Issue link: https://iconnect007.uberflip.com/i/1137649
108 SMT007 MAGAZINE I JULY 2019 There was not a way to create nodules arti- ficially; therefore, only PCBs rejected by the manufacturer due to nodules were used. Fig- ures 16 and 17 show examples of wire bonds made on nodules and scratches, respectively. Most nodules were almost completely covered with wire bonds so that it was not clear if nod- ules were under the wire bonds. Results and Discussion of Wire-bonding Experiment Pull test results of wire bonds made on nodules before reliability testing are shown in Table 2. These test results are from Prod- uct I. All of the wire bonds pull-tested before reliability testing showed a neck break fail- ure mode with acceptable wire pull strength (greater than 4.0 g). A neck break failure mode means that the wire bond was broken above the ball bond on the PCB bonding pad. Varia- tion in neck break strength in Table 2 was due to the difference in wire-bond length and loop height. An unacceptable failure mode would be lifted wire bonds on the PCB bonding pads instead of broken wire bonds during wire pull testing. Table 3 shows pull test results of wire bonds made on scratches before reliability testing. Wire bonds were made on scratches on PCBs rejected by the manufacturer. Pull test results showed a neck break failure mode with accept- able pull strength. Figure 16: Wire bond made on nodule. Figure 17: Wire bond made on scratch. Table 2: Pull test results of wire bonds made on nodules before reliability tests.