To improve the reliability and the yield of surface-treated components, new methods for the inspection and in-process quality control have to be developed to detect and avoid defects and inhomogeneities at the micrometre level.
Samples of electroplated contacts surfaces have to be inspected visually for defects such as pits, protrusions, burrs and discolourations as small as 15 um. In order to improve the reliability of inspection, an optical in process inspection system has been developed which detects the defects and analyses their size and 3-dimensional shape. The system consists of an optical sensor unit, a data acquisition unit, a real time transputer based image preprocessor and postprocessor and a system controller. The sensor unit contains a laser scanner and sensors which detect the direction of the scattered light. Due to the required optical 3-dimensional resolution and the high throughput the peak input data rate can reach 72 Megapixels.
The modular nature of the system allows for the separate exploitation of several modules, namely the signal conditioning module and the transputer modules. The latter use direct access to the computer busses, a novel and fast alternative to standard link only based boards.
The optical sensor concept can be applied to many different problems, where small 3-dimensional defects in planar surfaces must be detected.
Further exploitation in other in process inspection tasks is expected because of the low hardware costs.
Therefore the first main task is to develop an automatic optical inspection system which should work on-line due to high speed of plating process. The second objective is the investigation of the origin of defects which are detected by inspection system.
The exploitation of results will help to install a process control loop into the plating baths system in order to enhance the throughput and to reduce the costs.
2 Dublin 2