DocumentCode
2171666
Title
New CBGA package with improved 2nd level reliability
Author
Pendse, Raj ; Afshari, B. ; Butel, N. ; Leibovitz, J. ; Hosoi, Y. ; Shimada, M. ; Maeda, K. ; Maeda, M. ; Yonekura, H.
Author_Institution
ChipPac Inc., Santa Clara, CA, USA
fYear
2000
fDate
2000
Firstpage
1189
Lastpage
1197
Abstract
In the present work, we have studied several improvements in the materials, structure and design of the conventional flip chip-on-ceramic single chip package aimed at increasing the 2nd level reliability. The use of a novel ceramic substrate material (“HITCE Ceramic”), coupled with systematic changes in design and assembly materials resulted in an improvement of 2nd level reliability over the conventional alumina-based ceramic ball grid array (CBGA) package by approximately one order of magnitude. In the initial testing, a strong effect of the heat spreader (lid) structure on 2nd level reliability was seen. A careful finite element modeling (FEM) study was undertaken to understand the interaction of the package structure with 2nd level solder joint stress. The results of the study were validated based on empirical temp cycle data and by direct solder joint strain measurements using a novel strain measurement technique. Once validated, FEM was used as a tool for optimizing the package structure, namely, the lid material and thickness, the attach material between the lid and the the ceramic substrate, and the size and location of the attachment points. To minimize the impact on thermal performance and component level reliability, the die attach material was left unchanged. The optimized package structure was subsequently fabricated and subjected to 2nd level reliability testing. An approximately one order of magnitude improvement was seen, consistent with FEM predictions. It was necessary to ensure that the component-level reliability was not compromised as a result of the higher coefficient of thermal expansion (CTE) of the ceramic substrate material (HITCE), which presented a greater CTE mismatch between the die and substrate compared to the case of alumina ceramic. Therefore, a re-selection of the underfill material was performed and the component-level reliability with the chosen underfill material was verified through temp cycling and moisture tests (Cond B temp cyl and HAST)
Keywords
ball grid arrays; ceramic packaging; encapsulation; finite element analysis; integrated circuit packaging; integrated circuit reliability; moisture measurement; thermal expansion; CBGA package; CTE mismatch; FEM predictions; HITCE Ceramic; attach material; attachment points; ceramic substrate; ceramic substrate material; coefficient of thermal expansion; component level reliability; component-level reliability; die attach material; direct solder joint strain measurements; empirical temp cycle data; finite element modeling; flip chip-on-ceramic single chip package; heat spreader; lid material; moisture tests; package structure; second-level reliability; solder joint stress; strain measurement technique; thermal performance; underfill material; Assembly systems; Ceramics; Electronics packaging; Finite element methods; Materials reliability; Microassembly; Soldering; Strain measurement; Stress; Testing;
fLanguage
English
Publisher
ieee
Conference_Titel
Electronic Components & Technology Conference, 2000. 2000 Proceedings. 50th
Conference_Location
Las Vegas, NV
Print_ISBN
0-7803-5908-9
Type
conf
DOI
10.1109/ECTC.2000.853325
Filename
853325
Link To Document