in Proc. 29th
`Electronic Components Conf. (Cherry Hill, NJ, May 14-16, 1979).
`J. B. P. Williamson et al., “The influence of dust particles on the
`contact of solids,” Proc. Roy. Soc., A, vol. 237, pp. 560-573, 1956.
`M. Antler et al., “Base metal contacts: An exploratory study of
`separable connection to tin-lead,’’ in Proc. 20th Holm Seminar on
`Electric Contacts (Chicago, IL, Oct. 29-31, 1974).
`
`Page 8 of 8
`
`