Technical report Approaches to Life Estimation of Electronic Circuits

Similar documents
Yutaka Shikano. Visualizing a Quantum State

Agilent 4263B LCR Meter Operation Manual. Manual Change. Change 1 Add TAR in Test Signal Frequency Accuracy Test (Page 9-38) as follows.

Reactive Fluid Dynamics 1 G-COE 科目 複雑システムのデザイン体系 第 1 回 植田利久 慶應義塾大学大学院理工学研究科開放環境科学専攻 2009 年 4 月 14 日. Keio University

業績リスト 2006 年 1 月 ~2006 年 12 月, 各論文の最終行の ( ) 内はキーワード

Search and Digitalization of Maps at the National Diet Library

質量起源 暗黒物質 暗黒エネルギー 宇宙線 陽子崩壊 ニュートリノ質量 米国 P5 ニュートリノ CPV 宇宙背景ニュートリノクォーク レプトンマヨラナ粒子 ニュートリノ測定器 陽子崩壊探索. Diagram courtesy of P5. Origin of Mass.

2015 年度研究活動報告理工学術院 先進理工 応用物理学科小澤徹 Department of Applied Physics, Waseda University

近距離重力実験実験室における逆二乗則の法則の検証. Jiro Murata

ADEOS-II ミッションと海洋学 海洋気象学 (ADEOS-II Mission and Oceanography/Marine Meteorology)

二国間交流事業共同研究報告書 共同研究代表者所属 部局独立行政法人理化学研究所 創発物性科学研究センター

GRASS 入門 Introduction to GRASS GIS

BCR30AM-12LB. RJJ03G Rev I T(RMS) 30 A V DRM 600 V I FGT I, I RGT I, I RGT III 50 ma : PRSS0004ZE-A ( : TO-3P) 4 2, 4

誌上発表 1. 表面技術 特集 LED 照明と表面技術 61 巻 10 号 AlGaN 系殺菌用途紫外 LED の進展と今後の展望 平山秀樹 2010 年 10 月号.

Title 古典中国語 ( 漢文 ) の形態素解析とその応用 安岡, 孝一 ; ウィッテルン, クリスティアン ; 守岡, 知彦 ; 池田, 巧 ; 山崎, 直樹 ; 二階堂, 善弘 ; 鈴木, 慎吾 ; 師, 茂. Citation 情報処理学会論文誌 (2018), 59(2):

谷本俊郎博士の研究業績概要 谷本俊郎博士は これまで地球内部の大規模なマントルの対流運動を解明するための研究 および 大気 - 海洋 - 固体地球の相互作用に関する研究を様々な角度から進めてきた これらのうち主要な研究成果は 以下の様にまとめることができる

一般化川渡り問題について. 伊藤大雄 ( 京都大学 ) Joint work with Stefan Langerman (Univ. Libre de Bruxelles) 吉田悠一 ( 京都大学 ) 組合せゲーム パズルミニ研究集会

Report on the experiment of vibration measurement of Wire Brushes. mounted on hand held power tools ワイヤ ブラシ取付け時の手持動力工具振動測定調査の実施について

モータ用モデル予測電流制御における 予測用モータモデルの磁気特性表現の改善

外部電界によるグラフェン - 化学分子ファンデルワール ス相互作用制御と高機能素子応用. Citation 科学研究費助成事業研究成果報告書 :

Video Archive Contents Browsing Method based on News Structure Patterns

開催日 :2015 年 9 月 日会場 : 情報通信研究機構 ( 小金井 ) 主催 :SCOSTEP, WDS 共催 : 名大 STE 研 国立極地研 NICT SGEPSS

確率統計 特論 (Probability and Statistics)

I. Original. Lett., 39, 1382 (1977). Schneider, J. Chem. Phys., 68, 3401 (1978). Phys., 76, 5502(1982) (1985) (1987).

電子回路論第 12 回 Electric Circuits for Physicists

EDL analysis for "HAYABUSA" reentry and recovery operation はやぶさ カプセル帰還回収運用における EDL 解析

Evaluation of IGS Reprocessed Precise Ephemeris Applying the Analysis of the Japanese Domestic GPS Network Data

2018 年 ( 平成 30 年 ) 7 月 13 日 ( 金曜日 ) Fri July 13, 2018

CMB の温度 偏光揺らぎにおける弱い重力レンズ効果 並河俊弥 ( 東京大学 )

低温物質科学研究センター誌 : LTMセンター誌 (2013), 23: 22-26

Taking an advantage of innovations in science and technology to develop MHEWS

SOLID STATE PHYSICAL CHEMISTRY

結合および相互作用エネルギーの定量的 評価法の開発と新規典型元素化合物の構築

Ancient West Asian Civilization as a foundation of all modern civilization New Geological and Archaeological Academic Research in Japan

Illustrating SUSY breaking effects on various inflation models

研究業績リスト. 2) HOMO Energy Gap Dependence of Hole-Transfer Kinetics in DNA,

履歴書 文世一 1958 年 7 月 30 日生

マテリアルズインフォマティクスの最前線 吉田 亮. サイエンティフィック システム研究会 科学技術計算分科会 2017年度会合

Reactive Fluid Dynamics グローバル COE 環境共生 安全システムデザインの先導拠点 複雑システムのデザイン体系 第 2 回 力学系の非線形力学 理工学研究科, 開放環境科学専攻 教授植田利久 2009 年 4 月 16 日. Keio University. No.

むらの定量化について IEC-TC110 HHG2 への提案をベースに ソニー株式会社冨岡聡 フラットパネルディスプレイの人間工学シンポジウム

京都大学臨界集合体実験装置での加速器駆動システムにおけるウラン - 鉛領域炉心の中性子特性に関する実験ベンチマーク

The unification of gravity and electromagnetism.

Evaluation of Faults Near the Tsuruga Nuclear Power Plant

Parameter Estimation of Solar Radiation Pressure Torque of IKAROS

先端研究助成基金助成金 ( 最先端 次世代研究開発支援プログラム ) 実施状況報告書 ( 平成 25 年度 )

阪府 学 学院 学研究科松本章 研究業績リスト論 その他 ( )

井出哲博士の研究業績概要 井出哲博士はこれまで データ解析や数値シミュレーションの手法を用いることによって 地震の震源で起きている現象を様々な角度から研究してきた その主な研究成果は 以下の 3 つに大別される

16. Yamamoto, S. (in press). Non-reciprocal but peaceful fruit sharing in the wild bonobos of Wamba. Behaviour, in press.

C-H Activation in Total Synthesis Masayuki Tashiro (M1)

Method for making high-quality thin sections of native sulfur

EVERLIGHT ELECTRONICS CO.,LTD. Technical Data Sheet High Power LED 1W

Experimental and FE Analysis of Seismic Soil-Pile-Superstructure Interaction in Sand. Mahmoud N. HUSSIEN*, Tetsuo TOBITA and Susumu IAI

Review of Electrohydrodynamics in Corona Devices in Electrophotography. Kazuhiro Mori

生命科学科 3 年後期 分子細胞生物 I 相同アミノ酸配列の比較解析 藤博幸関西学院大学理工学部生命医化学科

Development of Advanced Simulation Methods for Solid Earth Simulations

TDK Lambda INSTRUCTION MANUAL. TDK Lambda A B 1/40

日本政府 ( 文部科学省 ) 奨学金留学生申請書

Title rays. Citation Advanced Materials Research, 409: 5. Issue Date Journal Article. Text version author.

研究会 ハイパー核物理の発展と今後の展望 2013年7月8日 和州閣(三重県志摩市)

統合シミュレーションコードによる高速点火実験解析大阪大学レーザーエネルギー学研究センター中村龍史

Study of Reliability Test Methods for Die-attach Joints on Power Semiconductors

テーマ名 ( 略称 ) : 沸騰 二相流体ループを用いた気液界面形成と熱伝達特性 (Two-Phase Flow)

平成 27 年度傾斜的研究費 ( 全学分 ) 研究報告書

Sources and further reading

Shingo NAKANISHI and Masamitsu OHNISHI

Hiroshi Ikeda a, *, Akiko Shimizu a, Makoto Ogawa b, Yasushi Ibaragi b and Shinobu Akiyama c : Lectotypification of Glaziocharis abei (Burmanniaceae)

原子核の弱電相互作用と超新星ニュートリノ

List of Publications Seiji Suga

田村直之 (Kavli IPMU) & PFS コラボレーション PFS 進捗報告 国立天文台三鷹 2015 年 9 月 14 日. Prime Focus Instrument PFI. Cobra engineering model module

研究期間全年度研究業績 熊谷直哉 ( 微生物化学研究所 A01 班公募班員 )

非弾性散乱を利用した不安定核 核構造研究 佐藤義輝東京工業大学

ROSE リポジトリいばらき ( 茨城大学学術情報リポジトリ )

Estimation of Gravel Size Distribution using Contact Time

超新星残骸からの陽子起源ガンマ線 放射スペクトルの変調機構

Listening Comprehension Sample 1

Day 5. A Gem of Combinatorics 組合わせ論の宝石. Proof of Dilworth s theorem Some Young diagram combinatorics ヤング図形の組合せ論

大原利明 算法点竄指南 点竄術 算 額 絵馬堂

[3] Tadafumi Adschiri: Supercritical Route for Super Hybrid Nanomaterials, ICMAT 2011, Suntec, Singapore,

Assessment of the SMT assemblies and Improvements through Accelerated testing methods. SMTA Chapter Meeting 18 th Jan 2014, India

EHP-A23/RGB33-P01/TR. Data Sheet. Materials. High Power LED 1W. 1 of 12 Release Date: :11:33.0 Expired Period: Forever

NIE を取り入れた教育実践研究 大学 2 年生を対象とした時事教養 Ⅰ の授業実践を通して. Educational Research based on Newspaper In Education (NIE)

テーマ名 ( 略称 ) : 沸騰 二相流体ループを用いた気液界面形成と熱伝達特性 (Two-Phase Flow)

川嶋辰彦教授略歴ならびに著作目録 1966 年 同大学大学院経済学研究科修士課程修了 1971 年 ペンシルバニア大学大学院博士課程修了. Ph. D. in Regional Science 取得 1972 年 同大学ウォートンスクール大学院地域科学部専任講師

車載用高効率燃焼圧センサー基板に最適なランガサイト型結晶の開発 結晶材料化学研究部門 シチズンホールディングス ( 株 )* 宇田聡 八百川律子 * Zhao Hengyu 前田健作 野澤純 藤原航三

10) Photophysicochemical Properties of Chlorophyll-a Adsorbed on Mg-containing Mesoporous Silica

3. Cautions for use 3-1. Voltage 3-2. Self heating 3-3. Mounting conditions 3-4. Storage conditions

Contents. Introduction Field Return Rate Estimation Process Used Parameters

Influence of MJO on Asian Climate and its Performance of JMA Monthly Forecast Model

J ap an. In t r o d u ct i o n. T h i s a n n o t a t ed b i b l i o g r a p h y c o l l e ct s al l m a t erial s a b o u t

( 一社 ) 日本応用地質学会 ( 公社 ) 日本地すべり学会 The Japan Society of Engineering Geology (JSEG) the Japan Landslide Society (JLS)

[ 粉体 砂丘の動力学 ] L.Guignier, H.Niiya, H. Nishimori, D. Lague and A. Valance: Sand dunes as migrating strings, Phys.Rev.E, Vol.87, (2013)

早稲田大学先進理工学部応用物理学科長谷川剛研究室 大学院先進理工学研究科物理及応用物理学専攻 大学院先進理工学研究科ナノ理工学専攻

EVERLIGHT ELECTRONICS CO.,LTD. Technical Data Sheet High Power LED 1W (Preliminary)

新技術説明会 ラマン分光 必見! AFM- ラマンによるナノイメージの世界 株式会社堀場製作所

9.Mistakes occur no matter ( ) careful you try to be. ( ア )how ( イ )what ( ウ )with ( エ )on

1. Packaging Outline Dimensions Specifications ) Absolute Maximum Ratings (Ta=25 C)... 4

京都大学臨界集合体実験装置における 100 MeV 陽子を用いた加速器駆動システムの固体 Pb-Bi の中性子特性に関する実験ベンチマーク

2011/12/25

Introduction to Multi-hazard Risk-based Early Warning System in Japan

EVERLIGHT ELECTRONICS CO.,LTD.

RQ1A070AP. V DSS -12V R DS(on) (Max.) 14mW I D -7A P D 1.5W. Pch -12V -7A Power MOSFET. Datasheet 外観図 内部回路図 特長 1) 低オン抵抗 2) ゲート保護ダイオード内蔵

Transcription:

Technical report Approaches to Life Estimation of Electronic Circuits Akira MOTOYAMA Representative, M.A. Reliability Technology Office Abstract Concerning the short-life problem of electronic application products in these days, methods of life estimation for an electronics circuit is explained with examples. As a starter, steps for life estimation are illustrated. First, allowable cumulative failure probability against operating time that customers expect, i.e. target value of life, is set. Next, the electronic components that underperform the target values are picked up without experiment, among those mounted on the electronic circuit. Possessing acceleration formulas for each component are used for it. For the components picked up as have short life, accelerated life test is performed to estimate the life. This time, taking film capacitors as an example, life estimation of component is explained by the Eyring model formula. This formula is composed of ambient temperature, humidity and voltage. In the case of film capacitors, the capacitance decrease by anodic oxidation of the electrode by water due to moisture ingress and an applied voltage. Finally, as a result of life estimation of picked-up electronic components, solder joint parts and boards, the target value is cleared after changing carbon film resistors that do not meet the target value, to metal glaze resistors. 1. Introduction Many electronic components and mechanical components mounted on the circuit board in an electronic circuit. To ensure the target life of the electronic circuit, it is necessary to extract and change the components and joint parts to become a short life. In this paper, after showing a life estimation step of the electronic circuit, we introduce the method of it with an example. 2. Procedure for life estimation of electronic circuit Figure 1 shows a procedure for the life estimation. Set a target value of life and pick up operating conditions of electronic circuits and components. Estimate life of electronic components, then. Though life of every component and part has to be estimated, it is practically impossible because number of components of an electronic circuit is too large. So, pick up short life components in advance on a desktop, run accelerated tests of the picked up components only, and estimate the lives. Then estimate the life as an electronic circuit by combining estimated lives from accelerated test results of joints (other than components) and PCBs, with those estimated above. Take countermeasures for components and parts that do not meet the target value of life. Test Navi Report No.20 (87th) 1

Cum. Failure prob. ESPEC Technical Information 1. Set the target value of life (Set the life expectancy and the allowable cumulative failure probability.) 2. Pick up operating conditions 3. Pick up short-life components 4. Estimate life of short-life components 5. Estimate life of the circuit board assembly (1) Set ambient temperature/humidity of the products and the components in the field. ( (2) Set electrical stresses (voltage/current/frequency) ) applied to the components. (1) Prepare a list of stresses to the components mounted on the target circuit. ( (2) Pick up short-life components based on equations for ) life estimation obtained in advance. (1) Apply accelerated test to the picket up components per failure mode. ( (2) Estimate time to failure and cumulative failure ) probability of the picked up components. (1) Apply accelerated tests to joint areas other than the components. ( ) (2) Apply accelerated tests to the PCB. 6. Verify the target value of life Figure 1. Procedure for life estimation of electronic circuit 3. Example of life estimation of electronic circuit (1) Set the target value of life (1) Estimate the life of the electronic circuit from the accelerated test results. ( (2) Pick up components and parts that do not meet the ) target values of life. A target value of life is usually decided based on the consideration of life expectancy by customer and cost for fixing the trouble. Here, as an example, we set the target value of life as 10 years of operation (cumulative with slightly longer than a half day operation per day), and the allowable cumulative failure probability as 0.1%. By the way, JEITA sets 0.1% as the allowable cumulative failure probability of semiconductors at useful life. Define the cumulative failure probability as a ratio of a number of the products returned as a failure so far and a number of the products inputted to the field. The cumulative failure probability vs. operating time in the field is shown in Figure 2. Assuming the relation as the Weibull distribution, it can be expressed by linear regression each of the three periods: early failure, random failure, and wear-out failure period. Here the target value of life is defined as a time when failure mode classified as wear-out failure starts to occur. Life expectancy Random failure Wear-out failure Early failure Target value of life 0.1% Operating time (h) 50,000 Figure 2. Set the target value of life Test Navi Report No.20 (87th) 2

(2) Pick up operating conditions Assume that the ambient temperature, operating time of an electronic circuit and ambient temperature of the components mounted in the product per year are shown in Table 1, after investigating the ambient temperature/humidity of a product/component and operating time in the field. Table 1. Operating environment and time of a product and target components in a year Ambient temp. of the product 5 C 10 C 15 C 20 C 25 C 30 C Ambient temp. of components 54 C 61 C 69 C 76 C 82 C 86 C Operating time per year (h) 400 500 800 1200 700 400 Information on electrical stresses such as voltages, currents and frequencies applied to the components is also required as a part of operating conditions. (3) Pick up short-life components Understand which types of the components are mounted in the electronic circuit so that short life components can be picked up. Mainly for passive components (L, C, R) assumed to have relatively shorter lives, summarize the manufacturer, product no., rating, and operating environment (such as temp-humidity, electrical stress) of each component and create a list of operating stresses per component. Taking a film capacitor as an example, electrode structure, voltage (rating and operating), current (rating and operating), and temp-humidity in the field are summarized per circuit code in Table 2 List of operating stresses of each component. Table 2. List of operating stresses of each component Carbon film resistors Transformers and coils Al electrolytic capacitors Film capacitors Code Electrode structure (type) Voltage (V) Current (A) Rating Opr. Rating Opr. Temp Hmdty Metalized Metalized Metalized Foil Foil Next, pick up short-life components by using life estimation equations of electronic components/parts owned by analysis service companies such as M.A Reliability Technical Office, and by using reliability test data of the component manufacturer. As an example, let s estimate the life of a metalized film capacitor of circuit code C12 in Table 2, under the temp-humidity conditions in the field. Test Navi Report No.20 (87th) 3

In general, estimated life against temp-humidity can be given by the Eyring model formula (1) below. L = L 0 exp(a1 (1/(273.15 + T) - 1/(273.15 + T 0 )) + a2 (RH - RH 0 )) (1) where L is median life (h) at ambient temperature T ( C), relative humidity RH (%) in the field, L 0 is median life (h) at ambient temperature T 0 ( C), relative humidity RH 0 (%) in the accelerated test, a1 is a temperature acceleration constant and a2 is a humidity acceleration constant. Then the operating condition of C12 shown in Table 2, i.e. ambient temperature 75 C and relative humidity 9%, and the reliability test condition of electronic component manufacturer, i.e. ambient temperature 60 C and relative humidity 90%, are entered into equation (1) to obtain equation (2). L = L 0 exp(a1 (1/(273.15 + 75) - 1/(273.15 + 60)) + a2 (9-90)) (2) For the two constant terms (a1, a2), use two C12 specific values at hand, and obtain C12 median life for the operating condition. Because this failure distribution is considered as a Weibull distribution, use the median life L obtained in the above and a form parameter m at hand in the equation (3) to get estimated life t = 52,800 hours at the cumulative failure probability 0.1%. ln(ln(1/(1-0.1%))) = m ln(t) + (-0.3665 - m ln(l)) (3) In this way, using life estimation equation of each component and data from electronic component manufacturers, estimate life against stresses (temperature, temp-humidity, and temp. cycle) corresponding to failure mode of each component. Here, seven components are picked up that may have short life based on the calculation result as an example. The result is shown in Table 3. As we have seen so far, without actual accelerated tests and by using proprietary data owned by data analysis service companies, life can be estimated in a short time on a desktop and short-life components can be picked up. Test Navi Report No.20 (87th) 4

Table 3. Pick up of short-life components by life estimation in advance Component name Al electrolytic capacitor Circ. Code C1 C2 Target life Operating time ON/OFF cycles - - Estimated lif e (Cum. failure prob.: 0.1%) Temp. 34,900h 57,300h Temp /Hmdy Temp. cycles No mode No mode C10 7,000 55,500h 54,600h Unknown Film capacitor C11 7,000 151,000h 57,700h Unknown C12 7,000 105,300h 52,800h 7,920 Trans. T1 Carbon film resistor R1 7,000-45,900h 34,700h No mode Fields of red number do not meet the target values and need the component evaluation without exception. Fields of blue number clear the target values but less than two times and need the component evaluation. According to Table 3, aluminum electrolytic capacitor C1 Trans. T1 and carbon film resistor R1 do not clear the target life of 50,000 hours, and need life estimation by accelerated tests. And other four components as well do not have enough margins and need life estimation by accelerated tests. As such, we need to have a life estimation equation in advance to pick up short-life components. In case the life estimation equation is not available, we may refer to the component manufacturer s test data (number of samples, characteristic values (such as capacitance change rate for a capacitor), failure criteria, etc.). Also we may pick up components operating at conditions closer to the component manufacturer s tolerances (temp-humidity, voltage, current, etc.), i.e. components with smaller derating, as short-life components. (4) Estimate life of short-life components Accelerated test per failure mode is required to have more precise life estimation of a short-life component. Test levels should be 3 or more here. Test items, test levels, failure criteria for short-life components picked up in the previous section are shown in Figure 4. For temp-humidity test, total 5 levels are required since 3 levels each are set for temperature and humidity. Test Navi Report No.20 (87th) 5

Table 4. Accelerated test items per component Component name Circ. code Test item Test level Failure criteria Al electrolytic capacitor C1 C2 Temperature 3 levels Ta: a C, b C, c C C/C <-a% Tanδ > spec x 2 C10 Temperature 3 levels Ta: a C, b C, c C C/C <-a% Tanδ > spec x 2 Film capacitor C10 C11 C12 Temp /Hmdty 5 levels Ta: a C, b C, c C RH: a%, b%, c% C/C <-a% Tanδ > spec x 2 C10 C12 Temp. cycles 3 levels T: a C, b C, c C Pulse tol. < aaop Tanδ > spec x 2 Trans. T1 Temperature 3 levels Ta: a C, b C, c C Layer short voltage < akv Carbon film resistor R1 Temp /Hmdy 5 levels Ta: a C, b C, c C RH: a%, b%, c% R/R >a% Obtain the Eyring model formula (1) in the above by the temp-humidity accelerated test, and obtain the acceleration constant a1 of the Arrhenius model formula (4) below, to estimate life in operating conditions. L = L 0 exp(a1 (1/(273.15 + T) - 1/(273.15 + T 0 ))) (4) where L is median life (h) at ambient temperature T ( C) in the field, L 0 is median life (h) at ambient temperature T 0 ( C) in the accelerated test, a1 is a temperature acceleration constant. That is, different from estimating life on a desktop, life is estimated by obtaining the constant (a1) of the Arrhenius model equation (4) and the constants (a1, a2) of the Eyring model formula (1). As an example, result of temp-humidity test of the film capacitor C12 based on the Eyring model is given below. Accelerated tests of 3 temperature levels (65, 75, and 85 C at humidity 85%) and 3 humidity levels (75, 80, and 85% at temperature 85 C), giving actual 5 levels, are conducted. From the test results, the constants (a1, a2) are obtained, giving life estimation equation (5) below. L = L 0 exp(10084 (1/(273.15 + 75) - 1/(273.15 + 85)) - 0.0805 (9-85)) (5) ln(ln(1/(1-0.1%))) = m ln(t) + (-0.3665 m ln(l)) (6) By getting median life L from equation (5) and performing Weibull analysis with equation (6), we finally get 57,295 hours as the estimated value of life for the operating conditions in the field at the cumulative failure probability 0.1%. Test Navi Report No.20 (87th) 6

Estimated lives of all the short-life components obtained by this procedure are shown in Table 5. The result reveals that the carbon film resistor among the 7 components picked up by the desktop life estimation, is the only one that cannot clear the target value. Table 5. Life estimation by accelerated testing (5) Estimate life of a circuit board assembly To estimate life of a circuit board assembly, select and perform accelerated tests (corresponding to each failure mode) to solders connecting between components and the PCB and to the PCB itself, Using the result, estimate the time to failure against the target cumulative failure probability 0.1%. Based on this approach, life of solder joint part can be estimated by temperature cycle test and using Modified Coffin-Manson formula (7). N = N 0 (f/f 0 )^a1 ( T 0 / T)^a2 exp(a3*(1/(273.15 + T) - 1/(273.15 + T 0 ))) (7) where N is the median life (cycle) in actual operation, N 0 is the median life (cycle) in the accelerated test, f is the actual operating period (cycles/day), f 0 is the test period (cycles/day), T is the temperature difference in actual operation, T 0 is the temperature difference in the accelerated test, T is the upper limit temperature ( C) in actual operation, T 0 is the upper limit temperature ( C) in the accelerated test, a1 is the acceleration constant of the period, a2 is the acceleration constant of the temperature difference, and a3 is the acceleration constant of the temperature. For the PCB, perform temp-humidity test as an accelerated test for migration and estimate life using the Eyring model formula (1) above. Test Navi Report No.20 (87th) 7

Cum. failure prob. ESPEC Technical Information (6) Verify the target value of life As we have discussed so far, life of short-life component can be estimated by the accelerated tests in Section (4), and life of joint part other than components and of PCB can be estimated by the accelerated tests in Section (5). Here, we use the results of those accelerated tests to estimate life of the electronic circuit. We also verify if the target value of life of the electronic circuit (50,000 hours at cumulative failure probability 0.1%) can be cleared. If we plot test results of short-life components/parts and the target value of life on a same Weibull probability sheet, we get Figure 3. Short-lif e components/parts R1 C1 C2 C12 T1 Target value of life 0.1% 50,000 Operating time (h) Figure 3. Weibull analysis of short-life components/parts and verification of the target value of life On the other hand, life of an electronic circuit can be estimated by obtaining the cumulative failure probability F(t) from equation (8). F(t) = 1 (R1(t)) R2(t) Rn(t) (8) where R1(t) to Rn(t) are reliabilities of each component, part and board at time t. If the value F(50,000) can be plotted below the target value of life, we have verified that the target can be cleared. For the picked up components/parts so far that do not meet the target value of life, investigate collectively by referencing Figure 3 and others whether to change them or not, then decide. In case of this example, the target has been cleared by changing the resistor R1, which does not meet the target value, from carbon film resistor to metal glaze resistor, of different material. As is seen in Figure 3, slope of a line (form parameter) varies per component/part. Referencing to Figure 3 is effective not only to verify the target value of life but also to guess increasing trend of cumulative failure probability of an electronic circuit when its operating time is extended beyond the target value of life. Test Navi Report No.20 (87th) 8

REFERENCES (1) 故障物性研究会編著 : 新しい信頼性技術と管理手法, pp.112-134, 3.2.5 実際の加速試験による寿命予測, 日刊工業新聞社, 2011 Society for the Study of Failure Physic (ed.), 3.2.5 Life Estimation by Actual Accelerated Tests, New Reliability Technologies and Management Methods, Nikkan Kogyo Shimbun Ltd., 2011, pp.112-134 (in Japanese) (2) 本山晃 : 信頼性 安全問題の実態と解析技術, 第 41 回信頼性 保全性シンポジウム報文集,pp.89-99, 日科技連,2011 Akira MOTOYAMA, Realities of Reliability and Safety Problems, and Analysis Technologies, Proceedings of the 41st Reliability and Maintainability Symposium, Japanese Union of Scientists and Engineers, 2011, pp.89-99 (in Japanese) (3) 本山晃 : 高信頼性を実現するセットメーカの試験 解析, 第 40 回信頼性 保全性シンポジウム報文集,pp.109-117, 日科技連,2010 Akira MOTOYAMA, Tests and Analyses by an Assembled Product Manufacturer Who Attains High Reliability, Proceedings of the 40th Reliability and Maintainability Symposium, Japanese Union of Scientists and Engineers, 2010, pp.109-117 (in Japanese) (4) 松岡敏成 本山晃 : 電子部品の信頼性試験ガイド ( 第 6 回 ), 日本信頼性学会誌 信頼性基礎講座, 日本信頼性学会,2009,3 月号 Toshinari MATSUOKA and Akira MOTYAMA, Guide to Reliability Tests of Electronic Components (No.6), The Journal of Reliability Engineering Association of Japan-Reliability Basic Course, Reliability Engineering Association of Japan, Mar., 2009 (in Japanese) (5) 大野浩一 本山晃 : 鉛フリーはんだの熱疲労寿命に関する一検討, Mate2002,p333-338,2002 Kouichi OONO and Akira MOTOYAMA, An Investigation on Thermal Fatigue Life of Lead-Free Solders, Mate2002, 2002, pp.333-338 (in Japanese) (6) 本山晃 : フィルムコンデンサの耐湿性に関する一検討, 第 36 回信頼性 保全性シンポジウム報文集, pp.259-262, 日科技連,2006 Akira MOTOYAMA, An Investigation on Humidity Resistance of Film Capacitors, Proceedings of the 36th Reliability and Maintainability Symposium, Japanese Union of Scientists and Engineers, 2006, pp.259-262 (in Japanese) Test Navi Report No.20 (87th) 9