PCB007 Magazine

PCB-July2016

Issue link: https://iconnect007.uberflip.com/i/701250

Contents of this Issue

Navigation

Page 13 of 97

14 The PCB Magazine • July 2016 chrome AOI systems. Illuminating the panel with different wavelengths at different angles makes it possible to accurately classify materi- als such as clean copper, oxidized copper and contamination as well as abnormal shapes such as dish-downs or shallow shorts. Two different defects might look the same under one illumination source. But the true dif- ference is revealed using multiple illumination sources, which simultaneously detect the sub- tle differences between actual defects and false alarms (Figure 1). This results in the highest possible detection results without compromis- ing on throughput. Support of the Most Challenging PCB Materials and Types For years, AOI solutions have been able to accurately inspect a range of designs and sub- strates. When it comes to flex and thin HDI layers, traditional AOI systems start to run into problems when trying to detect errors on the transparent and translucent layers. When im- aging multiple layers, the lower layers tend to show through the layers that are being inspect- ed, creating false defect reports. Multiple imaging technology eliminates the multilayer confusion by illuminating the in- spected layer with different light wavelengths. The multiple captured images provide the re- quired visual information for accurate inspec- tion with no false alarms from underlying lay- ers. An example of this is shown in Figure 2. Advanced Image Analysis Solutions and 'Oversensitivity' Reduction The accuracy of the AOI system inspection relies on the accuracy of the imaging, but it also needs accurate CAM data to compare with the image to ensure that the anomalies in the im- age are, in fact, anomalies. Moreover, advanced integration of CAM data with all its inherent information with advanced AOI algorithm adds another critical dimension to detection accuracy. Today, the most advanced AOI solutions significantly reduce oversensitivity—false posi- tive errors—by using intelligent panel under- standing algorithms. The AOI image analysis tools have the information about the detected error point including its location, specific role in the pattern and the defined dimension and FASTER, MORE ACCURATE AOI IS MORE IMPORTANT THAN EVER Figure 1: a) Microscopic image showing possible copper defect; b) using light source 1—still uncertain; c) light source 2—dust/pseudo-defect detected. Figure 2: a) Microscope image; b) light source 1—lower layers show through; c) light source 2—only inspected layer is seen for accurate inspection. A A B B C C

Articles in this issue

Archives of this issue

view archives of PCB007 Magazine - PCB-July2016