PowerPoint 簡報

Size: px

Start display at page:

Download "PowerPoint 簡報"

涛午权
5 years ago
Views:

1 Chapter 8 Itroducto of Relablty for Electroc Packagg

2 Geeral Relablty Testg for Electroc Package

3 Purposes Ths chapter descrbes a qualfcato methodology for semcoductor devces based o ed-use applcato evromet ad falure deas The tet of ths methodology s to provde a qualfcato framework that s geared toward the ed-use applcato ad rsk level a suppler ad ed-user s wellg to accept The methodology uses Stadard evromet & applcato from whch accelerato factors, sample sze, test durato, ad pass/fal crtera are calculated. 3

4 Idustry-wde Relablty Gudele Relablty requremets establshed by the Electroc Idustres Assocato (EIA) ad Jot Electro Devce Egeerg Coucl (JEDEC). Tests are performed uder extreme codtos usg equatos that descrbe the accelerato of falure mechasms. The tests used fro package relablty testg, cludg Precodto Temperature cycle Thermal shock PCT (pressure cook test) / u-bas HAST Temperature humdty Hgh temperature storage 4

5 Idustry-wde Relablty Gudele Item Stadard Codto Crtera Precodto.66JEDEC-STD-NO.- A3A (LEVEL 3) Bake 5 / 4 hrs30 / 60%RH/ 9 hrsir Reflow 0 / 3 cycles O/S pass,sat Temperature cycle MIL-STD-883E,Method 00.7,Codto C -65 / 50, Ar to ar O/S pass Thermal shock MIL-STD-883E,Method 0.9,Codto C -65 / 50, Lqud to lqud500,000cycles O/S pass Pressure cook test(wth hast ove) JEDEC-STD-NO.-A0-A, 00%R.H, 30PSI O/S pass Temperature Humdty JEDEC-STD-NO.-A08, (A0-A) 85 /85%RH500, 000 hrs. O/S pass Hgh temperature storage MIL-STD-883C,Method O/S pass 5

6 Mosture Effect 6

7 Package Reflow Process 7

Reflow Sestvty for Small Package The vapor pressure of mosture trapped sde of the package creases whe the package s exposed to hgh solder reflow temperature.

8 Reflow Sestvty for Small Package The vapor pressure of mosture trapped sde of the package creases whe the package s exposed to hgh solder reflow temperature. Ths pressure ca lead to teral delamato, cracks, bod damage, wre eckg, bod lftg, de lftg, th flm crackg. The most sgfcat s popcor effect Small, th package ca reach body temperatures greater tha 0C whe reflowsoldered to boards. To compesate for ths dfferece, some small packages have bee reclassfed to wthstad 35C 8

9 Classfcato Reflow Profles Package relablty ca also affected by the rate at whch a partcular devce reaches ts testg temperature. The reflow covecto methods: full covecto ad fra-red covecto 9

10 Falure Modes Ispecto 0

11 Scag acoustc mcroscope: SAM

12 Scag acoustc mcroscope: SAM

13 Temperature Cyclg Testg Temperature cyclg (JESD-A04-C) s performed to determe the resstace of packages to hgh ad low temperature extremes. The evromet stress test s desged to smulate the extesve chages temperature to whch packages are exposed. To pass the test, packages must show the sgs of damage such as crackg, chppg or breakg. 3

14 JEDEC Stadard for Temperature Cyclg Profle 4

15 JEDEC Stadard for Thermal Cyclg Test 5

16 CTE Msmatch Effect Temperature cyclg may expose weakesses packages that are composed of dfferet materals Expaso ad cotracto of package materals (CTE) ca occurs at dfferet rates whe exposed to temperature extremes. 6

Crack Itato ad Propagato Whe packages are repeatedly cycled betwee hot ad cold temperature extremes, the shear stress of the materals chage from compressve to tesle ad back.

17 Crack Itato ad Propagato Whe packages are repeatedly cycled betwee hot ad cold temperature extremes, the shear stress of the materals chage from compressve to tesle ad back. If the temperature extremes are great eough, crackg ad /or shear dsplacemet of the materals could occur, resultg dsplacemet of the metal traces below the packagg layers. 7

Thermal Shock Test (TST) Thermal Shock s performed to determe the resstace of the part to sudde chages temperature.

The parts are the exposed to a extremely low (or hgh) temperature ad, wth a short perod of tme, exposed to a extremely

Falure accelerato due to Thermal Shock ad Temp Cyclg depeds o the followg factors: ) the dfferece betwee the hgh ad low

18 Thermal Shock Test (TST) Thermal Shock s performed to determe the resstace of the part to sudde chages temperature. The parts udergo a specfed umber of cycles, whch start at ambet temperature. The parts are the exposed to a extremely low (or hgh) temperature ad, wth a short perod of tme, exposed to a extremely hgh (or low) temperature, before gog back to ambet temperature. Falure accelerato due to Thermal Shock ad Temp Cyclg depeds o the followg factors: ) the dfferece betwee the hgh ad low temperatures used; ) the trasfer tme betwee the two temperatures; ad 3) the dwell tmes at the extreme temperatures. Falure mechasms accelerated by thermal shock clude de crackg, package crackg, eck/heel/wre breaks, ad bod lftg. 8

Pressure Cooker Test (PCT) Autoclave Test, or Pressure Cooker Test (PCT), or Pressure Pot Test (PPOT), s a relablty test performed to assess the ablty of a product to wthstad severe temperature ad

19 Pressure Cooker Test (PCT) Autoclave Test, or Pressure Cooker Test (PCT), or Pressure Pot Test (PPOT), s a relablty test performed to assess the ablty of a product to wthstad severe temperature ad humdty codtos. It s used prmarly to accelerate corroso the metal parts of the product, cludg the metallzato areas o the surface of the de. It also subjects the samples to the hgh vapor pressure geerated sde the autoclave chamber. Autoclave testg cossts of soakg the samples for 68 hours at deg C, 00% RH, ad atm. Itermedate readpots at 48H ad 96H may also be employed Because of the extreme mosture codto durg autoclave testg, mosture-related electrcal leakage falures may be ecoutered after the test. Uless predefed to be so, such falures must ot be cosdered as vald PCT falures. Oly permaet falures, such as those arsg from corroso, are cosdered vald PCT falures. 9

Temperature Humdty Temperature, Humdty, Bas (THB) testg s a relablty test desged to accelerate metal corroso, partcularly that of the metallzatos o the de surface of the devce.

20 Temperature Humdty Temperature, Humdty, Bas (THB) testg s a relablty test desged to accelerate metal corroso, partcularly that of the metallzatos o the de surface of the devce. Asde from temperature ad humdty whch are eough to promote corroso of metals the presece of cotamats, bas s appled to the devce to provde the potetal dffereces eeded to trgger the corroso process, as well as to drve moble cotamats to areas of cocetrato o the de. THB testg employs the followg stress codtos: 000 hours at 85 deg C, 85% RH, wth bas appled to the devce. The bas appled s usually desged to smulate the bas codtos of the devce ts real-lfe applcato, maxmzg varatos the potetal levels of the dfferet metallzato areas o the de as much as possble. The ma drawback of THB s ts log durato, ecesstatg weeks before useable data are obtaed. Because of ths, a alteratve test, the HAST, has bee developed. HAST uses more severe stress codtos but ca be completed oly 96 hours. Its shorter durato s a advatage that makes t a more popular stress test the dustry. 0

21 Hgh-Accelerated Temperature ad Humdty Test (Bas-HAST)

Hgh Temperature Storage (HTS) The Hgh-Temperature Storage (HTS) test s performed to determe the effect o devces of log-term storage at elevated temperatures wthout ay electrcal stresses appled.

22 Hgh Temperature Storage (HTS) The Hgh-Temperature Storage (HTS) test s performed to determe the effect o devces of log-term storage at elevated temperatures wthout ay electrcal stresses appled. HTS s smlar to Stablzato Bake (Ml-Std-883 Method 008), except that HTS s doe over a much loger perod of tme (000 hours vs. 4 hours for Stab Bake at 50 deg C). The purpose of HTS s to assess the log-term relablty of devces uder hgh temperature codtos whle that of Stablzato bake s merely to serve as part of a screeg sequece or as a precodtog treatmet pror to the coduct of other tests.

23 Relablty Aalyss for Electroc Package 3

24 Bathtub Curve Ut: ppm/khrs Early Falure Rate (EFR) Wear-out Falure Rate (WFR ) Stable Falure Rate (SFR ) 4

25 Early Falure Aalyss Accelerated Agg Test Bur- EFR Weak Compoets Arrheus Reacto-rate Formula A, j E exp K j T T 0 IC Defects IC R Y EFR C Y S Y I Y D Y M /3 ~ 4/5 60 5

26 Amout of Gross Des 6 50 mm 8 00 mm 300 mm 8 00 IC IC N c π [ φ ( + θ ) A / θ ] 4A c c A c xy θ x / y y x 6

27 Example 8 x cm y 0.8 cm θ x / y / A c xy 80 mm N c π [ φ ( + θ ) A / θ ] c 4A c

28 Yeld Rate of KGD Y ( D) AD e f dd 0 Half-Gaussa Dstrbuto f ( D) πd 0 exp π D D 0 Y a + 3 t + at a3t f(d)/d 0 Good wthout Defects N π [ φ ( + θ ) Ac / θ ] ( 3 a t + a t a t ) y + 4Ac D/D 0 8

29 Example AD x cm y 0.5 cm θ x / y / 0.5 A c xy 50 φ ( + θ ) N t c π mm [ A / θ ] c 4A + 0.4AD 0 c Y a t + at + a3t % N y ( 3 a t + a t + a ) N t c 3 9

30 Yeld Rate of Packagg Fuctoal Test Full-speed Test DC Parameter Test AC Parameter Test Vsual Test Wafer Map Kow Good De, KGD KGD Y CS ( ) ( ξ ) % 00Y Y CS ( ) ( 3 + ) % t 0.t 0.75t ξ Y C ( ) ( 3 ) ( ξ ) NC % t 0.t t 30

31 Example AD x cm y 0.5 cm ξ 99.9% θ 50 A c mm t Y N c 75 Y C 3 ( ) ( ) ( ξ % ) C t t + t N 96.06% 95% 90% 99% 95% Bur- R EFR % 3

32 Stable Falure Aalyss Probablty Desty Fucto, PDF Cumulatve Dstrbuto Fucto, CDF 0 0 f ( x) 0 F( x) f ( x) dx ( x) f df( x) dx SFR CDF SFR CDF 3

33 Falure Rate vs. Relablty c F(x) R(x) R ( x) F( x) c 0 0 a 0 0 a F(x) a b bxc R(x) R(x) a b bxc 33

34 Istataeous ad Average Falure Rate Istataeous Falure Rate ( x) h f R ( x) ( x) h(x) SFR H ( x) l R( x) x x AFR ( x, x ) l R ( x ) l R( x ) H ( x ) H ( x ) x x x x 34

35 Theoretcal Model of Falure Lfe Expoetal Fucto PDF f ( ) x x λe λ CDF F ( ) λx x e x hours or days λ /hr or /day 35

36 Example F ( ) λx x e λ /day 00 R ( 00 0) F( 00) e R ( ) F( 00) e

37 Theoretcal Model of Falure Lfe PDF f ( x) exp σ π x µ σ CDF F ς µ σ x ( x) exp dς σ π µ x σ PDF exp π z z ( ) f z F ( x) CDF π ( x µ ) / σ e z / dz 37

38 Theoretcal Model of Falure Lfe PDF : f ( x) exp ( lς T ) σ π x l σ 50 CDF F l σ x ( x) exp ( lς T ) dς σ π x 50 µ lt 50 T50 50% Shape Parameter z PDF l x lt z 50 f ( z) exp F( x) σ π CDF π ( l x lt ) / σ e z 50 / dz 38

39 Example F ( x) π ( l x l T ) / σ e z 50 / dz T σ z l x lt50 σ l 000 l F π.309 z / ( 000) e dz

40 40 Theoretcal Theoretcal Model of Falure Lfe Model of Falure Lfe Webull Fucto Webull Webull PDF ( ) ( )( ) ( ) β θ β θ β / / / x e x x x f Webull CDF ( ) ( ) β /θ x e x F x β Webull Slope θ Characterstc Lfe ( ) X X Y X Y X β ( ) ( ) X X Y X X Y X C β θ e C / C X G Y + β l l ( ) x X l

41 Theoretcal Model of Falure Lfe G Johso s Theory ( z) z( z) ( ) ( z) L! z ( ) L( j + ) ( j )! z j+ ( z) G j j z j 50% G ( 50% ) j

42 Example x hours 30 [] ( ) ( ) ( ) G 50 % j.3 / +. 4 X Y X XY

43 Example 30 β C X ( X ) Y X ( X ) Y ( X ) X X Y X ( 30)( 7.43) ( 7.45)( 6.56) ( 30)( ) 7.45 X Y.68 ( )( 6.56) ( 7.45)( 7.43) ( 30)( ) θ / 0.3 /.68 e C β e 43 hrs F β ( x / θ ) ( x / 43) ( x) e e

44 Accelerato Factor (AF) Accelerato Factor x o AFx t x 0 x t AF PDF CDF f o AF o ( x ) f o t x AF F o o ( x ) F o t x AF AF AF f f o t m T T t o e ( / / ) 44 T o T t 44

45 Example PQFP -40 ~ cycle/hr 0 ~ cycle/day /4 cycle/hr Webull β 3.66 θ 6357 AF m /3 AF / e ( / 85 / ). F R o ( /44) ( ) x o x e o ( x /44) ( x ) o o e h o ( x ) o xo

1 1 68.% 1% 67% 6 1, [198] [199] [1986] [199] IoveIwasa Matsuo [1987]

1 1 68.% 1% 67% 6 1, [198] [199] [1986] [199] IoveIwasa Matsuo [1987] A Study o Applyg Physographc Iudato Model to Rafall - Ruoff Smulato Jug-Fu Ye Chag-Ju Yag Chag-Ta Tsa HEC-1 HEC-1,, ABSTRACT Tawa has lmted lad wth dese populato, so most watersheds are over-developed.