Here's our Policy:
Four to five questions and answers will be published.  They will be archived on ACES Int'l BROADBand.

Q & A:

Q1.  My customer’s IT department keeps asking me if the cabling we installed will support Ethernet Layer One, Two and Three devices on their network.  What are they talking about and why is it important to have a cabling system tested to be in "compliance" to the network’s requirements?

A1.  First, realize that most IT network managers have no real concept of cabling. Their misperception is that CAT 5 is the universal standard and fiber is too expensive. However, many are beginning to realize the importance of having a properly designed and installed structured cabling system. Their concern is that when they plug in and hook up the network devices and the PCs, they will work without experiencing downtime – since 60-70% of the time is related to cabling!

This is your opportunity to educate your customer on the pros and cons of the different types of cabling media from Cat 3, 5, 5E, 6 versus fiber. Alternatively, most cabling installers have little if any knowledge of networking. Once you have a firmer understanding of their network needs both current and for the future, the proper choice cabling type becomes a more viable solution. However, the need for speed or higher bandwidth requirements of these upper Layer devices they are referring to will most probably require the use of both fiber optics along with Cat 5E or 6.

It is very important to know the prime objective or the desired LAN application(s) when designing, installing, and testing a cabling installation. Lets start with a basic understanding of networking terms. Remember, one purpose of standards is interoperability or the ability of different active and passive components, technologies (LANs, MANs, WANs, etc), and multiple manufacturers products to work together.

The ISO or the International Organization of Standards is the internationally recognized organization for (networking) standards to achieve this goal. Standards are presented to them from worldwide members from ANSI or the American National Standards Institute representing the USA, the IEC or the International Electrotechnical Commission for Europe, CSA for Canada and so forth. In fact, you’ve probably seen "ISO 9000 Approved" marketed by companies. This doesn’t ensure a better product than a competitor without it; just that they have in place an ISO verified and approved quality control system.

In 1984, the ISO established the OSI or the Open Systems Interconnect Model of Networking that serves as the foundation for development of many standards for networking systems communications. This model has seven plus zero defined layers for data communications. For the purpose of this question, we will concentrate on just the first three layers and the related active equipment or hardware devices. Each layer serves a particular network function when sending a digital signal (e.g. an email). The most commonly used LAN application is Ethernet. The standards committee for this LAN application is the Institute of Electrical and Electronic Engineers, Inc. designated IEEE 802.3X.

The bottom of this model or Layer One (some will argue this is actually the "Zero" layer) is the ‘Transmission’ layer. "Bits" of digital information are converted from a frame into a pattern of ones and zeros for transport across a medium. In networking application, this is where active devices like a Hub functions – it broadcasts a digital signal out to all those PCs (nodes) that are connected to it and the bandwidth is shared or equally divided.

The description for an Ethernet application defines the type and speed of the signal plus the cabling medium. For example, "10BaseT"; "10" for 10 Megabits per second (Mbps), "Base" for a digital Baseband signal, and "T" for Twisted-Pair.

However, this Layer One also includes all the passive components (requires no power to operate) in a structured cabling system – inclusive of UTP and an optical link segment. Most if not all VDV or low-voltage manufacturers and contractors conduct business in this arena.

What standard is used for defining structured cabling? One in particular is ANSI/TIA/EIA 568B.3. It defines an optical link segment as "… the cable, connectors, and splices (if present) between two optical fiber patch panels". An optical fiber link is comprised of all multiple optical link segments inclusive of the backbone and horizontal cabling.

Traditionally, contractors installed and tested cabling systems to support only Layer One applications and related devices. Concentration was more on the horizontal cabling and with little if any regard to the devices connected to the backbone cabling. Install the cable and connectors, wall plates and patch panels, check for continuity -- perhaps CAT 3 or 5 test parameters on UTP --hopefully document, supply as-builts and leave. Not anymore.

The need for speed or for higher bandwidth along with lower attenuation or signal loss has given rise to fiber optics. Not just from the Service Providers (Outside Plant or WAN /MAN applications), but into the building LANs backbone and the horizontal with fiber-to-the-desk (FTTD). "Riser shafts" in high-rise buildings or "distribution" cabling in data centers and "Co-lo’s" (co-locations) are prime examples of networks with Layer One, Two, and Three active equipment and devices connected to the backbone.

What does this mean? Years ago, cabling/connector manufacturers recognized that the higher layers’ of the ISO/OSI Model of Networking and the related active equipment and hardware devices in the backbone were driving the need for more speed. By establishing a firm foundation of products in the Layer One physical level, their products would be able to support a variety of applications and related networking devices -- both at the workstation and in the backbone. To better attract customers and develop product loyalty, they began offering 7, then 10, then 15-year to lifetime warranties. (Some may interpret this a proprietary, but it’$ the name of the game!) Accurate testing had to follow this path too!

As a contractor, what have you been certifying for your customers these past years? It may have worked fine for those older, lower speed applications and Layer One devices, but will it handle the higher speed demands dictated from Layer Two and Three devices and equipment?

The next level or Layer Two of the ISO/OSI Model of Networking is the ‘Datalink’ layer, which provides a reliable transfer of data called frames across the media. Other factors include physical addressing, network topology, error notification, and flow control. The most commonly used active equipment or device at this level is the Ethernet Switch – instead of a broadcast shared signal like a Hub, it has a dedicated bandwidth path to each user connected to it.

The most common designations of a Switch are; 10/100BaseTX" or "100BaseFX"; auto sensing "10" for 10 megabits per second (Mbps) or "100" for 100 Mbps for the PC’s NIC cards, "Base" for digital Baseband signal, and "T" on Twisted-Pair and "F" on Fiber utilizing a "X" duplex signal.

Accordingly, a 100BaseTX application on UTP is tested to Cat 5 parameters. Fiber is tested by determining the attenuation or signal loss of the fiber link segment and is expressed in dB or decibels. A 100BaseFX device operating on 62.5/125um @ 850 nm has a hefty budget loss of 11.0 dB at a maximum distance of 2000 meters.

In the Ethernet world, have you ever noticed that LAN application speeds-- when comparing the horizontal to the backbone -- jump by a factor of ten? 10 Mbps PCs (usually Pentium I and II’s) with 100 Mbps backbones. Today, 100 Mbps PCs (Pentium III’s) to 1,000 Mbps backbones – 1,000 megabits is also called one Gigabit or one billion bits per second. Bringing "Gignet" to the workstation (Pentium IV’s) is helping drive the need for fiber in the horizontal or FTTD. A Gignet device is identified with a designation either 1000BaseT for UTP and 1000BaseSX and LX for fiber. The "S" designates the shorter and most price-desirable 850nm (nanometer) wavelength and "L" for the longer 1300nm wavelength with both on operating on multimode fiber.

The UTP cable used to support 1000BaseT Gignet is Cat 5E or the recently released Cat 6. These UTP testing parameters are found in the ANSI/TIA/EIA 568B.3 standard, which requires more detailed tests than Cat 5 for 100BaseTX.

This may be too technical, but reference ANSI/TIA/EIA 526-14A for testing multi-mode and 526-7A for single-mode fiber optics. The attenuation or budget loss parameters tighten significantly for Gignet with a maximum budget loss of only 3.2db at a maximum 220 meters or 7,218’ on 62.5/125um 200 MHz* km multi-mode fiber, but 3.9 db @ 550 meters or 18,046 ft. with 50/125um 500 MHz* km fibers using VCSELs light sources. This is still significantly a higher performance than UTP and at comparable costs.

Layer Three of the ISO/OSI Model of Networking is the ‘Network’ layer. It provides connectivity and the best path selection between the sending host and the receiving host. In this level, segments sent from the Layer Four are put into packets. The commonly used active equipment or device in Layer Three is an Ethernet Router.

The driving force for higher speeds usually starts with Layer Three active equipment or devices (routers) connected to the backbone and the MAN or WAN. It takes another 5-7+ years of price migration (mostly related to the cost of the chips) down through Layer Two to finally become a price competitive Layer One device to the workstation – the NIC card. Albeit, the cabling and connectors in Layer One must be able to support the desired application (usually Ethernet) and all related active equipment and devices in all three layers.

Whether it is a Layer Three Router, a Layer Two Switch or a Layer One Hub the relative speeds are increasing! "Moore’s Law" states that we double the processing power of a chip every 18 months. If you take into consideration the networking technologies of tomorrow, these higher speed requirements will be applicable within the cable/connector manufacturer’s warranty timeframe. You better ensure the cabling system is in "compliance" not only to ANSI/TIA/EIA 568B.3 and the cabling/connector manufacturer’s specifications, but also to the requirements of the customer’s network as dictated by IEEE 802.3X. Thus, you can dispel the IT manager’s misperceptions by knowing the limitations of UTP and justify the slightly added costs of incorporating fiber optics.

The goal of the ISO/OSI Model of Networking is to match the capabilities of all components to ensure a reliable transmission. It takes into consideration (1) the strength and characteristics of the launched signal by the transmitter, (2) the fidelity of the signal transmission through the fiber medium (the cabling plant), and (3) the capability of the receiver to capture and decode the signal.

Listen carefully to your customer’s needs! Ensure that the structured cabling system you or someone else has designed, but installed, tested and documented supports not only their desired LAN application for today and tomorrow, but all the associated Layer One, Two and Three active equipment and devices. Though Cat 5E and 6 satisfy today’s requirements in the horizontal, fiber must be considered in the backbone as well as the horizontal or FTTD to be future proof.

A2.  Simplex means one way only on one conductor or one fiber strand and functions like megaphone. A conventional simplex TX signal application is a video camera.

Half duplex means one-way both ways on two conductors or two fibers strands. It functions like a two-way radio. TX "tip" or transmit on one and RX "ring" receive on the other. Typical half-duplex signals applications are used in most fiber LAN Ethernets like 1000BaseSX and LX and UTP Ethernets up to 100BaseTX, and phone lines.

Full duplex means TX and RX both ways simultaneously on one conductor or one fiber strand and functions like a telephone conversation.. It is also called multiplexing by functioning like a telephone conversation. 1000BaseT requires splitting the signal along the four pairs with full duplex processing or multiplexing to achieve Gigabit transmission. Fiber optics uses WDM wave divisional multiplexing and DWDM (Dense WDM) and is typically used by long distance carriers.

A3.  The name of the cabling game is bandwidth! By definition, (optical) bandwidth is "a measure of the information carrying capacity" of a cabling system – the established performance criteria. Bandwidth is a frequency domain measured in Hertz (Hz) or the number of cycles per second (CPS) of an analog or sinusoidal (sine) wave. 1 KHz = 1000 CPS or 1,000 Hz. I MHz = 1 (mega) million CPS or 1,000,000 Hz. 1GHz = 1 (giga) billion CPS or 1,000,000,000 cycles per second.

Remember, that the rating and testing of a cabling system is measured in Hz or Hertz! Cat 3 UTP is factory-rated to 16 MHz, Cat 5E and 6 rated to 100 MHz with a maximum horizontal "Basic or Permanent Link" of 90 meters or 295 ft. distance. UTP field testers perform attenuation and crosstalk measurements across multiple sweep frequencies from 1KHz to the maximum rating of the cable (e.g. Cat 5E @ 100 MHz).

Multi-mode fiber optics is factory-rated from a minimum 160 MHz* km to 500 MHz* km (kilometer or 3,281 ft.) depending on the two operating wavelengths of 850 nm or 1300 nm. Single-mode fiber has theoretical unlimited bandwidth capability, as a direct function of the laser light source. Bandwidth or it’s "rating" is dependent on both optical glass quality (glass manufactures and cabling OEMs) and varies inversely with fiber length. (A long fiber will have less bandwidth available for use than a shorter fiber per MHz* km.). Field-testing fiber bandwidth is not a standard practice.

Alternatively, network applications are digital pulses or a square waves measured in bits per second or bps. 64 kbps is 64,000 bps. 1 Mbps is one (mega) million bps, etc. As mentioned in question #1, the most common type of LAN application is Ethernet. The 100BaseTX or 100BaseFX is 100 Mbps. Notice there is no mention of MHz or GHz.

The electronics or the encoding scheme of the Layer One, Two and Three devices will determine the required performance criteria or rating of the cable. For Gignet, it is now Cat 5E and 6. 50/125um multi-mode fiber is being strongly considered over 62.5/125 um fiber with the rapid growth of VCSEL’s or Vertical Cavity Surface Emitting Lasers over LED’s or Light Emitting Diodes as multi-mode light sources. The reasons are (1) minimum 500 MHz* km ratings, and (2) lower attenuation values, and (3) longer distances (300 meters to 500 meters).

Bandwidth is analogous to the RPM of your racing car engine (MHz) and LAN application speeds (Mbps) to the output of your wheels. If your tachometer reads 3,000 RPM does this mean you are going 3,000 mph? Of course not! The same holds true for a network. The UTP Cat 5E cabling has a bandwidth rating of 100 MHz, yet achieves 1Gignet or 1 billion bits per second. How is this achieved? The "transmission" of the racing car or the network’s digital electronics encoding scheme sophistication determines how fast the car can go (Mbps). A "higher-performance" car or cabling rating allows Layer One, Two, and Three devices to achieve higher digital data stream speeds. And, like racing cars, networks are getting much faster!

No matter how professional the installation looks, it has to be able to support the network it was designed for. Be aware that most customers – and, unfortunately, many of their consultants who design and write the specifications – will not have the foggiest idea what you’re talking about. However, you should be aware of the necessary steps to ensure that your Acceptance Test follows Acceptance Test Procedures to meet this goal. This will be discussed in the next edition.

Q4.   I have seen many references to "Acceptance Testing" in a number of articles and fiber optics technician textbooks.  What is an "Acceptance Test"?

A4.  An Acceptance Test is to verify that the installed optical transmission link segment with all it’s passive components meets the established performance criteria for successful deployment of the network LAN application – whether it is for one or a combination of Layer One, Two, or Three related devices. This verification can be defined as an Acceptance Test.

By definition, a structured cabling system’s Acceptance Test is verifiable test documentation to determine if the transmission cable and components meet and/or is in compliance to an industry (or government) standard. The standard defines both (1) the method of testing (or Acceptance Test Procedures) and, (2) the "acceptable" range between minimum and/or maximum specifications. Unfortunately to a customer, "unacceptable" testing generally isn’t realized until their network system crashes.

A5.  "The fireman is standing in a fire zone with firewalls and the room is equipped with a fire alarm system. He/she is firestopping with fire stop around a fire shield containing plenum-rated cables penetrating a fireproof wall lined with fire-rated plywood coated with two coats of fire retardant paint to provide a fire resistant rating to meet city code.... Give the tech a fiber break!"

Fire containment is an essential part of a professional cabling installers job. This includes installing the proper fire-rated cabling, firestop material applications, and the plywood in the closet. I suggest even taking a picture as documentation of your final firestop installation.

Let’s define the terms as they each have distinct definitions. 

Lets make the fireman a volunteer and his primary job is a professional VDV cabling technician.

A fire zone is a contained area completely enclosed by fire resistant walls, floors, and ceilings.

A firewall helps prevent fire spreading from one fire zone or area to another and runs all the way from the structural floor to the structural ceiling. It also refers to security mechanisms (and PC protection software for that matter). In either case, firewalls are also designed to keep unwanted and unauthorized traffic from a protected network.

A fire alarm system provides a reasonable amount of safety by reducing the probability of injury or loss of life from fire, smoke, and heat in buildings by providing (1) detection, (2) suppression, and (3) notification functions. (e.g. sprinklers)

Firestopping is the actual process of installing specialty material(s) into penetrations of fire-rated barriers to reestablish the integrity of the barrier. You cannot use the same construction material of the penetrated fire barrier or firewall (e.g. mortar, cement, or "mud") to obtain the same fire-rated integrity.

Firestop is the material, device, or assembly of parts in an architectural barrier to prevent vertical or horizontal passage of flames, smoke, water, or gasses the fire barrier. The two types are (1) "Mechanical" that consists of manufactured metal encased elastomeric components pre-sized to and shaped to fit around cable, pipes, conduit and tubes and, (2) "Non-mechanical" that generally are pliable (e.g. caulk, putty, sprays, etc.)

A fire shield is material, device, or assembly of parts used to prevent propagation of flames from one cable system or pathway to an adjacent cable system or pathway – e.g. between two cable trays.

Plenum-rated cable is usually but not always required by local code for plenum-rated ceilings (air chambers) and raised floor environments. The next lower fire rating is riser-rated, followed by general-rated, and the lowest is, ironically, residential-rated. The NEC allows a maximum 50 ft. from the entry point of an outside-rated cable to be fire-rated inside the building (unless it is in rigid metallic conduit).

Penetration is further defined as (1) "Membrane" penetration is through the outside surface of only one side of a fire-rated barrier (e.g. electrical box) and, (2) "Through" penetration penetrates both outside surfaces of a firewall (e.g. conduit, cable tray, etc.)

Fireproof does not support the combustion of flames even under accelerated conditions including masonry, block, brick, concrete and, gypsum board. However, no material is entirely fireproof (The former Chicago’s McCormick Place and the World Trade Centers burned to the ground and were a majority of these mentioned materials!)

Fire-rated cable, doors, walls, or "materials" -- including firestop itself -- is based upon a fire resistance rating (see below).

Fire retardant paint requires two coats over any telecommunications room’s plywood backboard (3/4" X 4’ X 8’ A/C void free). Check the local code, as water-based paint does not usually qualify)

Fire resistant rating is the time in hours or fraction of hours (e.g. gypsum board is 30 minutes per 5/8ths thickness) that a material or assembly of materials will withstand the passage of flames and the transmission of heat when exposed to fire under specified conditions of test and performance criteria.

City code or local code dictates over county, State, and federal codes (unless you are in a military base where they have their own set of codes). The local inspector’s job responsibility is to be cognizant of these other codes and, in particular, the NEC code.

A fire break is not a 10-minute break for coffee, but a material, device, or assembly of parts installed in a cable system (but NOT at a cable penetration of a fire barrier or firewall but in-between) to prevent the spread of fire along a cable.

How To Submit A Question:

e-mail: conradb@crossbowcom.com

Non-published questions will be answered off line within 10 days.

Bo

D.A. Bo Conrad, RCDD

President/Director

CrossBow Communications

ACES Int'l Home Page

Copyright© 05/15/2001 ACES L.L.C.

All rights reserved