Stress effects of epoxy adhesives on ceramic substrates and magnetics

Citation
Dw. Swanson et Lr. Enlow, Stress effects of epoxy adhesives on ceramic substrates and magnetics, MICROEL REL, 41(4), 2001, pp. 499-510
Citations number
2
Categorie Soggetti
Eletrical & Eletronics Engineeing
Journal title
MICROELECTRONICS RELIABILITY
ISSN journal
00262714 → ACNP
Volume
41
Issue
4
Year of publication
2001
Pages
499 - 510
Database
ISI
SICI code
0026-2714(200104)41:4<499:SEOEAO>2.0.ZU;2-T
Abstract
Conventional, high T-g adhesives can result in cracking of magnetics (trans formers and inductors) and ceramic substrates after curl or after thermal c ycling and extended stress screens. Cracking in substrates may result in el ectrical "opens" in underlying circuitry. Cracks in large(1.8 x 3.650 in.(2)) substrates under magnetics (ferrite cor es/inductors and wire-wound transformers) have been observed after screen t esting. Cracks in the substrates are most likely due to stress transfer in centrifuge in combination with "off-axis" stresses in thermal cycling and e nvironmental stress screening. Cracks are also observed in the epoxy fillet . Stress modeling is often difficult or impossible due to complex routines that are required and the lack of material characteristics such as Young's modulus, Poisson's ratio, and stress/strain curves for materials. Direct co mparison of materials is often not possible due to missing vendor data or d ata from different vendors performed under different curl conditions using varying test methods. A series of materials with varying room-temperature modulus characteristics ranging from over 1 x 10(6) to approximately 19,000 psi were evaluated for material characteristics and reliability in screen testing using sequentia l thermal cycling and centrifuge. Test vehicles were constructed using four sets of magnetics in packaged low-temperature co-fired ceramic (LTCC) subs trates. The specific objectives of this work were to select, evaluate, and recommen d a new adhesive for core-and-coil attachment. Evaluations performed for th e materials selected include (1) material testing and preliminary, first-or der stress calculations, (2) testing and evaluation of actual parts, and (3 ) correlation of results using NASTRAN (NASA structural analysis) mechanica l modeling. This material was, in part, previously presented at the 1999 International Microelectronics and Packaging Society (IMAPS), is copyright 1999 by IMAPS, and is reprinted with permission from the Proceedings of the 1999 Internat ional Symposium on Microelectronics, pp. 896-902, 26-28 October, 1999. (C) 1999 IMAPS. Published by Elsevier Science Ltd. All rights reserved.