Journal of Electronics Manufacturing

While the cost of computer hardware has rapidly decreased over the last decade, this has not happened to the cost of mechanical hardware. The cost of automatic optical inspection systems has thus remained high owing to the use of precision mechanics and optics. This paper describes a system which uses simplified mechanics to reduce cost, while keeping the performance acceptable by using a sub-pixel resolution inspection system. The lack of optical resolution can therefore be offset by more sophisticated software. The system inspects long thin PCBs, using a single-axis stepper motor to position the camera instead of the usual two-axis systems. The system uses sub-pixel resolution techniques to achieve better than 25 μm repeatability with a pixel size of 75 × 70 μm. The system inspects bottom-side SMT components including chip capacitors, resistors, transistors, diodes, SOICs, PLCCs and air-wound coils at a rate of 25 components per second. It uses a transputer-controlled frame grabber hosted in a PC clone. One of the principal design aims was to create a system flexible enough to handle a large variety of board types and SMT components associated with these boards.

This paper describes one of the modules in integrating printed circuit board (PCB) design and manufacturing databases. The module deals with cutting of circuit boards from larger rectangular boards known as panels. When the circuit boards are rectangular, the problem of deciding where to cut the panel is similar to the rectangular cutting stock problem in operations research. None of the existing solution approaches to this problem can be used because of the unique requirements of the PCB manufacturing process. After extensive consultation with industry experts involved in solving the problem manually, a new algorithm has been designed which emphasizes the grouping of similar pieces and patterns in addition to minimizing the trim losses in a panel. The resultant software code implementing the algorithm is the module that integrates design and manufacturing information.

This paper reports on a survey of a sample of surface-mount technology (SMT) users in Singapore. The problems and challenges of this sample of SMT practitioners in adopting and adapting to current and emerging technologies are presented. The conclusion drawn is that there are a number of problems which are common to all the users and which have to be overcome; at the same time implementation of the newer technologies cannot be ignored in order to increase Singapore's competitiveness in electronics manufacturing.

This paper presents mathematical models that are intended to assist in the design of anisotropic conducting materials. Such materials are usually random distributions of hard conducting spheres in epoxy resin matrices. Conducting joints are constructed in soft conducting materials by the plastic deformation of the conductor on the printed circuit boards. An earlier paper (Williams et al., 1993) describes some of the average behaviour of the materials during joint assembly and after assembly. This work explores the effects of the random distribution of particles in the material on resistivity variation and the probability of shorting between fine-pitch interconnections.

We evaluate the performance of several dispatching rules in a semiconductor testing environment and examine the effects of uncertainties in problem data and job arrival patterns. A series of simulation experiments shows that a single dispatching rule seldom performs well for several different performance measures simultaneously, non-homogenous job arrival patterns significantly affect several aspects of system performance while uncertainty in problem data does not, and the choice of dispatching rules for a given performance measure is robust to both arrival patterns and uncertainties.