|
Institute of Circuit Technology Chairman Steve Payne welcomed delegates to the 34th ICT Annual Symposium at Tweed Horizons in the Scottish borders, and reflected upon the early years of the Institute, in an era before the introduction of personal computers, mobile phones and the Internet, when the Institute was the leading body in the industry and was a founder member of the Printed Circuit World Convention. Although numbers had declined as the industry regressed in the late 1990s and early 2000s, membership had continued to grow steadily throughout the last 12 months and now stood at approximately 230. An encouraging demographic was that Institute is getting “younger” in terms of average age of members. Payne was confident that the UK would continue to play an important role in the innovation, design and fabrication of printed circuit boards from flexible circuits to the most complex multilayer structures, and emerging technologies such as printed electronics.
The keynote address came from Dr Peter Hughes OBE, Chief Executive of Scottish Engineering and leading spokesman on matters affecting the Scottish manufacturing engineering sector. His infectious enthusiasm for the opportunities that exist in industry in Scotland was evident from the outset. With clearly presented statistics he reminded the audience just how significant was the Scottish manufacturing industry and its contribution to exports. It was enlightening to learn that that electronics is bigger business than whisky! But the key to the future success of industry lay in education and encouragement of young people. In recent years an “obsession with qualifications” had led to a disparity between what Scottish employers needed and what education and skills training youngsters could access. A university education was not an essential; the actual requirement was for a system for developing skills that met everyone’s objectives and equipped individuals with abilities flexible enough to meet the needs of today and respond to the demands of tomorrow. Many routes were available for young people to fulfil their potential, the strategic expansion of apprenticeship being a practical means of filling the skills gap amongst Scotland ’s future workforce. Not many keynote speakers include a surprise musical feature. Dr Hughes, with the help of some audience participation, rounded off his presentation with a demonstration of his skills on guitar, mandolin, autoharp - even the McNally Strumstick!
Data management in electronic design and manufacture was the topic discussed by
Carl O’Roche of Quantum CAD. In a competitive industry where the differentiation between best-in-class SMEs and average companies was their success in meeting product development objectives and consequently driving profitability by driving product to market. Collaboration was of utmost importance, particularly in a context of increasing outsourcing of activities and increasing numbers of external partners. The control of product realisation and project management in the development and engineering phase was becoming increasingly difficult, with numerous modifications, revisions and product versions and the increasing stream of communication involved. In the specific example of a design bureau acting as an external partner, efficient control of data was crucial. O’Roche described how Quantum had adopted a proprietary dynamic data management system as their collaboration platform for controlling and tracking all of their PCB data. It was a flexible web-based server/client system with full project flow control and event logging. All files were securely stored on their server, with nothing left on susceptible file systems, and everyone involved in the project had their own username and password to log on to the server and access their design data in a central shared area. Every visit was recorded and notification alerts were sent automatically to inform users of changes to a file. All previous versions of drawings and documents were retained. The system had shown measurable benefits in visibility of changes and verifying completion of changes. O’Roche’s comment that “20% of your time every day is spent looking for documents” provoked a unanimous murmur of agreement from members of the audience.
Jonathan Smuga from Napier University reported the progress of a project to develop conductive polymer composites for EMI shielding applications, being conducted in collaboration with an industrial partner specialising in the manufacture of pigments, metal powders and metal flakes. Having explained the basic mechanism by which voltages and currents in a device can induce corresponding voltages and currents in a neighbouring device, he discussed the theory of shielding and the factors determining whether electromagnetic radiation was transmitted, absorbed or reflected by the shielding material. A significant parameter was the intrinsic impedance of a surface, a ratio of the electric to magnetic field amplitudes, which decreased with increasing conductivity. Of the options available for creating an EMI shield, solid metal enclosures were prohibitively expensive for most applications and plastics offered a more cost-effective solution provided they could be made sufficiently conductive. Thermal metal spraying gave only line-of-sight coverage and could cause degradation of the plastic surface. Electroless copper plating could give uniform coverage but involve a relatively slow multistage wet process. Vacuum deposition and sputter coating techniques had the limitation of chamber size, and applied metal foils were too labour intensive for other than prototypes. Having discounted these options, attention was focused on conductive plastic composites, which offered the potential to combine metallic conductivity with plastic processability. Conductive fillers were judged to be a more practicable choice than inherently conductive polymers, charge transfer complexes or organometallic compounds. Candidate materials were filament nickel, flake nickel and expanded graphite, added to the “percolation threshold”, at which the material becomes conductive. Coating formulations were prepared by dispersing the fillers in solvent to prevent agglomeration, then blending them into polymethyl methacrylate resin. Samples were coated and dried then measured for surface resistivity, EMI attenuation and reflectivity. The resistance values of the graphite-filled samples were too high for the material to be effective as an EMI shield. The nickel-filled examples showed gave more encouraging results. Future work would include testing at higher frequencies – up to 40 GHz, and modelling with Comsol Multiphysics software.
Alan Colquhoun, Principal DfM Engineer at BAE Systems Hillend, presented a worked example in design for manufacture. He demonstrated what could be achieved by a DfM/DfT integrated project team in improving the manufacturability and testability of a design. The engineering methodology facilitated the design of products such that they were easy to manufacture. General objectives were to reduce the number of parts to minimize the opportunity for a defective part or an assembly error, to reduce the total cost of fabricating and assembling the product, to design verifiability into the product to provide a natural test or inspection of the item, to avoid tolerances beyond the natural capability of the manufacturing processes, to design for ease of assembly by minimizing fastenings and hand-soldered joints, and to design for ease of servicing the product. The DfM/DfT team also made the effort to understand more about the capabilities and limitations of the production system, in order to refine design rules to further guide and optimise the product for production. Colcquhoun used a hypothetical assembly to demonstrate how, with logical reasoning in a team environment a stage-by-stage cost reduction of 30% could be achieved. In the real example of the Commander radar system, a total saving per system of £200,000, at a cost of 21 man days work, had been made by his DfM/DfT integrated project team.
In a presentation entitled Fine lines with LDI, Uwe Altmann from Orbotech began with the question: What do you mean by fine lines? Many manufacturers would refer to their own capability and answer maybe 100 microns or 75 microns. Orbotech defined “fine lines” as 50 microns and below, and already had the machine capability to achieve 15 microns, although developments in photoresist capability were awaited before these dimensions could be realised in production. He listed the benefits of laser direct imaging, “photolithography without a mask”, as the elimination of repeat defects and vacuum-contact effects, and the minimisation of surface topography effects by the +/- 300 micron depth of focus of their Large Scan Optics, which had been developed in collaboration with Zeiss in Jena. He explained in detail how the optical system had a total beam path of 9 metres, with the laser being transformed into multiple beams and reflected on to the panel by a polygon. Each pixel was independently addressable, and the on-the-fly dynamic-scaling system could achieve registration to +/- 12 microns, with compensation for panel orientation, and dimensional distortion. Typical throughput was 80 double-sided panels per hour, using an 8-watt laser and a resist of 10-16 millijoule photospeed. Altmann demonstrated the precision of the system with photographs of a pattern of 25 micron lines in a dry film 100 microns thick, which showed extremely cleanly defined vertical sidewalls. Additional features of the LDI system were its ability to mark panels with individual serial numbers, date and scale-factor stamps and bar codes.
Francesca Stern of BPA achieved a breakthrough in communications technology, with the Institute’s first-ever remote presentation. Her PowerPoint was in the conference room, her voice was in the conference room, but Francesca herself was in BPA’s office, nearly 400 miles away. She explained in detail how BPA went about their business of analysing and forecasting business information and producing technology roadmaps. She used the analogy of driving a car: the driver needs to see very clearly where he’s going, whilst at the same time knowing what’s going on around him, and he needs to know critically when, and which way, to turn - forecasting helps the driver make that decision. Forecasting was all about accumulating data from the past and intelligently extrapolating it into the future. Francesca demonstrated the various ways in which growth curves could be presented, and techniques for smoothing the curve so that underlying trends became more clearly visible. She discussed the use of “leading indicators”, economic indicators that change before the economy has changed, such as average weekly hours worked by manufacturing workers, new housing starts, unemployment figures,money supply, inventory changes, new orders for capital goods and stock exchange prices. But these were only useful for short-term forecasting. National statistics, input from trade associations, response to questionnaires from companies in the industry – all contributed to the information from which forecasts were prepared. Factors like international conflicts, currency effects and national disasters could all push data out of shape. With few exceptions, history showed that BPA’s forecasting had been consistently accurate, and the forecast continued to operate as tools to highlight trends and turning points so that executives could swing the steering wheel in the right direction at the right time, and know when to tread on the accelerator or the brakes.
The final presentation came from Mike Osmond of Intrasys Design, on the subject of “Maintaining the Balance”, with reference to DfM, or, as he put it, DfX - design for all desirable attributes or, more simply, design for excellence. PCB design was a physical definition of the electrical requirements – a transition for concept to the real world. But in today’s real world it was no longer a “join-the-dots” exercise, because if the dots weren’t joined by interconnections with the right characteristics, the device would not work! The designer had a heavy burden of responsibility – the quality of his design could substantially affect the yields of both PCB fabricator and assembler, and efficient design processes were critical to ensure that the product got to market early. Osmond emphasised the importance of analysing the design early, to save revision spins, eliminate scrap and reduce labour and hardware costs. DfX methodologies ensured an end product fully optimised for the highest overall performance at the most economical unit cost. He concluded his presentation by connecting a CAD system to the projector and giving a live demonstration of how the engineer interacts with a real design.
Outside of the lecture theatre, during breaks in the proceedings, delegates made the most of the opportunity to visit the tabletop exhibition area and to network with their peers. All round, a very successful and well-balanced event and a credit to the efforts of ICT Technical Director Bill Wilkie and the staff of Tweed Horizons.
Pete Starkey
ICT Council
June 2008
|
