1、 EIA STANDARD TP-112 CONTACT RESISTANCE AND CURRENT RATING OF PARALLEL A CIRCUITS TEST PROCEDURE FOR ELECTRICAL CONNECTORS, CONTACTS AND SOCKETS EIA-364-112 JANUARY 2010 ANSI/EIA-364-112-2010 Approved: January 14, 2010 EIA-364-112 EIA Standards Electronic Components Association NOTICE EIA Engineerin
2、g Standards and Publications are designed to serve the public interest through eliminating misunderstandings between manufacturers and purchasers, facilitating interchangeability and improvement of products, and assisting the purchaser in selecting and obtaining with minimum delay the proper product
3、 for his particular need. Existence of such Standards and Publications shall not in any respect preclude any member or nonmember of EIA from manufacturing or selling products not conforming to such Standards and Publications, nor shall the existence of such Standards and Publications preclude their
4、voluntary use by those other than EIA members, whether the standard is to be used either domestically or internationally. Standards and Publications are adopted by EIA in accordance with the American National Standards Institute (ANSI) patent policy. By such action, EIA does not assume any liability
5、 to any patent owner, nor does it assume any obligation whatever to parties adopting the Standard or Publication. This EIA Standard is considered to have International Standardization implication, but the International Electrotechnical Commission activity has not progressed to the point where a vali
6、d comparison between the EIA Standard and the IEC document can be made. This Standard does not purport to address all safety problems associated with its use or all applicable regulatory requirements. It is the responsibility of the user of this Standard to establish appropriate safety and health pr
7、actices and to determine the applicability of regulatory limitations before its use. (From Standards Proposal No. 5179 formulated under the cognizance of the CE-2.0 National Connector and Socket Standards Committee) Published by: ELECTRONIC COMPONENTS ASSOCIATION 2010 EIA Standards and Technology De
8、partment 2500 Wilson Boulevard Suite 310 Arlington, VA 22201 PRICE: Please call: Global Engineering Documents, USA and Canada (1-800-854-7179) http:/ All rights reserved Printed in U.S.A. PLEASE ! DONT VIOLATE THE LAW! This document is copyrighted by the EIA and may not be reproduced without permiss
9、ion. Organizations may obtain permission to reproduce a limited number of copies through entering into a license agreement. For information, contact: Global Engineering Documents 15 Inverness Way East Englewood, CO 80112-5704 or call USA and Canada (1-800-854-7179), International (303-397-7956) CONT
10、ENTS Clause Page 1 Introduction . 1 1.1 Scope . 1 1.2 Object 1 1.3 Background . 1 1.4 Prerequisites 2 2 Definitions 2 3 Test method . 2 3.1 Method A, central limit theorem parallel contact resistance 2 3.2 Method B (reserved for future use) 3 4 Details to be specified . 3 5 Test documentation . 4 Ta
11、ble A.1 Test groups A-1 A.2 Test results A-2 A.3 Monte Carlo simulation results . A-3 A.4 Histogram results A-4 A.5 Distribution of parallel resistance means results A-6 A.6 Unit-to-unit variation relative to the pin-to-pin variation . A-7 Annex A Method A, example central limit theorem parallel con
12、tact resistance (informative) A-1 A.1 Test sequences A-1 A.2 Results . A-2 A.3 Validation A-3 A.4 Histogram results A-4 A.5 Discussion . A-4 A.6 Implementation . A-7 A.7 Summary . A-8 A.8 Notes . A-8 i (This page left blank) ii EIA-364-112 Page 1 TEST PROCEDURE No. 112 CONTACT RESISTANCE AND CURRENT
13、 RATING OF PARALLEL CIRCUITS TEST PROCEDURE FOR ELECTRICAL CONNECTORS AND SOCKETS (From EIA Standards Proposal No. 5179, formulated under the cognizance of EIA CE-2.0 Committee on National Connector Standards) 1 Introduction 1.1 Scope This procedure applies to connectors and sockets when multiple ci
14、rcuits are electrically connected in a parallel configuration and there is a need to determine the expected parallel resistance and or current rating. 1.2 Object 1.2.1 This standard establishes a procedure for determining the predicted parallel contact resistance of a connector or socket when multip
15、le circuits are connected in parallel. A statistical approach is employed to determine the parallel or effective contact resistance using empirical data from EIA connector test sequences such as EIA-364-1000. 1.2.2 The procedure for determining current rating of a connector or socket when multiple c
16、ircuits are connected in parallel will be added in the near future to this document. 1.3 Background Traditional methods for determining low level contact resistance limits for parallel circuits such as a worst case parallel resistance calculation employ extremely conservative calculations that produ
17、ce limits unlikely to be seen in actual application. This procedure outlines a more realistic method that is based on sound statistical principles and determines the parallel resistance limit with a percentile derived from calculations made using easily-obtained data. The limit calculated by this st
18、andard may be used by system designers to plan their voltage drop budget. Unrealistically high limits calculated with the previous method result in inefficient use of power resources in an application. Method A applies the Central Limit Theorem to parallel resistance distributions for low power leve
19、l applications. Method B applies to current rating of parallel circuits using distribution fitting and Monte Carlo simulation. EIA-364-112 Page 2 1.4 Prerequisites The proposed methods are not a substitute for common connector qualification practices. These methods rely upon said qualification test
20、data. In such tests, a maximum resistance and/or a maximum delta resistance limit will be applied to individual contact data. It is not appropriate the apply theses methods to products with open circuits or high resistances that do not meet these criteria. Connector products that do not meet these c
21、riteria should not be considered qualified for use in an application based solely upon its calculated parallel resistance. The connectors must meet a clear individual contact limit before a parallel resistance is calculated. 2 Definitions 2.1 Parallel circuit Two or more connector or socket circuits
22、 electrically connected together on both sides of the separable interface. 2.2 Central limit theorem A mathematical theorem that is central to the use of statistics. It states that for a random sample of observations from any distribution with a finite mean and a finite variance, the mean of the obs
23、ervations will follow a normal distribution. This theorem is the main justification for the widespread use of statistical analyses based on the normal distribution. 3 Test method 3.1 Method A, central limit theorem parallel contact resistance See annex A for an example and additional information. 3.
24、1.1 Identify the end-of-life low level contact resistance measurements, R, for each stress group in a typical connector qualification test. Calculate the mean ( R ) and standard deviation (sR) of the distribution of R for each test group. 3.1.2 Determine the number of contacts (n) in the parallel ci
25、rcuit for the connector of interest. 3.1.3 Calculate the expected parallel resistance ( pR ) through the connector of interest using the following formula. nRR p EIA-364-112 Page 3 3.1.4 Consideration may be given to analyzing the data per the actual use case. The computations for Rpwould use the ex
26、act contact positions planned in the final application. Alternately, if other contacts in the product are identical to those used in the final application, then that data could also be utilized. 3.1.5 Use the following formula to calculate the expected parallel resistance standard deviation (sp), fo
27、und by applying the central limit theorem. nnssRp3.1.6 Per the central limit theorem, assume the distribution of pR is normal. 3.1.7 Calculate the percentile of interest using pR and sp and normal probability according to the following formula. pp szRP +=993.1.8 The value of z comes from a z-table.
28、For the 99th percentile, z = 2.326. For a more conservative calculation, a t-table may be used rather than a z-table. When a t-table is used, the degrees of freedom equals n 1. 3.2 Method B, current rating of parallel circuits (Under development) 4 Details to be specified The following details shall
29、 be specified in the referencing document: 4.1 Analysis method to be used 4.2 End of life contact resistance data for each test sequence from a connector qualification test program such as EIA-364-1000 (see the annex(s) for an example) EIA-364-112 Page 4 5 Test documentation Documentation shall cont
30、ain the details specified in clause 4, with any exceptions, and the following: 5.1 Title of test 5.2 Test equipment used, and date of last and next calibration 5.3 Results of the analysis (see the annex(s) for an example) 5.4 Name of operator and start/finish dates of test EIA-364-112 Page A-1 Annex
31、 A Method A, example central limit theorem parallel contact resistance (informative) The following data illustrates the method using data from a sample of 25 FB DIMM parts, divided into five groups and subjected to the following stress tests: A.1 Test sequences Table A.1 Test groups Test Group Test
32、or Examination A B C D E 1 Low Level Contact Resistance 1,4,6 1,4,6,8 1,3,5,7 1,4,6,8,10 1,4,6,8 2 Durability 2 2 2 2 2 3 Temperature Life 3 4 Thermal Shock 3 5 Physical Shock 6 6 Random Vibration 4 7 Cyclic Temperature &Humidity 5 8 Reseating 5 7 9 7 9 Mixed Flowing Gas 5 10 Thermal Cycling 5 11 Th
33、ermal Disturbance 7 12 Temperature Life (Preconditioning) 3 3 Specimen quantity (pieces) 5 5 5 5 5 NOTES 1 Group C did not compete testing on all contacts. 2 The data included here is for those contacts on which testing was completed. EIA-364-112 Page A-2 A.2 Results The method produces the results
34、given below for circuits using 4, 22, and 59 contacts in parallel. Table A.2 Test results Group Number of Data Points (Readings) R sRn (Number of contacts in parallel) pR sp99th%tile A 1200 7.4409 1.0617 4 1.86022 0.13272 2.16891 B 1200 7.7255 0.714 4 1.93136 0.08925 2.13895C 290 5.7362 0.4641 4 1.4
35、3406 0.05801 1.56900 D 1200 7.0192 0.5723 4 1.75479 0.07153 1.92118E 1200 7.676 0.7201 4 1.91900 0.09001 2.12836 A 1200 7.4409 1.0617 22 0.33822 0.01029 0.36215B 1200 7.7255 0.714 22 0.35116 0.00692 0.36725 C 290 5.7362 0.4641 22 0.26074 0.00450 0.27120D 1200 7.0192 0.5723 22 0.31905 0.00555 0.33195
36、 E 1200 7.676 0.7201 22 0.34891 0.00698 0.36514A 1200 7.4409 1.0617 59 0.12612 0.00234 0.13157 B 1200 7.7255 0.714 59 0.13094 0.00158 0.13460C 290 5.7362 0.4641 59 0.09722 0.00102 0.09961 D 1200 7.0192 0.5723 59 0.11897 0.00126 0.12191E 1200 7.676 0.7201 59 0.13010 0.00159 0.13380 EIA-364-112 Page A
37、-3 A.3 Validation A.3.1 A Monte Carlo simulation can validate the results given above. The simulation method uses bootstrapped data points from the original raw data set with replacement. For a simulation meant to validate a 4-pin circuit, the simulation randomly selects 4 readings from a single str
38、ess group. The simulation calculates the parallel resistance of these 4 points and saves it. The simulation repeats this until 2,500 parallel resistances for the given pin count and stress group have been calculated. A.3.2 The Monte Carlo simulations produced the results given below. The percentile
39、here is the actual percentile from the simulated parallel resistances, not a calculation based on the mean and standard deviation of those resistances. The is the difference between the prediction percentile given in the formulaic method above and the simulation percentile. Table A.3 Monte Carlo sim
40、ulation results Group Simulations n pR sp99th%tile A 2500 4 1.83133 0.12860 2.15800 0.01091B 2500 4 1.91569 0.08984 2.11717 0.02178C 2500 4 1.42808 0.05240 1.57970 -0.01070D 2500 4 1.74697 0.07120 1.91877 0.00241E 2500 4 1.90938 0.08609 2.13530 -0.00694A 2500 22 0.33213 0.00951 0.35617 0.00598B 2500
41、 22 0.34803 0.00707 0.36431 0.00294C 2500 22 0.25931 0.00403 0.26997 0.00123D 2500 22 0.31711 0.00542 0.32952 0.00243E 2500 22 0.34621 0.00643 0.36226 0.00288A 2500 59 0.12373 0.00213 0.12894 0.00263B 2500 59 0.12983 0.00157 0.13336 0.00125C 2500 59 0.09666 0.00080 0.09866 0.00095D 2500 59 0.11820 0
42、.00122 0.12094 0.00096E 2500 59 0.12905 0.00145 0.13260 0.00120EIA-364-112 Page A-4 A.4 Histogram results The charts below graphically depict the results. The histograms represent the data for each group and pin count. The red lines represent the calculated limit from the prediction percentile. The
43、blue lines represent the actual percentile from the simulations. Table A.4 Histogram results Group Raw n = 4 n = 22 n = 59 A 50100150200250Count.117 .119 .121 .123 .125 .127 .129 .131501001502001.5 1.6 1.7 1.8 1.9 2 2.1 2.2 2.3100200300.31 .32 .33 .34 .35 .36 .37501001502006 7 8 9 10B 100200300.32 .
44、33 .34 .35 .36 .37100200300Count.124 .126 .128 .13 .132 .1341002003001.7 1.8 1.9 2 2.1 2.2501001502006 7 8 9C 100200300Count.094 .095 .096 .097 .098 .099 .11002003004001.3 1.4 1.5 1.6 1.750100150200250.25 .26 .272550751005 6 7 8 9D 50100150200Count.115.116.117 .118.119 .12 .121.122.1231002003001.6 1
45、.7 1.8 1.9 2100200300.3 .31 .32 .33501001502005.5 6 6.5 7 7.5 8 8.5 9E 1002003006 7 8 9 10 111002003001.7 1.8 1.9 2 2.1 2.2 2.3100200300400.33 .34 .35 .36 .37100200300400Count.125 .127 .129.13 .131 .133NOTE The method was replicated at two other manufacturers with comparable results. A.5 Discussion
46、A.5.1 The method used here begins with the common formula to calculate individual parallel resistance values, Rp. It is typically used to find the total resistance across n resistors in parallel or, in the case of this paper, n contacts used in the particular parallel circuit. =niipRR111EIA-364-112
47、Page A-5 A.5.2 The form for calculating the expected value of the distribution of m of these parallel resistances is below. In the case of this paper, this distribution is of m connectors. 1mRRmjpjp= or substituting =mjjniipRmR11111. A.5.3 This form could be custom programmed into a script, though t
48、he approximation of its value as 1/n that of the arithmetic mean of the end-of-life resistances is close enough for this application and doesnt require any custom scripting or programming. Therefore we use the simpler version. nRR p A.5.4 The form for calculating the variance of the distribution of
49、Rp is: 1)(122=mRRsmjppjpor substituting 111111121112=mRmRsmjmjjniijniip. EIA-364-112 Page A-6 A.5.5 One attribute that the method leverages is that the Central Limit Theorem applies to pR the same as it would to R . Evidence that this is true comes from the Monte Carlo simulation used to validate the formulae. As the number of contacts in the circuit increases, the distribution of the parallel resistance means becomes more and more normal. Table A.5 Distr
