Selviah, D.R.; Fernández, F.A.; Papakonstantinou, I.; Wang, K.; Bagshiahi, H.; Walker, A.C.; McCarthy, A.; ... Milward, D.; + view all Selviah, D.R.; Fernández, F.A.; Papakonstantinou, I.; Wang, K.; Bagshiahi, H.; Walker, A.C.; McCarthy, A.; Suyal, H.; Hutt, D.A.; Conway, P.P.; Chappell, J.; Zakariyah, S.S.; Milward, D.; - view fewer (2008) Integrated optical and electronic interconnect printed circuit board manufacturing. Journal of the Institute of Circuit Technology , 1 (3) pp. 9-12.

Preview

PDF 7907.pdf Download (635kB)

Abstract

Introduction: At high bit rates copper tracks in printed circuit boards (PCBs) suffer severe loss and pulse distortion due to radiation of electromagnetic waves, dispersion and bandwidth limitations. The loss can be overcome to some extent by transmitting higher power pulses and by changing the dielectric constant and loss tangent of the PCB substrate material. However, high power pulses consume power and can cause electro-migration which reduces the board lifetime, although the copper tracks can be surrounded by another metal to prevent this at the expense of further processing steps. The use of special board materials can be costly and some materials containing high dielectric constant crystallites can cause poor adhesion. The pulse distortion, dispersion and bandwidth limitations can be overcome to some extent by the use of pulse pre-emphasis and adaptive equalisation at further cost. Electromagnetic waves are radiated efficiently at high bit rates removing power from the track so causing loss, but more importantly they are also received efficiently by other nearby and distant copper tracks on the same PCB, or on adjacent PCBs, or PCBs and other electrical conductors outside of the system enclosure. This EMI crosstalk causes increased noise and so degrades the signal to noise ratio and the bit error rate of the copper track interconnections. Therefore, the main forces driving the development of alternative interconnect technologies are the EMI crosstalk, which becomes increasingly more serious as bit rates increase for longer and denser interconnects, and secondly the cost of overcoming the other problems that occur in copper interconnects at high bit rates. Optical fibres have replaced copper cables for long distance, backbone and submarine applications where they offer wide bandwidths for low loss, produce and receive no electromagnetic interference, and are relatively low cost. Optical interconnects are beginning to penetrate the markets at shorter distances, such as in local area networks, and as their cost is reduced, will be used within the system enclosure. The use of optics is expected to occur first where the problems for copper are most significant which is for high bit rate, dense interconnections in large area backplanes within non-conducting enclosures. Optical fibres are not the most convenient for interconnections within a system as they can only bend through a large radius of about 10 cm, otherwise light escapes from the fibre core into the cladding resulting in loss and signal corruption. Fibre connectors form a major part of the cost of the optical interconnect and a system with many fibres has many costly connectors. The fibres must be individually routed and errors in routing are time consuming to debug and correct. The fibres can be laid flat on the PCB plane and even bonded together within an epoxy layer, but this is not suited to low cost mass production. An alternative technology suitable for low cost mass production is that of multimode polymer buried channel optical waveguide interconnections within layers in the multilayer PCB formed by the same, or slightly modified, processes already available within PCB manufacturing facilities. Copper tracks are still required in such substrates to transmit power through the backplane (or motherboard), Figure 1, in order to power mezzanine (or line, or drive, or daughter) boards and copper is still a practical and low cost option at low data rates. Hence, there is a need to develop a new type of multilayer hybrid PCB in which optical waveguide interconnects are used for the highest data rates, with copper tracks for lower data rates and for power lines and earth planes. These issues have been anticipated by system design companies such as Xyratex Technology, IBM Zurich and Siemens C-Labs, microprocessor designers such as Intel and materials development companies such as Dow Corning, NTT, Rohm and Haas and Exxelis, who have instituted research in their own laboratories and in associated universities into optical waveguide interconnect technology. Leading Universities and Research Institutions such as Cambridge (CAPE), University College London (UCL), Heriot Watt University, Loughborough University, National Physical Laboratory (NPL), IMEC - Ghent University, TFCG Microsystems, Belgium, Paderborn University, Germany, Helsinki University of Technology, Espoo, Finland and ETRI, South Korea are developing novel polymer materials, developing fabrication techniques, discovering design rules for waveguide layout and carrying out precision characterisation. Optical buried channel waveguides usually have a core with an approximately square or rectangular cross section made from a high refractive index (slow speed of light) material and a cladding surrounding the core of a lower refractive index (higher speed of light). They operate by total internal reflection (TIR) in a similar way to optical fibres. The cost of waveguide connectors is minimised by choosing to use multi-mode waveguides which typically have cores of 40 - 70micron width which can tolerate more misalignment than single mode waveguides. The optical buried channel waveguides are formed on a plane by a variety of fabrication techniques which can be implemented, after slight adaptation, in PCB manufacturers. Arrays of low-cost vertical cavity surface emitting lasers (VCSELs) emitting 850 nm wavelength and arrays of photodiodes operating at 1 0 Gb/s are readily available at low-cost for use in optical transmitters and receivers. At this wavelength, polymer is a convenient low-loss material for use as the core and cladding. Polymers can be chosen or designed which can be easily processed to form waveguides at low temperatures, have low cost, and can withstand subsequent high temperature reflow soldering processes. For optical printed circuit boards to be brought into widespread use, layout tools must be made readily available which design both the copper tracks and the optical waveguides [1]. In 2006 David R. Selviah of UCL, formed a large consortium of complementary universities and companies and led a successful bid to carry out a Flagship project entitled “Integrated Optical and Electronic Interconnect PCB Manufacturing (OPCB)” in the Innovative Electronics Manufacturing Research Centre (IeMRC). The consortium companies represented a complete supply and manufacturing chain and route to market for the polymer waveguide technology including companies manufacturing PCB layout tools, computer programs for modelling the behaviour of multimode waveguides, developing and supplying low loss polymer formulations, manufacturing multilayer PCBs, supplying printer fabrication equipment together with end user system companies who require optical printed circuit boards. The following sections describe the project’s objectives, the approaches being taken and some examples of what has been achieved so far in the project with an indication of future directions.

Download activity - last month

Export as JSONCSVXML

Download activity - last 12 months

Export as CSVXMLJSON

Downloads by country - last 12 months

Export as JSONCSVXML

Archive Staff Only

View Item