Issue link: https://iconnect007.uberflip.com/i/408601
32 SMT Magazine • November 2014 new PLacement tecHnoLogy For rework systems continues can be achieved with none to little additional effort: Treating more than one component au- tomatically (for desoldering or placement and soldering) will be an easy task for such a system. Handling the component automatically and having cameras on board allows using the sys- tem for quality assurance aspects, by generating and evaluating images. For example, a compo- nent could be inspected if all pins are in order, before its treatment. It is possible to apply flux or solder paste onto the component with com- paratively simple means in very precise por- tions, by the dipping method. The dipping method requires the precision of the gantry system in combination with a precise cavity allows the supply of accurate and small portions of solder paste to each individu- al solder ball (in the case of a BGA). The previ- ously dipped component is aligned to the target position, placed and then is ready for soldering. The result of this rework process would be com- parable to a soldering result created by a SMT production line. outlook As already mentioned, there will be obsta- cles in the way of this technology in terms of new and unknown components that cannot be handled by the system initially. To address this question, there must be a deeper look into the methods of image process- ing. The key idea is to generate a learning system which will improve its own capability instead of being trained to new challenges by human pro- grammers. Some questions may still demand an- swers, but if the human programmers concen- trate on a learning system which is dealing with a flexible algorithm library instead of adapting a set of algorithms to new data, this vision might not be out of reach for the next generation of rework systems. summary Compared to the existing methods of com- ponent placement and treatment, this new placement technology implemented in a re- work system opens up a field to repair complex electronic assemblies. At the same time, com- petitive cameras and stepper motor axes allow this technology to be available at a more widely accepted price level. feaTure Figure 6: Side view image of a bGa ball, placed after the component was dipped into solder paste. Figure 7: Method of learning system in image processing. (Source: Math & Tech engineering Gmbh [3] )