Prediction of wave solder machine parameters based on printed circuit board design characteristics Conference Paper uri icon

abstract

  • Automatic wave soldering has been a well-established manufacturing technique for about twenty years, and yet many basic concepts of this process are not understood or controlled. This results in much higher rates of defects, something that is becoming unacceptable in today's highly competitive environment. Effective control of the soldering process requires an understanding of the basic characteristics of both the materials used and the production equipment involved. The goal is to achieve a consistent formation of perfect solder joints. Machine parameters such as: preheat temperatures, conveyor speed, flux density, and wave height must be set correctly to ensure proper solder joint quality. The selection of the optimal combination of these parameters is done through experimentation. However, the use of trial and error for parameter setting often requires using several (expensive) production boards and seizing the resources (operator and machine) for an extensive time period. Hence, there is a need for a method to determine the set of machine parameters given the design characteristics of the board such as component density, board thickness, number of leads, board size, etc. In this paper we present preliminary modeling results based on a neural networks model.

published proceedings

  • Intelligent Engineering Systems Through Artificial Neural Networks

author list (cited authors)

  • Malave, C. O., Sastri, T., & Johnson, R.

complete list of authors

  • Malave, CO||Sastri, T||Johnson, R

publication date

  • December 1992