(images: courtesy of Trevor Galbraith)
In the evolving Industry 4.0 debate for machine-to-machine communication, one could easily make the analogy if machines are the limbs of the body, then inspection systems act as the brain and software (MIS) systems are the central nervous system.
To enable the machines (limbs) to receive correctional feedback from the inspection systems (brain), inspection systems need to provide accurate data. Inspection systems can be divided into two principal categories: positional systems (non-intelligent), such as sensors that may report on the location of a board within the process and monitoring systems (intelligent) that measure positions before, during and after a given process and report off-sets, missing components, wrong polarity and other anomalies to the preceding machine in the process (printer or pick and place machine), (or management information system) which then makes an intelligent decision on how to handle these data.
Inspection systems pre-reflow can be classified as ‘failure prevention’ systems and those that are post-reflow can be classified as ‘failure detection’ systems. Put simply, after you have soldered the components to the board, any defect detected beyond that point will be subject to the rework and repair process. It is therefore essential to detect defects as early as possible in the production process. The earlier you detect the defect, the less money and time it will cost to correct it.
Two main failure prevention systems provide the process data to the printers and pick and place machines. These are Solder Paste Inspection systems (SPI) and 3D Automatic Optical Inspection systems (AOI). Given that over 60 percent of defects can be traced back to the printer, the SPI system performs a critical function by detecting and measuring the amount of paste deposited on the board. This volumetric data is then fed back to the printer through the MIS system. The printer can decide to change the position of the stencil, increase the pressure on the squeegee, adjust the angle of the blade, volume of paste or simply stop the process.
The position of the AOI is critical. In a perfect environment manufacturers should have both pre-reflow and post reflow AOI. Pre-reflow AOI should be 2D and 3D. Both perform different functions – 2D is used for scanning barcodes and identifying nomenclature on components, while 3D can perform ‘z’ height measurements to detect defects such as lifted leads or coplanarity issues. Pre-reflow 3D AOI machines can verify the dimensions of the component or device against the parameters loaded from the BOM list, and positional accuracy data from CAD information. This could allow the Pick and place systems to adjust x, y, or z pressure, correct polarity issues, replace missing components, clean nozzles or simply stop the process.
If production planning can only afford one 3D AOI, then it must go post-reflow to provide comprehensive protection. However, for the reasons mentioned earlier, the ability to detect the defect before it is soldered makes a strong argument for pre-reflow 3D AOI. AOI can detect a much wider range of potential defects, but to date, most systems use comparative images against a ‘Golden Board’ rather than metrology data that can be passed to the printer or placement machine for corrective action.
Unlike the high accuracy SPI systems, traditional AOI determines the location of a box or area on the PCB relative to a fudicial or datum point on the board. Consequently, it can measure the location or detect PCB stretch, comparing it to the Gerber data, but it lacks the ability needed to provide large amounts of actionable data to the placement machines and printers. New 3D AOI systems that use metrology are beginning to emerge, these claim to have accuracy in the micron region of measurement. This leads to accurate z height inspection helping with lifted leads and warped QFNs etc.
Thermal Profiler systems provide valuable feedback on the behavior of the printed board, components and any thermally sensitive areas of the assemblies as they transition through the elevated thermal excursions in the reflow oven or selective soldering system. Oven monitoring systems provide information on the status of the energy consumption, fans, airflow and nitrogen.
© 2013 Janóczki M, Becker Á, Jakab L, Gróf R, Takács T. Published in Automatic Optical Inspection of Soldering under CC BY 3.0 license. Available from: http://dx.doi.org/10.5772/51699
All components move during the reflow process. However, modern solders, fluxes and tight control of the peak reflow temperature have minimized the disruption in modern systems. In addition to most of the defects listed in pre-reflow AOI, post-reflow AOI is also looking for solder shorts and opens, tombstoning, lifted leads and other defects where the solder has not fully coalesced.
In-Line X-ray is capable of finding most of the defects that can be detected by the AOI systems, plus void measurement and joint inspection which AOI cannot perform, e.g., BGAs. However, this is post-reflow which will result in rework and is an extremely slow inspection process, unless it is combined with an optimization technology, such as ‘SmartLink’.
This is starting to be used as a complimentary technology to In-line inspection processes and some systems are able to make quite accurate measurements which provide real benefits to process control and improvement.
Fig. 3a: Defects detected during failure prevention and Process Monitoring
Fig. 3b: Defects detected after reflow
Further work is needed to accelerate the use of metrology in 3D SPI and 3D AOI to provide accurate correctional data to the printer, MIS and pick and place equipment. Thermal profiling software has to work more closely with the materials community to understand, plot and report on the behavior of materials during reflow. Every effort should be made to reduce failure detection rates. After a board has passed through the reflow oven, it adds cost to rework it and the potential for field failures increases.