Page 113 - Wire Bonding in Microelectronics
P. 113

92     Cha pte r  F o u r




                                                             X
                        140                                    4


                                               3
                      Pulling force (gf)  100  2






                              1
                        50






                         0
                          0        10      20       30      40
                                    Rise in pulling hook (mils)

              FIGURE 4-9  Bond-pulling force (related to stress) versus rise in pulling hook
              after wire contact (related to strain and elongation) for a 200 µm (8 mil)
              diameter aluminum wire bond on a power device. (1) is the triangular loop
              formation and elastic wire tensioning region—the dotted lines indicate the
              typical variations that are observed in this region; (2) is the wire elastic limit;
              (3) is the region of inelastic wire deformation; (X) is the region where the wire
              necks down rapidly and then breaks at (4). This curve was corrected for
              measuring apparatus nonlinearities.



              limit of the wire, and point 3 is the region of inelastic (plastic) elonga-
              tion, which, in this case, begins at approximately 60% of the bond
              pull force. At point X, the wire necks down rapidly and then breaks at
              point 4. The elongation of the wire in region 3 was determined to be
              10.5% by using data from Fig. 4-7 and the measured bond geometry.
              More explanation, as well as stress-strain curves of typical bonding
              wires, is given in Sec. 3.2.


         4.3 Ball-Bond Shear Test

              4.3.1 Introduction
              The wire-bond pull test is universally used to assess the strength and
              to determine bonding machine setup parameters of wire bonds used
              in microelectronics (see Sec. 4.1). Often, technicians and engineers assume
              that pull test data, which are adequate to determine wedge-bonding
   108   109   110   111   112   113   114   115   116   117   118