|
|
|
|
|
|
|
|
Comparison of SAC105 and SAC305 Solders
Materials Tech |
|
Authored By:Gregory Henshall, Hewlett-Packard Co. Jasbir Bath, Bath Technical Consultancy Sundar Sethuraman, Jabil David Geiger, Flextronics International Ahmer Syed, Amkor M.J. Lee, Xilinx Keith Newman and Livia Hu, Sun Microsystems Dong Hyun Kim and Weidong Xie, Cisco Systems Wade Eagar and Jack Waldvogel, Motorola TranscriptMany BGA and CSP component suppliers have begun shipment of components with a variety of "second generation" lead free solder ball alloys. Much of the motivation for the alloy changes has been to improve mechanical shock resistance. Several publications have established the improved performance of such 2nd level BGA/CSP sphere alloys; however, much less has been published regarding the thermal fatigue resistance of components with these new lead free ball alloys. In this paper, the thermal fatigue performance under accelerated test conditions is compared for three common BGA ball alloys. The implications of findings and areas for further study are discussed. SummaryMany BGA and CSP component suppliers have begun shipment of components with a variety of "second generation" Pb-free solder ball alloys. Much of the motivation for the alloy changes has been to improve mechanical shock resistance. Several publications have established the improved performance of such 2nd level BGA/CSP sphere alloys; however, much less has been published regarding the thermal fatigue resistance of components with these new Pb-free ball alloys. As these components and alloys become mainstream, their use in situations where thermal fatigue resistance is critical to product life will become an important consideration. Therefore, an understanding of thermal fatigue performance for new alloys is necessary for OEM/ EMS/ ODM companies to make design and procurement decisions, and for component suppliers to ensure the reliability of their products under a range of field use conditions. In this study, the thermal fatigue performance under accelerated test conditions is compared for three common BGA ball alloys: SAC105, Sn-3.5Ag, and SAC305 as a control. Accelerated thermal cycle (ATC) testing was performed using 676 PBGA components with 1.0 mm pitch and electrolytic Ni/Au finished component pads. These components were assembled to high-temperature rated Cu-OSP coated printed circuit boards using SAC 305 solder paste, which represents one of the most common assembly practices in industry today. ATC testing was performed using the IPC-9701A TC1 condition of 0/100°C with 10-minute dwells (nominal); 3 different failure criteria were used in constructing the Weibull failure curves. The data indicate that SAC105 has the lowest thermal fatigue resistance among the alloys tested, with Sn-3.5Ag and SAC 305 having similar and superior performance. The impact of failure criterion on the Weibull curves is also presented. The implications of these findings and areas for further study are discussed ConclusionsThis investigation focused on the thermal fatigue performance of solder joints made using three common BGA ball alloys: SAC105, Sn-3.5Ag, and SAC305. Accelerated thermal cycle (ATC) testing was performed using a 1.0-mm pitch, 676 PBGA daisy-chain test vehicle. The BGA packages with electrolytic Ni/Au-finished component pads were assembled to high temperature rated Cu-OSP coated printed circuit boards using SAC 305 solder paste. ATC testing was performed using insitu electrical monitoring and the IPC-9701A TC1 condition of 0/100 XC with 10-minute dwells (nominal). Three different electrical failure criteria were used in constructing the Weibull failure curves: 20% resistance rise, 500°, and hard open (infinite resistance). The following conclusions were drawn from this study. 1. SAC105 ball with SAC305 paste has the lowest thermal fatigue resistance under the conditions tested, with Sn-3.5Ag and SAC 305 having similar and superior performance. Failures for all alloys were confirmed to be due to thermal fatigue in the bulk solder near the package pad. 2. Use of the IPC-9701A standard failure criterion of 20% resistance rise provides the most sensitive measure of failure among those studied. Failure was typically detected 200 to 500 cycles sooner than for the coarser criteria, though somewhat more scatter was evident. The two high resistance failure criteria gave very similar Weibull curves. 3. The relatively poor performance of the SAC105 ball alloy joints does not necessarily mean this alloy has insufficient thermal fatigue resistance for specific applications. Comparison of the data from the current study with those from an earlier one using the same test vehicle suggest that SAC105 ball alloy joints may perform worse than those with eutectic Sn-Pb solder; however, it is suggested that a direct comparison within the same study be performed tovalidate or invalidate this conclusion. 4. Areas for further study of the thermal fatigue performance of solder joints have been identified, including assessments of the impact of dopants, varying surface finishes, and the development of acceleration models Initially Published in the IPC Proceedings |
|
Comments
|
|
|
|
|
|
|
|
|