Effect of Hot Solder Dipping on Part Stresses

Hot solder dipping is an accepted method to mitigate whisker growth on electronic components plated with pure tin. To provide the best protection from tin whiskers, the solder coating needs to be up to the package body. The conventional wisdom for solder dipping maintains that dipping up to the package body can cause component damage. If the specific construction differences of ceramic and plastic-encapsulated parts are considered, it appears that while temperature gradients are important for ceramic parts, temperature is the major factor in failure of plastic parts. Finite-element analysis (FEA) is used to determine internal stresses resulting from solder dipping. Transient thermal FEA analysis is used to determine the temperature distribution inside the part and a static structural FEA is used to calculate internal stresses. For a 128-lead plastic part, maximum stresses from dipping the side of the package are 35% of those expected in reflow and 37% higher than that expected if only the lead tip is dipped. For a 184-lead ceramic part, dipping until the package side touches the solder increases stresses by a factor of more than three over dipping the lead to within 10-mils (0.25 mm) from the package body. Plotting the maximum relative stress as a function of maximum semiconductor die temperature shows a strong linear correlation between maximum temperature and stress for plastic – a correlation not observed for a ceramic part. INTRODUCTION Tin whiskers are needle-like crystals of tin growing from pure tin or high-tin alloy surfaces, which may grow long enough to cause electrical shorts. Hot solder dipping is an accepted method to mitigate whisker growth on electronic components. To provide the best protection from tin whiskers, the solder coating needs to be up to the package body but conventional wisdom for solder dipping maintains that dipping up to the package body can cause component damage. Although this damage has been observed for ceramic parts and some plastic parts, it is clear that plastic and ceramic parts are constructed differently. A ceramic part typically has a good thermal expansion match between the leads and ceramic whereas a plastic part is comprised of materials with varying expansion rates. The use of expansion-matched materials in a ceramic part typically results in selection of a lead material like Alloy-42 that has thermal conductivity comparable to the ceramic. In contrast, plastic part types typically use copper alloy lead materials that have substantially higher thermal conductivity than the plastic part encapsulants. In a plastic part, the good thermal conductivity of the copper lead frame compared to the alloy 42 may tend to reduce the impact of varying dip depths. Winslow [1] provided an excellent analysis calculating and measuring the thermal gradients in a plastic part and made reference to thermal analysis followed by structural analysis but the structural analysis results were not provided. The present analysis evaluates the role of temperature gradients as well as maximum processing temperature in the generation of package stresses. NOMENCLATURE FEA – Finite Element Analysis TQFP – Thin Quad Flat pack TSSOP – Thin Shrink Small-Outline Package Tg – Glass Transition Temperature CTE – Coefficient of Thermal Expansion QFP – Quad Flat Pack CSAM C-Mode Scanning Acoustic Microscopy APPROACH Finite-element analysis (FEA) was used to determine the internal stresses resulting from solder dipping. Once a finite-element model was developed for leaded plastic and ceramic parts, a transient thermal analysis was used to determine the temperature distribution inside the part. When the thermal FEA was completed, the thermal elements were converted to structural 1 Copyright © 2008 by ASME elements and a static structural analysis was conducted to calculate internal stresses. The thermal/structural analysis was conducted for solder dipping at varying depths and the standard surface mount convection reflow soldering process. Plastic Part Finite-Element Model Model Development Prior work (Sengupta [2]) indicated potential durability issues with thin-quad flat pack (TQFP) and thin-shrink small outline package (TSSOP) parts because of the ratio of die area to package thickness. In the present work, the model for the plastic part was based on a 128-lead plastic TQFP manufactured by Lattice Semiconductor [3]. This part was selected due to the expected risk with TQFP parts from the prior study and availability of parts for characterization. Specific dimensions (Table I) of the part were based on the package drawing [3], Xray measurements (Fig. 1), and cross sections (Fig. 2). Leads were modeled as straight (i.e. the lead form was not considered) which will be explained in the development of the Boundary Conditions. Table I – Dimensions Used in 128-lead plastic TQFP Analysis