Masters Abstracts (2003)

JADHAV, HRUSHIKESH

Email: hrushi_j@yahoo.com

DFM for PCB Assembly

In the electronics-manufacturing environment, the ability to eliminate design defects during the early design phase is of paramount importance. Design inconsistencies that are not addressed during the conceptual phase can increase the potential for manufacturing defects. Also, correcting problems during the manufacturing phase is costly and severely affects the time required to introduce new products into the marketplace. Typically, EMS providers receive customer requirements (designs) from the Original Equipment Manufacturers (OEMs) through a third party CAD designer. During this data transfer process, especially in a low-volume-high-mix electronics-manufacturing environment, the probability of receiving inconsistent PCB designs is high. Hence, it is very important that every design is scrutinized before it is released to production. This research demonstrates the use of a knowledge-based expert system for validating complex PCB designs. Several measurements are performed on the PCB designs with the help of a DFM tool. The expert system processes the measured data and identifies all 'design violations'. It also provides an explanation for the violations, their criticality levels and recommendations for improving the PCB design based upon the rules stored in the knowledge base. The rules referred to by the expert system are in-line with the Institute for Interconnecting and Packaging Electronics Circuits (IPC) standards. The output of the expert system can be given to the CAD designer so that critical design violations can be eliminated before the commencement of actual production. The scope of this research is primarily the validation of land pattern designs. The Graphical User Interface (GUI) provides instructions and menus that can assist even a novice engineer/user to understand and execute the validation of PCB designs. The system also has a user-friendly mechanism for updating the knowledge stored in the knowledge base. The software used for developing the expert system was Visual Age Java (VAJ). VAJ provides an Integrated Development Environment (IDE), which facilities modularity during the system development. This research has demonstrated the capability of Artificial Intelligence (AI) based systems for decision making vis-¨¤-vis the design validation of PCBs. DFM engineers can accurately check large and complex PCB designs. Moreover, the time and effort required for performing the DFM checks is also considerably reduced.