Design of a vision-assisted robotic cell for assembly of surface-mounted printed circuit boards

Abstract

Surface mount technology (SMT) is defined as placement of components on the surface of printed circuit boards (PCB). In this technology, accurate placement of components is a crucial matter. This is due to (1) the existence of errors in the different elements of a placement system (such as a placement machine, board, and component) and (2) the reduction in the component/pad pitch size and the increase in the number of termination leads. The currently available surface mount component (SMC) placement systems use some kind of Component-to-Pad alignment technique. They use feedback from sensory devices such as a vision system to compensate for the positional and orientational errors of components or pad patterns. However, these systems are unable to compensate for the errors (such as size, angular and positional errors) associated with each individual lead/pad. As the number of leads/pads increases and the pitch size decreases, the placement criterion must change from Component-to-Pad alignment to Lead-to-Pad alignment to meet the high accuracy requirement. Systems which utilize a Lead-to-Pad matching technique, compensate for errors in leads/pads by matching each single lead with the corresponding pad. In this thesis, the development of a vision-assisted SMC placement system which utilizes a Lead-to-Pad matching technique is discussed. The placement problem is defined in a mathematical form and modeled as a nonlinear optimization problem. A solution procedure is developed to solve the mathematical problem. The comparison of the solution procedure to a nonlinear optimization technique shows its superiority. The placement system uses two cameras for locating leads and pads. The proper illumination techniques provide sharp images of components and pads which require very little image processing. The vision software programs accurately locate leads and pads and detect any error in size, orientation and position of individual leads and pads. A simulation model is developed to study the effect of the placement machine error on the placement of SMCs with different characteristics. This model can also be used to investigate the placement machine accuracy requirement based on the placement objectives.