当前位置:首页 >> 能源/化工 >>

BGA Underfill Performance - Reliability


Underfill Experience and Reliability

Underfill Application Examples

Cellphone
Parts with red arrows are underfilled

Optics Module

Celestica C

onfidential

2

Underfill
BGA are typically not underfilled
? Mechanical stress (vibration, impact) usually most important for BGA

applications
? Larger solder ball size and higher standoff already increases reliability ? More difficult to rework a component with underfill ? May not be able to clip on a heat sink

Flip Chips, CSPs often are underfilled
? Flip chip reliability is very bad without underfilling ? Fine pitch CSP often fail very quickly without underfilling ? Underfill couples chip to substrate, relieves stress/strain from joints ? CSP normally looking for mechanical improvement, sometimes

thermal cycling
Celestica Confidential 3

Underfill Selection Considerations
? Non-reworkable and reworkable formulations available ? Selection depends on understanding system requirements
? Underfills used to either improve long term reliability and/or increase
mechanical strength of interconnect
Type
CTE (ppm/C)

Pros

Cons

? Designed for applications where

Nonreworkable (ie. Silica filled)

20-30

high CTE mismatch between component and laminate it is mounted to exists (ex. Flip chip, WS-CSP) ? Low ionic content, high adhesion, low CTE, high Tg
? Improved flow speed, cure time ? Lower cost ? Improves shock/drop performance

? Higher propensity for voiding

? Substrate must be heated to >

800C for guarantee good flow

Reworkable

50-70

? Less mechanical strength ? Poorer ATC performance

Celestica Confidential

4

Underfill Material Experience
Material
Loctite 3513* (reworkable) Loctite 3549 (reworkable) Zymet X6-82-5LV50* Emerson & Cuming* Stycast A312-20 Namics Ohmcoat 1572 Namics 8437 Hysol FP4546 Zymet 1432 (reworkable) Zymet CN-1453 (reworkable)
* - Current Production Material Celestica Confidential 5

Product
Bluetooth Headset Test Vehicle Test Vehicle Optical Card Experiment Cellphone Cellphone Test Vehicle Test Vehicle Test Vehicle Test Vehicle Pre-production

Carrier
FR4 FR4 Rigid-Flex FR4 FR4 FR4 MCM-SLC FR4 FR4 FR4 FR4

Component
CSP CBGA CSP Connector BGA CSP CSP BGA Flip Chip Flip Chip FCA CSP CSP (Pb-free)

Pitch (mm)
0.8 1.0 0.8 1.27 0.8 0.8 1.27 0.2 0.25 0.2 0.25 0.25 0.8 0.75

Underfill Design Considerations
? Silkscreening can interfere with underfill
spreading

? Via proximity can cause underfill material to
“leak” to bottomside of board

? Need local fiducials to facilitate machine
programming

? Reworkable material available however will
traditional rework tools require modifications
? Material must be heated to soften, then part
torqued once solder melted
Celestica Confidential 6

Underfill ATC Reliability Testing
? Reliability testing of underfills has been performed on packages ranging

from FCAs, WS-CSPs, PBGAs to CBGAs
? Both reworkable and non-reworkable formulations evaluated ? Reliability of reworkable materials currently inferior to non-reworkable

formulations
? Reworkable underfills primarily intended to enhance shock resistance instead

of ATC performance.
? To facilitate rework, reworkable underfills lack fillers and therefore have higher

CTE value (typ. >50 ppm/C). This results in large CTE mismatch between
underfill material and the underlying solder / laminate system resulting in poor ATC performance.
? Non-reworkable underfills are specifically designed to improve ATC

performance and have CTE’s closer to the materials they are bonded to.
Celestica Confidential 7

FCA Reliability Results
? Due to high CTE mismatch between silicon chip and laminate carrier substrate, only non-reworkable underfills recommended for this application ? Non-reworkable underfills significantly improve ATC performance ? Sample data for 3599 I/O FCA device
?Bump Diam:120mm Bump height 65 mm Bump number: 3599

UF b h R(6000) T(50%)

Hysol FP 4549 1.07 8689 51% 6098
Weibull Plot
c umu la tive pe rc e nt
9 9 .9 99 90 70 50 30 20 10 5 1 0 .5 0 .1 10 100 1000 10000 100000 E st . : M L E S h ap e: 1 . 1 2 3 3 9 S cal e: 8 2 0 5 . 0 5 O ri g i n : 0 . 0 F ail u res: 1 7 S am p l e si ze: 3 5

Hysol FP 4549 1.23 6867 43% 5098
W e ibu ll Plot
99. 9 99 90 70 50 30 20 10 5 1 0. 5 0. 1 10 100 1000 10000 100000 Est . : MLE Shape: 1. 22892 Scal e: 6867. 7 Ori gi n: 0. 0 Fai l ur es : 5 Sam pl e si ze: 9

Zymet X6-82-5LV 2.91 5623 30% 4958
W eibull Plot
c umulative pe rc ent
9 9 .9 99 90 70 50 30 20 10 5 1 0 .5 0 .1 100 1 0 00 1 0 00 0 1 0 00 0 0 E st .: M L E S h ap e: 2 . 9 0 9 2 8 S cale: 5 6 2 3 . 7 4 O ri g i n : 0 . 0 F ai lu res: 7 S amp l e si ze: 1 0

Zymet X6-82-5LV 7.31 8322 89% 7360

nt rce cu m u lativ e p e

C ol_1

C ol_1

Col_1

Celestica Confidential

8

WS-CSP / PBGA Reliability Study
?30 Boards assembled as part of HDPug consortia WS-CSP project ?13 Boards underfilled to provide reliability comparison with nonunderfilled boards (non-reworkable formulation: Namics Ohmcoat 1572) ?All components underfilled including BGA control
Celestica Confidential 9

WS-CSP Reliability Results
?Marginal improvement in reliability for underfilled BGA device (approx. 25%)
?Less dramatic than increase seen in WS-CSP (higher CTE mismatch)

Celestica Confidential

10

WS-CSP Reliability

Underfilled WS-CSPs exhibit
approx. 4X improvement in attach reliability (nonreworkable formulation)

Celestica Confidential

11

Reworkable vs. Non-Reworkable Reliability Comparison
?Evaluate direct comparison of ATC reliability for larger

I/O BGA using:
?No underfill ?Reworkable underfill – Loctite 3513 ?Non-reworkable – Zymet X6-82-5LV50

?Component :
?CBGA 625 I/O ?SAC Solder Balls

?ATC as per IPC-9701 TC-1
?0-1000C ?6000 cycles
Celestica Confidential 12

Underfill Material Properties

Celestica Confidential

13

BGA Underfill Reliability Results

?Reliability of reworkable underfill at best equal to or slightly worse than no underfill at all ?Non-reworkable underfill improved reliability approx. 4X over no underfill
Celestica Confidential 14

Reworkable Underfills
?Current evaluations being

performed using Loctite 3549
?Exhibits better flow properties than

Loctite 3513
?Less prone to voiding

Celestica Confidential

15

BGA Drop Test Performance
? Failures observed very early on BGAs subjected to drop

testing when not underfilled
? Corner joints first to fail. Also observed solder joint cracks

after multiple drops

Celestica Confidential

16

Drop Test Setup

Celestica Confidential

17

Drop Test Performance
?Both reworkable and non-reworkable formulations

evaluated:
?Non-reworkable: Emerson & Cuming E1216, Loctite 3593 ?Reworkable : Emerson & Cuming XE1218, Loctite 3550, 3551,

FP6101

?Combination of 100 drop pre-conditioning followed by

ATC testing
?As before, reworkable underfilled components saw lower

reliability than non-reworkable counterparts
?Direct comparison to non-underfilled no possible, since they would

have failed the 100 drop pre-conditioning and never made it to ATC
Celestica Confidential 18

Reliability after Drop Testing

Data Coutesy UIC Consortia

Celestica Confidential

19

Adhesion to Soldermask
?Need to evaluate performance of

targeted underfill material with
other chemicals and soldermask being used
?Presence of flux residue can also

potentially lower ATC or drop performance
?Possible explanation for lower

performance of non-reworkable 3593 on previous reliability plot

Example of Interaction with Soldermask resulting in no stick

Celestica Confidential

20

Backup

Celestica Confidential

21

Reworkable Underfill Rework

PCB Side After Removal

Torque Handle for Rework Nozzle

“Heat and torque” nozzle can remove
Component Side

component and break underfill bonds without

any pad damage. However, UF residue still
remains on PCB which must be removed.
UF Residue to Remove
Celestica Confidential 22

Reworkable Underfill Rework
While underfill is still warm, use “solder
Using “solder sucker” to remove remaining residue

sucking” attachment
on rework station to simultaneously

remove underfill
residue and excess solder from pads.
Good wetting after reflow

Celestica Confidential

23

Reworkable Underfill Process flow
Reflow Component with UF
Remove Component with UF by heat and torque nozzle Remove solder and UF residue from pad using site redress tool Dispensing paste/flux on PCB site Replace new compnent on PCB 5DX X-ray

Celestica Confidential

24

Dispenser AVL
?Automated machines for dispense currently on our AVL

and in production include:
?Asymtek M2020 & X1020 ?Camalot XyflexPro DLM

?Speedline about to release new dispensing platform

called FX-D
?Less expensive than XyflexPro ?All features / options field upgradeable

?Slightly slower and less accurate than XyflexPro however for

large BGA applications this should not be a concern
Celestica Confidential 25

High Volume Dispensing Applications
?Dual nozzle configuration for capacity ?Approx. 100K units / week encapped

?Pail pump system for material flow
?2 systems for overall cycle time ?Fixture design for multi-up/orientation req’d ?Inspection/attachment bench req’d for 50HD ?Convection oven/IR Oven for curing in-line Valve DV-03

?Rework process/tooling identified

Valve DV-05

Pail-pump system
Celestica Confidential

IC Application example
26


相关文章:
Asymtek S-920用户培训手册_图文
18 四. 程式编写 我们举例介绍一种编程方法: 一快 PCB 板上有 BGA 零件需要做 underfill,对于 BGA 我们采用“L”型点胶方式,并分四循环点完所需胶量,点胶时...
pcb&smt专用术语
Reliability/Service 质量/可靠度/服务 P/N Part ...Challenges at BGA and CSP Assembly and Rework ...(空洞) Underfill Voiding(底部填充空洞) Bridging(...
芯片互联技术的研究现状与发展趋势
high performance and meet the requirements of ...面阵型芯片(如 BGA)键合可以通过焊料焊凸键合和...同时还需要 底部填充胶 (Underfill)填充在芯片与基板...
手机CPU加胶维修参考
的维修过程(无铅工艺) underfill 的维修过程(无铅工艺): 1、取下 BGA:以逐步升温的模式,让 BGA 上的温度达到 345 度,并保持 70 秒,使用镊子或其他尖 细的东...
深圳市达邦德科技有限公司-Underfill(底部填充胶)的功...
DABOND 深圳市达邦德科技有限公司 Underfill(底部填充胶)的功能与应用 1:为什么要用底填胶? 解决 PCBA 上的一个关键问题,CSP/BGA 存在的隐患-应力集中; 1) ...
回流焊工艺常用中英文术语
Challenges at BGA and CSP Assembly and Rework Stage (BGA、CSP 组装和翻修...Underfill Voiding(底部填充空洞) Bridging(桥连) Open(开路) Underfill Crack(底部...
回流焊炉温测试通用工艺
(Underfill) : 回流焊是 SMT 最后一个关键工序, 是一个实时过程控制, 其过程...但是,公司 PCB 组装上大量使用 BGA、PLCC、FINE PITCH 等 元件,加上 PCB ...
SMT中氮气气氛下的焊接
Underfill Voiding(底部填充空洞) Bridging(桥连) Open(开路) Underfill Crack(底部...(BGA)到不断更密间距的零件,要求 新的回流焊接炉来提供更精确控制的热传 导...
IQC PCBA检验规范
BGA 器件 天线 模块 焊点 焊点宽度小于 3/4 引脚宽(或焊盘宽其中较小者);...Underfill √√√ 生效日期 2012 年 05 月 25 日 核准 审核 制作 东莞市**...
有限元仿真技术在板级工艺可靠性设计中的应用
PCBA 高温翘曲变形、 PTH 疲劳可 靠性评估、CSP/Underfill 可靠性评估等方面。...针对 FCPBGA,供应商给出了两种 PCB 焊盘尺寸的可靠 性数据,见表 5。表明 ...
更多相关标签:
underfill | underfill胶 | underfill工艺 | underfill胶3811 | underfill点胶方式 | underfill点胶机 | underfill胶外观标准 | molded underfill |