Automated Optical Inspection (AOI)
Automated Optical Inspection (AOI) was a step implemented several decades ago to improve the PCB product yields. It was originally started to inspect PCB layers. The layers could be scrapped at an early stage reducing cost of manufacturing. The defective layers could also be reworked to an acceptable level.
AOI, automated optical inspection is an automated vision inspection of PCB during the manufacturing process. It is used to scan the inner layers and outer layers of a PCB after the processes of etching and stripping. After scanning by an AOI machine, the defects which do not meet the manufacturer’s requirement will be identified by the machine. In this way, AOI can detect problems early in the production process, so faults would not be passed to next production process and production cost could be saved. Therefore, AOI can increase the quality of PCBs and reduce the production cost.
About the AOI Machines
The basic principles of the AOI machine are to have a motion system in combination with a camera and specialized software. The market has older generation machines utilizing monochrome cameras that can detect transition from substrate to copper surface finish, this allows for the image to be compared to the CAD data or the learned image.
There are colour and monochrome camera types. For AOI machines, they only need to classify copper and substrate, so a monochrome type of camera is enough for them. Modern AOI cameras are digital using the full colour spectrum. This allows for finer and finer lines/traces to be inspected. In digital imaging, a pixel (or picture element) is a single point in a raster image. The pixel is the smallest addressable screen element; it is the smallest unit of a picture that can be controlled. Each pixel has its own address. The address of a pixel corresponds to its coordinates. Pixels are normally arranged in a two-dimensional grid, and are often represented using dots or squares. Each pixel is a sample of an original image. More samples typically provide more accurate representations of the original.
The intensity of each pixel is variable. In colour image systems, a color is typically represented by three or four component intensities such as red, green, and blue, or cyan, magenta, yellow, and black.
The number of pixels is also an important parameter of cameras. Larger number of pixels does not mean it can detect smaller defects. To detect smaller defects, smaller pixel size is required. Pixel size is calculated by scan width / number of pixels.
For example, the scan width is 2 inches and number of pixels of camera is 2000. Pixel size = 2i inches/2000 = 1mil.
Some machines use cameras with fewer pixels. But they can use more cameras to compensate. Normally the exposure time for more pixels CCD is longer. So, the scanning time of higher pixel CCDs is longer too. Therefore, some machines use multiple CCDs with less pixels rather than using one camera with more pixels.
monochrome and color difference
The camera will only work with suitable light to enhance the surface finish of the inne rlayer or outer layer. Older systems will use halogen lamps, infrared lighting or LED’s for modern machines. Each light has a distinct advantage depending on the type of product you are trying to inspect.
The light of AOI machines consists of specular and diffused light.
Specular light and diffuse light on flat surface
For specular light, light ray incident on the surface in a single angle and reflected into a single direction by the surface like mirror-reflection. For diffuse light, the light incident on the surface is from many directions and reflected in many directions.
Consider a dent on a line, the surface of the dent is not flat, most of the specular light will not be reflected to the camera. But diffuse light will be less affected by the surface and there will be some part of diffuse light being reflected to the camera.
Figure 25 – Specular light and diffuse light on a dent
If the diffused light intensity of the dent is high enough, the brightness of the dent will be similar as the brightness of flat surface. This small brightness difference may cause the machine miss detecting the dent.
On the other hand, if the intensity of diffused light is reduced, then it may cause some false catching, such as oxidation, dust, oil stain, etc.
Some machines use a basic pixel by pixel comparison. Although effective, it can increase processing time.
Detect Defects by pixel by pixel
Some machines use logical methods to detect the defects. They use different methods such as complex contour comparison and design base inspection to detect the defects on different features. So, the process time would be longer. To maintain short process time, the machine needs faster computers for comparison using pixel by pixel method.
For AOI, the main two factors that determine whether the defects can be identified are the size and brightness/color of defects. The pixel size of the AOI machines determine the minimum size of defects it can detect. The pixel size can be seen as the basic unit of the scanned image. Therefore, the minimum defect size detectable must be larger or equal to pixel size. But even if the defect size is the same as the pixel size, the defect might be very difficult to detect. For example, in the following case, defect size is same as pixel size.
For case A, it is the best situation where the defect is inside the pixel, so the defect is easily detectable. For case B, it is the worst situation as each of the four pixels can only see the 25% of the defect, so it may not detect the defect. Therefore, if the defect size is very small and is close to the pixel size, then it may not be detected by the machine.
Case for defect size equal to pixel size
Case for defect size equal to pixel size
Another factor which makes the defect difficult to identify is brightness/color of defect. If the brightness/color of the defects is close to normal area, then it will be difficult for the machine to detect the defects. The common missing defects on AOI and AVI are scratches and dents.
A verification station is used offline to the AOI to verify or repair defects.
Why Gardien’s Solution?
Your Gardien Local Service Centre may offer this service. The AOI tooling and production services are controlled by Gardien’s proprietary job flow system called Ontrack (link to Ontrack). This creates a seamless and error free flow of information from the customer supplied data to the Service Floors at Gardien.
Gardien’s team is trained and qualified on all internal process as documented Quality Management System. The internal process has specific inputs for incoming, certified, and not good boards as well as descriptive educational programs on various board types, and surface finishes.
Gardien certifies each order processed with a Certificate of Compliance with details about how the order was processed, what specifications were used to certify the product, equipment used with calibration expiry date, team member who processed the order, quantity, and failure analysis.
Gardien strongly recommends that the test sequence and parameters are clearly stated on the manufacturing drawing or at a minimum agreed upon during the quoting or contract realization phase. In addition, any PCB sent to us for processing should have a unique identifier on each peace for electronic traceability.