One critical issue with the automated assembly of printed circuit boards (PCBs) is to screen for bad parts and eliminate any risk that boards will actually be assembled and soldered with defective components. A simple problem with a bent pin on a surface-mounted integrated circuit, for ex-ample, can become a major problem involving rework on the board.
The pin-position data provided by the optoNCDT 2000 sensor forms the basis of an algorithm calculation to assess pin shape and alignment.
Engineers at Siemens Dematic Electronics Assembly Systems, who manufacture PCB assembly equipment, are familiar with this issue. In their high-speed board assembly machines, a dedicated vision system checks every component for the right shape, orientation, and for defects. Large ICs, such as 256-pin quad flat packages (QFP 256), with fine pitches down to 0.4 mm, are particularly sensitive. Misshapen or misaligned pins may cause the IC to tilt, producing defective connections when soldered. Such defective pin arrays can be caused by faulty IC production or through damage in transit or handling. To overcome this problem, Siemens Dematic offers an optional component pin inspection module under the designation Coplanarity Module, which is linked to the vision system on the machine. The module is available for the SIPLACE F5HM end-of-line machine.
The Coplanarity Module utilizes an optoNCDT 2000 laser triangulation sensor from Micro-Epsilon. The pick-and-place head on the machine passes the component through the sensor's beam so that each row of pins is scanned. A laser diode projects visible light onto the pins and a charge-coupled device (CCD) array in the sensor picks up the diffuse radiation reflected from the pins. The CCD array supplies a position-dependent, digital signal proportional to the distance between sensor and the pin. Fluctuations in intensity are compensated automatically during the processing of a single measurement. The sensor is therefore able to cope with variations in the sur-face properties almost instantaneously, and supplies measurements with a resolution down to 0.25µ. The pin-position data is fed to the optoNCDT controller containing a single-board embedded PC. A program developed by Siemens Dematic runs on the PC and analyzes the pin data, checking for misalignment and bent pins. ICs that fail this analysis are rejected by the assembly machine.
Petra Klein-Gunnewigk, product manager at Siemens Dematic, says, "With a large IC such as QFP 256 in 0.5 pitch, or even larger application-specific ICs (ASICs), correct pin alignment is essential and we have a tolerance of only twenty micrometers. The software algorithm evaluates the pin positions affecting the component's center of gravity. It then assesses whether tilting will cause some of the pins to make inadequate contact with the solder paste." She explains that, for this size of IC, the inspection time required by the Coplanarity system is about 2 to 3 sec. Once the line production has stabilized and the incidence of faulty pin arrays drops to acceptable levels, the inspection stage can be easily omitted and included again later if quality problems arise.
For industrial control applications, or even a simple assembly line, that machine can go almost 24/7 without a break. But what happens when the task is a little more complex? That’s where the “smart” machine would come in. The smart machine is one that has some simple (or complex in some cases) processing capability to be able to adapt to changing conditions. Such machines are suited for a host of applications, including automotive, aerospace, defense, medical, computers and electronics, telecommunications, consumer goods, and so on. This discussion will examine what’s possible with smart machines, and what tradeoffs need to be made to implement such a solution.