ICT Seminar - "Digital Manufacturing".
10th November 2004
The Institute of Circuit Technology (ICT) held a symposium on Digital Manufacturing in South Shields on 10th November 2004 which was attended by an audience of over forty. Ten fabricator companies were represented. ICT Chairman Andrew Hall welcomed delegates and acknowledged the hospitality of Circatex in hosting the event.
Circatex Chairman Steve Jones opened the proceedings with “A European Strategy for the 21st Century”, a blunt reminder that so far as Europe was concerned, there remained no future in high-volume manufacturing of printed circuits. The European industry could only survive and prosper by adopting a service-orientated philosophy and needed very quickly to establish the ability to turn around small-to-medium batches of complex product in cycle times as short as 24 hours, with no leeway for the costs of yield loss or material wastage. Achieving consistently acceptable registration on large panels with HDI design-rules had become the greatest challenge, and the realities of in-process dimensional changes in both materials and artworks had pushed the registration requirement beyond the capability of traditional photolithography.
Using skilful mathematical modelling, together with a real-time demonstration of the yield-prediction procedures which Circatex use to determine whether a job is manufacturable to a particular dimensional tolerance at a particular panel size, Jones made a very convincing case in justification of digital imaging. He saw “digital” as the way forward for primary imaging, not just for dimensional reasons, but also to support his quick-turn short-run concept of sequential inner layer production and single-panel lamination. Digital direct imaging techniques offered a means of eliminating artwork effects and the opportunity for in-process panel-by-panel dimensional compensation for material movement. The only robust technique currently available was laser direct imaging, but he saw an enormous opportunity for ink-jet printing of etch resist, even though line resolution and edge definition were not yet to a standard which would support contemporary design rules. He declared his willingness to cooperate actively in the development of ink-jet processes to assist in driving imaging technology forward.
David Wayness of Rohm and Haas Electronic Materials gave a comprehensive overview of digital imaging techniques, explaining the principles of laser direct imaging and laser direct structuring before focusing on the subject of ink-jet printing. It was clear that Rohm and Haas saw the potential of ink-jet as a key imaging technology of the future and had committed to a major development programme in partnership with Xaar and Schmid. Wayness explored material and pattern-formation issues, and discussed objectively some of the physical realities of drop formation and flight, illustrated with stroboscopic slow-motion video. He introduced the concept of the grey-scale, as compared with the binary, printhead and the contribution it could make to improvement in edge definition
Adriano Blason of equipment manufacturer New System made a cost justification for ink jet legend printing as alternative to conventional screen printing, which indicated a potential five-year saving of $350K for a fabricator setting-up 25 different legend images per day for production batches of 25 panels. He explained how New System had developed their system based on drop-on-demand ink-jet printing using low-viscosity inks, with on-the-fly UV curing as the method of controlling flow-out once the ink had wet the substrate surface. Their bi-directional printers each had a pair of printheads with an independent UV lamp on each side, one for each printing direction. New System would shortly be in a position to offer 100 micron line and space etch resist capability, based on the availability of the grey-scale printhead, and they were in the process of optimising ink adhesion, etch resistance and strippability. Blason described a reel-to-reel etch resist printer which had been custom-built for automotive-industry flexible circuits, where the design characteristic of extremely long conductors meant that conventional imaging processes could not be used. Progress continued to be made in solder resist application by ink-jet. Ink adhesion and printing resolution were improving, and ink-jet had the capability to selectively deposit different thicknesses on different features, produce solder dams by flooding up to the edges of etched features, and economise on ink consumption by not printing on to bare substrate.
John Scott of Patterning Technologies gave a broad insight into industrial inkjet applications. He particularly emphasised the reality and complexity of process development, with examples of problems and limitations regarding data rasterisation, resolution, accuracy of image placement and interactions between inks and substrate surfaces. “We have yet to find a material we can’t print, provided we do the right things...” All was not as easy as it might have appeared: many questions remained to be answered and the reliable deposition of functional materials from industrial piezo-electric print-heads was a world apart from the deposition of water-based ink on absorbent paper by the desk-top bubble-jet machine. He described a process route for ink-jet printing large-format LCD displays, and looked at PCB printing processes – legend, solder mask and etch resist, then examined other potential printed circuit applications: printed passive components, resistors and capacitors, direct writing of conductors, conformal coatings and selective application of solder, referring to printhead capability limits and the potential of the grey-scale head to make a major step-improvement in edge-definition.
Robert Harvey, Technology Manager of printhead manufacturer Xaar, challenged the audience to “think outside the box”. He examined options for the creation of printed interconnect, with or without the incorporation of integrated passive components, by ink-jet techniques: subtractive, additive and direct-write, and suggested that circuit technologists seriously consider alternatives to traditional concepts so that the industry could form some consensus on what would be the preferred routes. This might involve some radical revision of design rules and the industry’s characteristic conservatism could present an obstacle to progress.
One of his illustrations was of a 20-layer flexible circuit recently produced by Epson by all-inkjet deposition of conductive and dielectric inks.
Having reviewed the fundamentals of print-head design and operation, and explained the principles of Xaar’s newly-introduced grey-scale head, on which were founded many system integrators’ ambitions for next-generation image quality, he called upon the PCB industry to define its requirements in the light of what was prospectively and realistically achievable by ink-jet printing. The printhead is the key enabling component for integration into printed circuit ink-jet imaging systems, but current demand for print-heads by the printed circuit industry represented only a tiny fraction of Xaar’s output. To put this in perspective, Harvey commented that Xaar manufactured 180,000 printheads per annum, which were used 74% in graphic arts, 15% in packaging, 2% in industrial and considerably less than 1% in PCB applications. Xaar saw the potential of PCB applications to grow to 30%, but would not commit resources to speculative development. They were prepared to build partnerships with paying customers in order to drive development, and would prioritise their programmes in line with clearly defined needs.
The symposium succeeded in presenting a broad spectrum of information about the present status of, and future scope for, ink-jet techniques in the printed circuit industry. It provoked much thought and discussion and, most importantly, provided a vehicle for potential collaborators in future development projects to meet, exchange ideas and form the beginnings of new interactive partnerships.
Pete Starkey
ICT Council
November 2004