WCA July 2015
using equal DC current values. Installations using PoE technology differ greatly in terms of their configuration and environment. With an increasing demand for more power, which is likely to require powering over all four pairs, further work is required to investigate cable and component performance in areas such as bundled cables, thermal behaviour in different ambient temperature environments, cable reach under PoE, and connector demating under load. [2] M Gilmore, ‘The impact of copper clad aluminium (and steel) conductors within balanced pair cables (intended for use within implementations of generic cabling),’ FIA-IAN-002, 2011 [3] IEC Subcommittee 46C, ’Proposal for measuring of heating of data cables by current,’ 46C/906/NP, 2009 [4] COMSOL Multiphysics: www.comsol.com [5] G J Anders, Rating of Electric Power Cables in Unfavorable Thermal Environment, Wiley-Blackwell, pp 2-4 (2004) [6] National Instruments LabVIEW: www.ni.com/labview [7] Minitab: www.minitab.com [8] J Wilson and C Hernández-Hall, Physics Laboratory Experiments, Brooks/Cole, p 361 (2009) [9] F S Akinnuoye, H Sasse, V Kang, A Duffy, ‘Heating Effects on channel performance for Power over Ethernet (PoE) applications,’ Proceedings of the International Wire & Cable Symposium (IWCS), November, 2013 [10] H Congdon, B Davis, ‘Mythbusting takes on shielded cabling,’ Bicsi Presentation, 2009 [11] Four-pair PoE study group: www.ieee802.org/3/4PPOE Paper courtesy of the 63 rd IWCS Technical Symposium, Providence, Rhode Island, USA, November 2014. References [1] IEEE Standard 802.3at, 2009
Further consideration should be given to the correlation of simulated data and that of the installation environment. On one hand, the simulation may be based on a worst case scenario. However, in reality, the duty cycle may dictate that the power is only supplied for a fraction of the time. Good installation practices should be implemented wherever possible, such as minimising bundle sizes, accounting for temperature rise for maximum cable lengths, and keeping pathways and spaces free of thermally insulating materials. It is important to note that, while excellent correlation was seen between simulated and measured results for a single cable setup, this work was not intended to replicate the behaviour of cables in bundles. However, it is anticipated that good correlation between theory and practice will also apply to bundled configurations in free air and in a variety of cable containment systems, ie tray, trunking, conduit, etc. A comparison between the CCA 24 AWG UTP and Cat6A 26 AWG U/FTP samples show that it is possible for cables with smaller conductors to radiate less heat than those of larger conductors when supplied with identical DC current values. It is also known that the conductive foil in screened cables act as a heat sink which helps to reduce the amount of heat radiated from the cable [10] . Therefore, it is important to take into account the construction of cable, and not only the conductor diameter for PoE system deployment. The increasing demand for more power supplied from the Power PSE to the PD is obvious, with powering over all four pairs expected to be IEEE standardised [11] and a variety of non-standard products currently available on the market which provide power levels in excess of those stated in IEEE 802.3at. Higher power levels will increase performance risk, but will also allow PoE to be used for a wider range of applications. Conclusions A two-dimensional model was set up using COMSOL Multiphysics software in order to replicate measured results. A test method proposed by IEC subcommittee 46C to assess cables for Power over Ethernet was followed, which outlined the measurement for heating data cables by DC current. Excellent correlation between simulated and measured results for a single cable has been demonstrated. This correlation encourages use of the software for the thermal prediction of cables in densely populated networks. It is also expected to provide a strong indication of the temperature rise for varying bundle sizes, ambient temperatures, and containment systems, etc. The thermal performance of a CCA cable subjected to DC powering has also been proven to radiate more heat than cables which use solid copper conductors
Brand-Rex Ltd Glenrothes Fife United Kingdom Tel : +44 1592 778459
Email : ssimms@brand-rex.com Website : www.brand-rex.com
56
www.read-wca.com
Wire & Cable ASIA – JulyAugust 2015
Made with FlippingBook