Although many techniques used in building automation (such as light and climate control, security and surveillance systems) are also used in home automation, additional functions in home automation can include the control of multi-media home entertainment systems, automatic scenes for dinners and parties, and a more user-friendly control interface.

It is the philosophy of Ashurst Systems to engineer individual systems in a way that the technology introduced to the residential environment is unobtrusive and simple to use, whilst enabling the end user to enjoy the latest advances in home technology and automation.

Ashurst Systems were responsible for the installation of a Cat 5e/Cat 6 and OM3 Fibre backbone system, as well as the Horizontal Cat 6 cabling of the Business & Management Suite at one of the most exclusive apartment developments in the world.

With a hybrid Backbone System consisting of Belden Cat 5e/Cat 6 and Excel OM3 fibre this prestigious project was completed on time and within budget.

Ashurst Solutions was responsible for the design and installation of a structured cabling system at the London HQ of an International Utilities Company.

The building comprised five floors with SCR on each floor providing diversity for both horizontal Cat 6 and vertical Data and Fibre backbone cabling.

The fast track project was completed through a major Fit-Out Contractor.

These were supported with a voice and fibre backbone system diversely routed from and to MER.

The outlets services supported included BMS, WLAN and Patient Entertainment Services.

Ashurst Systems were responsible for the installation of circa 23,000 Copper Cat 6a cabinet and switch harness links within two Data Halls at a leading Bank’s Data Centre to deliver a network infrastructure capable of supporting data rates up to 10Gb.

The Project was completed within an extremely tight timeframe to a high standard and has become the client’s benchmark for its future projects.

Civica, as part of the QED Wates consortium, was selected as the prime ICT contractor for a £270M, 10-year contract awarded as part of the UK’s Labour Government schools improvement initiative.

The deal involves the building and refurbishment of secondary high schools.

Working in conjunction with the QED Wates consortium, Ashurst Systems were chosen to supply and install a Brand-Rex Cat 6 Plus structured cabling solution to all schools.

In the equivalent US standard, the ANSI/TIA/EIA-568-C series, the transmission performance of balanced cabling channels and links is defined in terms of Category – with Category 5e, 6 and augmented Category 6 being approximately the same as Class D, Class E and Class EA of the ISO/IEC standards.

In all the standards the component requirements, covering cables, connecting hardware and cords, are specified in terms of Category.

One way of achieving the desired channel performance is to use components of the correct Category in the correct configuration (sometimes known as a “reference implementation”). Using such reference implementations Category 5 components may create Class D channels (in the ISO/IEC and BS EN world) or Category 5 channel (in ANSI/TIA/EIA world.

However, two important phrases are used in the preceding paragraph and are marked in italic text. The first is the use of the term “one way”. There are in fact three separate routes to conformance with the desired ISO/IEC and BS EN channel performance and only one requires the use of components of a defined Category. The second important issue is the use of the word “may”. This is because the use of components of a given Category in a reference implementation does not guarantee the required channel performance. Figure 1 shows the relevant text in the relevant BS EN 50173-x standards (it is essentially the same in ISO/IEC 11801). The key terms are in the third bullet and are “based upon a statistical approach to performance modelling” which undermine the traditional, and arguably perfectly reasonable, assumption that if cables, connecting hardware and cords conform to a specific Category then any resulting cabling will also meet the requirements for links and channels respectively. To understand why this text is included in the standards one has to remember that cabling and component performance requirements are in a continual state of development. In the 1995 edition of ISO/IEC 11801 (and BS EN 50173) we only had to consider Class D:1995 channels created using Category 5:1995 components. In 2002 the Class D channels and Category 5 components were updated – harmonising them with the then new Category 5e requirements specified in the North American Standards. In addition, we introduced channel Classes E and Class F along with Category 6 and 7 components. It was this that forced the universal amendment of the conformance clause exemplified in Figure 1 after the detailed performance modelling use to determine the performance of the components showed that channel performance could not be guaranteed in all circumstances – for all Classes.

The situation has not only not improved but has got worse with the current introduction of Category 6A and 7A components that may be used to create Class EA and FA channel respectively. Now the modelling indicates that statistical risk has increased and, even worse, that certain configurations of Class FA cabling requires the use of components of performance significantly in excess of Category 7A.

So, in the face of this rather unwelcome news, how should customers specify their needs?

Quite clearly, simply specifying components of a given Category is not the way to proceed unless the specifier has afull understanding of the situations under which the statistical risk to either link to channel performance applies. In fact since two of three routes to conformance do not require the use of specific components (and the third requires technical knowledge or advice) then a simple and dogmatic reliance on component Categorisation would seem to represent a demonstrably poor solution -particularly as the required channel performance increases.

The problems with relying on component performance alone begins with the structure of a channel. As shown in The problems with relying on component performance alone begins with the structure of a channel. As shown in Figure 2, a channel is created by adding some cords to a fixed installation.  The cords are added at the telecommunications outlet (TO), connecting the fixed cabling to the equipment in the work area (work area cord) and at the panels in the distributor, either as direct interconnection to equipment (using and equipment cord) or an indirect connection via cross-connect using both an equipment cord and a patch cord). If a consolidation point (CP) is used then two cords are needed to connect the CP to the work area equipment.

It is not only the presence of cords attached to the fixed cabling that creates the challenge to channel performance but their number, length and performance. It will be noted that the second route to conformance described in Figure 1 states that “channel performance shall be assured when adding more than one cord to either end of a conformant link”. This means that just because the fixed installation has been tested and shown to be conformant (e.g. a Class E link) there is no guarantee that Class E channels are automatically created by adding more than one cord of Category 6 at one or both ends.

Instead, the standards require the attachment of “appropriate” components. This means “appropriate” to the design of the link and the resulting channel. A requirement of BS EN 50173-2 (applicable to all premises adopting office cabling structures) is to design horizontal cabling to provide a minimum of Class D channel performance – allowing the customer the option to specify a higher Class if required. The key thing is to have a design that ensures that the required channel Class can be created. This means that the supplier should advise the client of the conditions under which the desired Class will be achieved taking into account the configuration of the cabling and the environment to which the cabling is subjected.

For example, for a given length of equipment cord at the distributor, which lengths of patch cords should be avoided if resonance-related failures are to be prevented. Similarly, what combinations of CP cord and work area cord lengths must be avoided for the same reasons. Furthermore, are there recommended restrictions of minimum fixed cabling lengths to prevent link test failures where CPs are used – and, finally, what is the impact on fixed cabling lengths of using long cords or where the cabling experiences elevated temperatures – such as those generated by Power over Ethernet.

The answers to these types of question are significantly more important than whether a specific component meets a particular Category. Moreover, it is impossible to determine from the results obtained from a link or channel test:

• Whether or not the components within the cabling met a specific Category
• Whether that Category of performance was achieved by those components in the installed condition.

Therefore, whilst it may be desirable to specify components of a given Category, it has to be considered to be a secondary consideration.
 

Bibliography
ANSI TIA/EIA-568-C Generic Customer-Owned Telecommunications Networks series
BS EN 50173-x:2007 Information technology – Generic cabling series
BS EN 50173-2:2007 Information technology – Generic cabling – Office premises
ISO/IEC 11801 Information technology – Generic cabling for customer premises
This White Paper has been produced by Mike Gilmore, e-Ready Building Limited, on behalf of Excel.

Authorization for reproduction for Enzcom Solutions Ltd with thanks to Excel.

This paper is designed to clarify a number points relating to the use of Horizontal Structured Cabling to support 10G Ethernet. With the ratification of ISO/IEC 11801 Class EA the IEEE agreed to use this for all future development involving 10G Ethernet over copper.

ISO/IEC 11801 and BS/EN 50173-1 outlines a channel length based on 100m, the performance limits for balanced cabling channels are given in 6.4. These limits are derived from the component performance limits of Clause 9 and 10 assuming the channel is composed of 90 m of solid conductor cable, 10 m of cord(s) and four connections for Class EA (Cat 6A ).

BS/EN 50173 references BS/EN50288 (Multi-Element Metallic Cables used in Analogue and Digital communication and control), which specifies the horizontal cable for both unscreened and screened cables, for example 50288-5-1 is the screened horizontal cable and 50288-5-2 defines the work area and patch cord cables for Class E.

At this moment in time the standard for Class EA cables has not been ratified however the respective cables are referenced as follows: 50288-10-1 for the screened horizontal cable and 50288-10-2 for the screened work area and patch cord cables.

The major difference between *-1 and *-2 in both documents is the diameter of the conductors, the first has to be a minimum of 0.5mm and the second limit being 0.4mm, this being defined by their intended use.

Alternatives
There are some alternative solutions for providing 10G Ethernet horizontal system, they include Fibre to the Desk, Class E (Cat 6) cabling over a limited length as well as a smaller diameter Class EA solid conductor work area cable being used in the Horizontal.

Each one of these has their pros and cons however only one of these options is truly compliant with the standards we have outlined. That being, Fibre to the Desk whilst its performance cannot be questioned it does come with a price and is usually deployed in high security environments.

A few years ago the TIA/EIA came up with TSB155 which was an interim proposal to allow ‘existing’ installations of CAT 6 to run 10G, this involved a long list of actions to be taken to mitigate the risks, however it was limited to very short runs (below 55m). It was also only ever developed to existing’ installations and was not designed for New Installations.

The third alternative looks at utilising a cable that complies with 50288-5-2 and designed to meet 50288-10-2 (work area and patch cord) these are typically 26AWG cable with a higher attenuation therefore the distance needs to be de¬rated accordingly.

Attenuation = Reduction of signal strength during transmission. Attenuation is the opposite of amplification, and is normal when a signal is sent from one point to another. If the signal attenuates too much, it becomes unintelligible; Attenuation is measured in decibels (dB).

Solid conductor cable meeting these standards was originally designed for use as Harness and Switch Links, predominantly in Inter Cabinet links within major communication rooms and Data Centres.

In this environment it is not a major issue that some of these cables have an Attenuation factor/rating of 1.5, which means a reduction in the permanent link length from 90metres to 60metres as anything over this length in a data centre would typically involve fibre. There are some that state they are ‘zero loss’ which has been achieved by a higher quality construction but they can still suffer one of the following key issue if not deployed in a suitable manner.

Furthermore the reduced dimensions and lower cost can be seen as a distinct advantage in some high density applications within the DC and Major Comms Rooms.

Now comes the key issue, possibly the largest concern about using 26 AWG cable throughout; it involves Power over Ethernet (PoE) it is being increasingly deployed to power devices such as a phones or cameras etc. The standard PoE has the capability to handle 25.5 W over 4 pairs.

In September 2009 IEEE 802.3at (PoE Plus) was approved and devices are now starting to appear on the market that utilise enhanced PoE to support 25.5W for 2 pair powering and up to 51W across all 4 pairs. This doubling in power will have a dramatic knock on effect in ways we have never previously considered when designing structured cabling installations.

The ISO/IEC TR 29125 lays out guidelines for remote powering and gives worst case temp rise for cable bundles of different Category vs. current carried per pair however it only covers down to 24 AWG Cat 5e cable. The table above gives an indication of the potential temperature rises.

At 600mA, which is the upper limit of PoE Plus, a bundle of 127, 24 AWG Cat5e cables will see a temperature rise equivalent to over 13º Centigrade. Using the above, it is not hard to imagine the heat produced by the not un-realistic number of 300 cables in a run.

To date no detailed analysis has been done to estimate the temperature increase that will result from using thinner 26 AWG cable, however if, as we have already established, the cable has a greater resistance, it is only reasonable to assume this will be reflected in an additional temperature increase.

As all twisted-pair cables are referenced in the cabling standards at 20º C +/- 3º C, beyond this the Attenuation is adjusted by a factor of 0.2% per degree Celsius. In turn the performance of the cable could be dramatically reduced as temperatures within bundles could be well above this level.

Conclusions
Consideration must be given to the first paragraph of this paper. The IEEE agreed to use Class EA as defined by ISO/IEC 11801, BS/EN 50173 as the basis for any future developments around 10G Ethernet. Trying to implement a non standards based solution runs the risk of not being able to support future applications as they are developed.

To give an example, if IEEE were to come up with a High Definition Interactive Video Solution over 10G Ethernet, over a Class EA system, there are no guarantees this would work over a bespoke solution using either Cat 6 cabling or the 26AWG work area cable.

It is imperative to weigh up the very limited numbers of benefits against a large number of potential risks by going with either of the two options mentioned. The IEEE is constantly developing new applications and relevant standards, and are doing this based upon the physical medium i.e. the Cabling System meeting what ISO/IEC and BS/EN have defined.

The result may be a risk too far.

This Technical Note has been produced by Paul Cave, Technical Manager, on behalf of Excel.

What’s the problem?

Fibre basics
Fibre optic cabling carries pulses of light between transmitters and receivers. These pulses represent the data being sent across the cable. In order for the data to be transmitted successfully, the light must arrive at the far end of the cable with enough power to be measured. Light loss between the ends of a fibre link comes from multiple sources such as the attenuation of the fibre itself, fusion splices, macrobends and loss through adapter couplings where end-faces meet.

In lower data rate networks with shorter lengths, loss budgets may be generous enough to allow for significant attenuation throughout the link and still the link will function properly. However, there is one perpetual trend in structured cabling: the constant push for greater bandwidth. As fibre links are pushed to carry higher data rates, loss budgets get correspondingly smaller, requiring all loss events to be minimised.

Enemy #1 — a dirty face
Among key sources of loss that can bring a fibre network down, dirty and damaged end-faces are the threat most underestimated. In a survey commissioned by Fluke Networks, dirty end-faces were found to be the #1 cause of fibre link failure for both installers and private network owners. Contaminated end-faces were the cause of fibre links failing 85% of the time. It’s astounding and yet easy to prevent. Nevertheless, there continues to be a lack of appreciation for this crucial issue and lots of misinformation about proper techniques.

What to look for and when
Network professionals need to know what to look for when evaluating end-face conditions. There are two types of problems that will cause loss as light leaves one end-face and enters another inside an adapter: contamination and damage.

Contamination
Contamination comes in many forms from dust to oils to buffer gel. Simply touching the ferrule will immediately deposit an unacceptable amount of body oil on the end-face. Dust and small static-charged particles float through the air and can land on any exposed termination. This can be especially true in facilities undergoing construction or renovation. In new installations, buffer gel and pulling lube can easily find its way onto an end-face.

Ironically, protective caps – also called “dust caps” – are one of the most common contributors to contamination. These caps are made in high-speed production processes that use a mold release compound that will contaminate end-faces on contact. Further, as the plastic cap ages the plasticizers deteriorate resulting in an outgas residue. Last, airborne dust itself will find its way into the protective cap and will move to the end-face when the cap is pushed onto a ferrule. It’s a very common mistake to assume that end-faces are clean when patch cords or pre-terminated pigtails are removed from a sealed bag with protective caps in place.

Inspection of the end-face should verify that no contaminants are within the field of view. The most crucial area to ensure is clean is the core of the fibre, followed by the cladding. Yet contamination on the ferrule – outside of the end-face – could slide towards to core as the fibre is mated or handled. Therefore, all visible contamination should be removed if possible.

Damage
Deciding to mate every connection first and then inspecting only those that fail is a dangerous approach as the physical contact of mated contaminants can cause permanent damage. This permanent damage would require more costly and time consuming re-termination or replacement of pre-terminated links.

Damage will appear as scratches, pits, cracks or chips. These end-face surface defects could be the result of poor termination or mated contamination. Regardless of the cause, damage must be evaluated to determine if action is required as some of it can be ignored or remedied. Up to 5% of the outer edge of fibre cladding generally may be chipped as this is a common result of the polishing process. Any chips on the core are unacceptable. If scratches or excess epoxy bleed is found, repolishing with fine lapping paper can eliminate the problem. If the end-face is cracked or shattered, then the fibre must always be re-terminated.

In every instance, all end-faces should always be inspected before insertion. If a connector is being mated to a port, then the port should be inspected as well. Inspecting one side of a connection is ineffective as contamination inside a port can not only cause damage but also migrate to the connector being inserted. Too often equipment ports are overlooked not only as contaminated themselves but also as a source of contamination for test cords.

How to inspect

Fibre microscope choices
From the first days of fibre optic cabling, microscopes were used to inspect end-faces. Initially stereo bench top microscopes were modified to handle the task in manufacturing environments. Over time new microscopes were designed specifically for the task, resulting in smaller units that could be taken down the hall to the cabling closet or outside into the field.

Microscopes can be divided into two basic groupings: optical and video. Optical microscopes incorporate an objective lens and an eyepiece lens to allow you to view the end-face directly through the device. Today, the barrel shaped microscopes are ubiquitous in termination kits and used to inspect patch cords during troubleshooting. The best feature of these microscopes is their price as they are the least expensive way to see endface details. Their drawback is that they are unable to view end-faces through bulkheads or inside equipment. As a result, you will sometimes here these microscopes referred to as “patch cord scopes.”

Video microscopes incorporate both an optical probe and a display for viewing the probe’s image. Probes are designed to be small so that they can reach ports in hard-to-access places. The screens allow images to be expanded for easier identification of contaminants and damage. Because the endface is viewed on a screen instead of directly, probes eliminate any chance of harmful laser light from reaching a person’s eye.

Microscope evaluation
What matters most about a microscope is what it shows the user. In the case of fibre optic inspection, the goal is to identify all contaminants and damage of a minimum size and within a critical area. Users must first identify the appropriate minimum size contaminant or defect that will affect their system. The smallest-sized item that a microscope can detect is referred to as its detection capability. Next, look for the microscope that has the largest field of view while also maintaining the necessary detection capability. It is preferable to see as much of the surface area as possible while maintaining requisite detection capability. Detection capability and field of view require a trade-off as improving on one dimension tends to require a detriment to the other.

If detection capability and field of view are the most appropriate measurements of a microscope, then why is magnification the prevalent metric. Magnification is perfectly applicable to optical microscopes as their performance is a direct function of the objective and eyepiece lens inside the device. Where magnification becomes less applicable is in video microscopes where the size of the image is a function of both the magnification of the lens as well as the size of the screen. Complicating matters further is the effect of contrast on the ultimate goal of detection capability. Magnification specifications for video microscopes are a vestige of the historical prevalence of optical microscopes. Though magnification is directly related to detection capability, it is a less precise measure of a fibre microscope’s capabilities than detection capability and field of view.

How to clean

Beware of bad habits
Because cleaning has been part of fibre maintenance for years, most people have their own approaches for cleaning end-faces. However, beware of bad habits as many have developed in the industry over time. With an evolving base of knowledge, the industry has moved recently towards new best practices. One common approach to cleaning end-faces is to blast them with canned air, either on a connector or inside a port. Canned air is only effective on one type of contaminant: large dust particles.

Canned air is ineffective not only on oils and residues but also on smaller, charged dust particles. Moreover, canned air will tend to blow large particles around inside ports rather than carefully remove them.

Use of solvent
Another suboptimal approach is to clean without use of a solvent. Solvents provide multiple benefits, the most being their ability to dissolve contaminants that have dried or adhered onto the end-face. In addition, solvents will envelop particles and debris to effectively lift them from the ferrule surface so that they can be carried away without damaging the end-face. Last, solvents will prevent a static charge from developing during cleaning with a dry wipe or reel. There are many stories of end-faces becoming statically charged during solvent-free cleanings such that they were strongly attracting
static-charged dust floating in the air. The developed charge can be so strong that static dust accumulates on the end-face during the short move from a microscope into port.

Solvent selection
Isopropyl alcohol (IPA) has been used for years in the fibre cabling industry to successfully clean end-face and continues to find broad use today. But there are solvents now available specially formulated for fibre end-face cleaning that are far superior to IPA in every way. These new solvents are more effective at dissolving virtually every contaminant than IPA. Further, these custom solvents will dissolve non-ionic compounds such as pulling lube and buffer gel that IPA will not.

With a specified lower surface tension, the specialised solvents will do a better job of enveloping debris for removal than IPA. When cleaning inside ports, evaporation rates become important as lingering solvents can become trapped during mating, resulting in a harmful residue. Fibre-specific solvents have tailored evaporation rates that give them time to work yet disappear before mating.

Last, IPA is highly hygroscopic which means it will draw water moisture from the air and onto the end-face. This water mixes with the IPA and leaves a residue if it dries on the end-face. To be safe, leave the IPA in the medicine cabinet.

Cleaning tools
There are a wide variety of tools available to clean end-faces. The most basic tools are wipes and swabs used to clean patch cords and inside ports, respectively. More involved approaches include mechanical, hand-held contraptions designed to make easier work of cleaning. The most complex devices incorporate blasted solvents or ultrasound in water to achieve the best result. While the more complex systems may achieve better results, they cost far more money.

Individuals should determine the best approach for their application and budget. The one key criterion for wiping materials is that they be lint-free. Shirtsleeves are unacceptable!

Best practices
Whatever approach is selected, certain truisms apply to fibre optic end-face inspection and cleaning. First, inspection must occur not only before but also after cleaning to ensure a good result. If a post-cleaning inspection shows remaining contamination, then a second cleaning must follow. Second, both sides of any connection need to be inspected as every mating involves two surfaces coming into contact. And last, it is almost always easier and cheaper to inspect and clean as a preventative measure than as reactive response.

Consistent inspection and cleaning up front will avoid unexpected and costly downtime in the future.