Issue link: https://iconnect007.uberflip.com/i/721932
96 SMT Magazine • September 2016 using a high resolution scan of 1024x960 pix- els. Gating, region of set AE echo, was on the interface between the die face and the underfill material; therefore, the condition of bonding pads onto the die face was revealed. In anoth- er case, the flip-chip package was imaged from the bottom substrate side. Gating was on the interface between the cured underfill and the substrate; therefore, bonding the solder balls to their pads. This shows difficulty of implement- ing such condition for a real application when a large number of other package and materials interfere with acoustic signals. Sakuma, et al., 11 used non-destruction tech- niques for flip-chip improvement as well as ver- ification for such improvement. Both C-SAM and X-ray NDE images were presented for an as- sembly prior to its optimization by a differential heating/cooling chip joint method. The C-SAM investigation detected fractures in the ULK lay- ers, whereas X-ray techniques identified solder joint bridging. In a recent investigation, Phom- mahaxay et al., 12, 13 stated that even though the 200 MHz transducer can detect gross defects, it does not have sufficient resolution to detect voids in a through silicon vias (TSV). Micron size defect detection required the development of a C-SAM transducer with 1 GHz capability. Such a high frequency transduced allowed good and bad TSVs to be distinguished in a number of test samples. The new SAM with resolution and depth sensitivity and defect resolution >>10 µm range enables localization and measure of de - fects in z-3D approach. So, new GHz SAM can be utilized as a new approach for semiconduc- tor failure analysis in 1 µm range with potential for in line tool TSV inspection development for complete 300 mm wafer inspection. Experimental Evaluation by C-SAM Test Plan and Evaluation Approaches This section covers evaluation performed by C-SAM using a number of advanced packages and assemblies before and after various environ- mental exposure. Representative flip-chip plas- tic and ceramic area array (ball/column) pack- ages and assemblies from the previous investi- gations were subjected to C-SAM evaluation. It also included recently acquired land grid array packages and fine pitch assemblies. Key packag- es evaluated included are: • A plastic land grid array after assembly. Because of extremely low stand-off, it was ex- tremely difficult to visually inspect solder inter- connections. It was thought that C-SAM tech- nique may provide an insight into integrity of solder interconnections. • A ceramic flip-chip LGA package with 1517 I/O, which were previously assembled onto PCB and then removed, was subjected to C-SAM evaluation. So, the flip-chip die was ex- posed to two reflow cycles. Since the flip-chip die had no heat sink attachment, its back was exposed. Also, a CGA assembly version of this LGA package, which were previously assembled onto PCB and subjected to thermal cycling, was included in the C-SAM evaluation. • A flip-chip CGA1752 I/O package assem- bly was also subject to C-SAM evaluation. It was realized that this package has an extra heat sink attachment; therefore, the acoustic signal from bonding materials will be a dominant signal. • A flip-chip BGA1704 I/O package assem- bly was also subjected to C-SAM evaluation. The flip-chip die of this package, similar to its ceramic CGA 1752 counterpart, also had an ex- tra interface due to heat sink attachment. This package assembly previously was subjected to a number of thermal cycles. • The FC-CGA 1752 I/O after its heat sink was removed was re-examined. This CGA was re-evaluated by C-SAM for integrity of the flip- chip solder joints since original package showed only integrity of heat sink bonding materials. • A hermetically sealed CGA with 1272 col- umns, which had die wire bonded, was also subject to C-SAM evaluation to determine if in- ternal integrity of wire bonds could be assessed. • A large number of fine pitch and stack package assemblies were subjected to C-SAM to evaluate their integrity and appropriateness of C-SAM. • The FC-CGA 1752 package with no heat sink along with a fine pitch package was sub- jected to cycles up to 20 times of solder iron touch to induce defects. These were re-scanned to determine the level of damage and their de- tectability by the C-SAM technique. DEFECT FEATURES DETECTED BY ACOUSTIC EMISSION