PCB007 Magazine

PCB-Apr2015

Issue link: http://iconnect007.uberflip.com/i/490885

Contents of this Issue

Navigation

Page 63 of 76

64 The PCB Magazine • April 2015 "pristine area" of the landing pad when gauging the acceptability of the witness mark. For a rect- angle SMT landing pad, the pristine area is de- fined by the central 80% of the land width x 80% of the land length. For round SMT landing pads, the pristine area is defined as the central 80% of the diameter. Electrical test probe witness marks within the pristine area for Class I, Class II and Class III are considered cosmetic in nature and are acceptable provided the requirements of the final finish are met (IPC-6012C). As noted earlier, witness marks may be dif - ferent depending on the type of test solution be- ing used. Fixtures use either solid pins or spring pins with different options for head styles. These can be round (conical), spear point or even chisel point. The main goal for Electrical test is to perform the test required while minimizing the witness mark to the PCB. Many are under of the opinion that fixtures are the main source of pad damage resulting from electrical test and in many cases this is true. Due to the mechanics of the fixture, translator pins, stripper plates (cas - settes) there are many factors in play that can cause a pin to "lock" and apply excessive pres- sure/force to the delicate landing pad. Figures 1 & 2 provide examples of what can result from excessive pressure applied to the PCB. The above examples are typical of "pin lock" or the pin stuck when the fixture is compressed. In Figure 1 the damage is in the pristine area and is severe enough that copper is exposed. In most cases, this PCB is scrap due to the excessive pin hit unless reworked if allowed. In Figure 2, the hit is less severe and is just on the borderline of the pristine area. This pin hit is not exposing copper and in most cases will be allowed. Flying probe machines are much more deli- cate when it comes to witness marks, but they are not totally immune either. Although they do provide a very light touch to the PCB, other factors can increase witness marks from a fly- ing probe. Compression values, X/Y/Z velocities and accelerations all come in to play with flying probe equipment. Figure 3 gives an example of a witness mark created from a flying probe where one of the above noted parameters may be slightly out of tolerance. Figure 4 shows the common witness mark left by a flying probe. The most delicate of all is the test of wire- bond. In most cases the direct probing of wire- bond pads is not recommended, as there is no copper structure under the gold and the pads are extremely vulnerable to damage. Many times, no witness mark in the pristine area is allowed. Solutions to this vary from up-line on the trace outside of the pristine area, the first available pad in the net closest to the wire-bond termina- tion or a 2-pass test solution where the shorts test is performed on a stronger test point within the net and the continuity test is performed by covering the wire-bond area with a conduc- tive material, creating a special test program to check the wire-bond terminations to another point in their net to validate continuity. This testIng toDD ELECTRICAL TEST: SURFACE FINISH VS. WATER MARKS continues Figure 3: FP hard hit. Figure 4: FP standard hit.

Articles in this issue

Archives of this issue

view archives of PCB007 Magazine - PCB-Apr2015