ICT Evening Seminar

Circuit World

ISSN: 0305-6120

Article publication date: 1 December 2004




Coultard, F. (2004), "ICT Evening Seminar", Circuit World, Vol. 30 No. 4. https://doi.org/10.1108/cw.2004.21730dac.001



Emerald Group Publishing Limited

Copyright © 2004, Emerald Group Publishing Limited

ICT Evening Seminar

ICT Evening Seminar

Flexible Technology in PCBs

The Arundel Lodge, Arundel, 25 February 2004

Keywords: ICT, Printed-circuit boards, Electronics industry, Conferences

In February, the ICT held another of its South Coast evening seminars at the Arundel Lodge in Arundel. There was an excellent turn-out for what has turned out to be a popular event, with over 30 attendees and a splendid atmosphere ensued.

The evening began with a paper from Graphic's Kerry Hewlett, which covered the history of flexible circuitry. Flexible circuit technology goes back to the early 1970s and was then, and is now, used as a solution to challenges in many military designs. Graphic entered this technology in the mid-1980s. The early designs used heavy plating and all dimensions were big! Flex material was unstable (and still is) and there were problems with adhesives, drilling and plating. De-lamination was common and peel strengths were low. By the early 1990s the technology had matured somewhat with better materials and tighter stability. Boards commanded higher prices but needed longer lead times than the norm. New materials with higher glass transition temperatures and better temperature resistance became available. Kerry distributed special examples of flex-rigid work. The future will lead to much smaller dimensions and more complex designs with microvia flex-rigid being the greatest growth area. Three-dimensional circuitry is more common with step edge designs, which are very challenging. There is still a relatively high price in this area and controlling assembly costs is a key to success. Good design is essential. The future challenge is definitely stability.

Next from Graphic was Adam Wellington who gave a paper on Design and Manufacturing issues. Adam emphasised the criticality of discussing the concept before starting any project. Design in this field should address, typically, the plated circuit to flex interface, radius calculations, copper type (e.g. rolled, annealed, electro-deposited), bend radius, substrate material selection, number of flexible layers, adhesives and symmetry to name a few. Key manufacturing issues are registration, resin types and related challenges, material movement, copper weights, pattern, telegraphing, drill hole quality, de- smear selection, metallisation, imaging, solder mask, electrical test and profile. Design should focus on collaboration and correct technology selection. There is a need to invest in high technology solutions to stay ahead.

The next paper was from Steve Jones of the Invotec Group entitled "Flex-Rigid Solutions: Living with Dimensionally Unstable Solutions". Steve had crafted a comprehensive coverage of this topic in his usual professional and enthusiastic style. The introduction pointed to general design rule drivers, yield predictions and capability and process benchmarking. The military aerospace environment is a key flex-rigid design-driver. Characterisation of the flex-rigid process involves reducing process variations and coping with variability.

Steve then took us through high density interconnect (HDI) design-rule drivers and the benefits of large panels in this technology. Drivers for flex rigid include the facts that volumes are much smaller than for multilayers and the cost driver being in the processing rather than materials with yield being a major factor. It is essential to compare processing and use benchmarking. He showed both yield prediction and measurement models. Emphasis was put on the huge benefits in measuring the PCB process but also including the artwork stage. Up to 75 percent of design tolerance is lost before drilling, whereas less than 50 percent of variation is due to materials. Considerable engineering effort is essential. Every process step must be characterised and its error contribution minimised. In conclusion, one should recognise that there is no one solution to registration issues, attend to detail and have zero tolerance to variation.

Frank CoultardIntellect

Related articles