| 摘要 |
A machine vision camera observes components to be placed at a selected location on a substrate from a given angle relative to the orthogonal to the substrate through a mirror and then observes the selected location on the substrate from the same angle but with the mirror displaced in order to measure, register and align under-side contact and edge features of the component to corresponding substrate features. The camera moves with the pick-up head of a placement machine that picks up the component from a component feeder or component store and transports it to a location above a mirror where its bottom surface and edges are imaged as it is held stationary. The mirror may be carried with the pick-up head and may be retractable. Component feature coordinate locations are calculated and used to calculate the coordinates of the component features relative to the pick-up head. The selected location of the substrate is imaged by the same camera but with the intervening mirror displaced and coordinates for features corresponding to those of the component features are obtained in a similar fashion. A difference vector is then calculated to determine how the pick-up head should be moved (excluding the Z- or vertical direction) so that the corresponding component and target substrate features are brought into physical contact. Additionally, an iterative alignment of component to substrate significantly improves alignment accuracy beyond the dead reckoning accuracy of the placement machine's mechanics. Finally, the component is moved in the Z-direction to bring it into contact with the substrate.
|
