Andrew Wilson, Editor, [email protected]
IBM Corp. (Rochester, MN, USA; www.ibm.com) develops, manufactures, and programs servers and other computer-system components such as printed-circuit boards (PCBs). Key components in the company's iSeries Model 890 server include connection modules—flexible array interconnect sockets—that are used to mount single-chip module components to a multilayer PCB. These modules measure 36 × 36 mm, with connection contacts 1 mm apart. Before installation, however, each module must be visually inspected for defects such as wire shorts, raised loops, and contamination.
Previously, this inspection was performed manually, resulting in a process that gave inconsistent results between operators. To automate this process, IBM engineers wanted to develop a system capable of inspecting multiple types of interconnect modules, some which contain more than 4000 contacts. Test time was also an important issue. IBM's goal was to inspect an entire module for all defects in less than 5 min. With little capital available to improve this inspection process, IBM engineers implemented a low-cost system based on a Model 542C Color SmartImage Sensor camera from DVT Corp. (Norcross, GA, USA; www.dvtsensors.com) and a Cartesian robot and a Super SEL controller from Intelligent Actuator (Torrance, CA, USA; www.intelligentactuator.com).
"With DVT's machine-vision software, Framework, the integrated system provided seamless communications to ensure that every IBM interposer module contains the necessary number of contacts and that each contact meets stringent quality requirements," says Brian Kreifels, an IBM industrial engineer. "Since the interposer modules had a nonuniform background and there are variations in part size, shape, and color among models, a color machine-vision solution was required," he says.
DVT's 542C camera features a color-matching tool that distinguishes gold wire from a black background. With traditional black and white cameras, the contrast between wire and background was not high enough to detect single wire defects. Also, the DVT camera features a built-in microprocessor that can cut inspection time to an acceptable value.
The camera's customizable user interface and its ability to log inspection data were added bonuses for IBM. With support from DVT's local distributor, Automation Inc (Minneapolis MN, USA; www.automationinc.com), IBM used features found in DVT's Windows-based, point-and-click software, Framework. This software offers built-in ActiveX controls that allow integration with Visual Basic software.
With this tool, IBM developed a custom user interface to view and record inspection images, create password protection for process control and security, and log inspection data to a spreadsheet. These data include results of all inspections by position number. "Now, using a spreadsheet, we can archive the inspection data for future analysis and for a record to reference specific part serial numbers," Kreifels says.