Issue link: https://iconnect007.uberflip.com/i/806706
34 SMT Magazine • April 2017 4. Die-Attach Sandwich Setup The high-power die-attach setup was simulat- ed with the use of two Cu coupons, which were to be joined with a flux-coated solder preform. A 3-inch x 3-inch ceramic plate was used as a car- rier. The simulated die-attach sandwich (Cu cou- pon on a solder preform, then placed on another Cu coupon) was placed on the carrier. At die-at- tach with solder preforms, many manufacturing processes use fixtures with some weight on the die to secure the sandwich. In this study, metal weight varying from 0 to 100 grams was placed on the top of the sandwich to simulate the fixture weight. Paper cardboard was placed on top of the die for heat insulation, if a weight was used. 5. Reflow Peak Temperature The sandwich on the carrier was placed on a 5-zone, in-line contact reflow oven, and then sent through the reflow oven with various peak temperatures as follows: • SAC305: 240°C, 250°C, and 260°C • 63Sn/37Pb: 205°C, 215°C, and 225°C • 57Bi/42Sn/1Ag: 160°C, 170°C, and 180°C 6. Reflow Under Various Pressures For SAC305, various weights were placed on top of the sandwiches at reflow: 0, 10, 30, and 100 grams. For this set of assembly, freshly cleaned Cu coupons, 0.5% flux concentration, and a peak temperature of 240°C were used. Ten sandwiches were reflowed for each condition. For 63Sn/37Pb and 57Bi/42Sn/1Ag, 0, 10, 20, and 30 grams of weight were used for the pres- sure study. 7. Assessment After reflow, the sandwiches were examined under X-ray to determine the voiding area percent- age. For SAC305, three sandwiches for each weight condition were cross-sectioned to determine the bondline thickness under various weights. Results Effect of Weight on SAC305 System Using X-ray Images Figure 1 shows the X-ray images of SAC305 sandwiches under various weights. Here, the flux concentration was 0.5%, the peak temper- ature was 240°C, and the coupons were etch- cleaned prior to use. Although 10 samples were prepared for each combination of conditions, three sandwiches were removed due to large opens in the solder joints because the image analysis software was unable to determine the voiding percentage properly. Most of the sam- ples showed full wetting to the perimeter of coupons, and only a few showed a small frac- tion of non-wetting near the edge of the joints. Most of the voids showed plain vacancy, ex- cept for a few samples where some spotty light- VOIDING CONTROL AT HIGH-POWER DIE-ATTACH PREFORM SOLDERING Figure 1: X-ray images for SAC305 soldered sandwiches processed with various weights on top of samples.