Cable Testing 101

The cable tester can also measure several characteristics of the cable, such as attenuation, impedance, Near End Crosstalk (NEXT) and noise, to determine the location of a fault.

Where wiring problems occur

Short circuits and open connections can happen months or years after installation, especially if cheap insulation dries out, becomes brittle and cracks. Conversely, a water-soaked cable won’t carry traffic very well. Sometimes during installation a person will pull a wire around a corner and part of the insulation will scrape off (this condition is called a “shiner”). The cable problem might not immediately manifest itself, but may wait months before causing a network outage. Some wiring problems happen during cable manufacture and connection. Once in a great while, the factory or the installer will put connectors on the cable with the wrong wire leading to the wrong pin, and the new cable won’t work at all. Or the person may mix up the wire pairs by attaching connectors in a way that causes one of the wires to carry a signal that the other wire pair should carry (a condition known as “reversed pairs”).

Even with perfectly manufactured, carefully connected wire, you can inadvertently cause cable problems. In planning a network installation or enhancement, you might overlook the published limitations of the wiring specification. The result is a segment with cables that are too long or with too many nodes in a segment. It is easy to overlook distance and number-of-nodes limitations when you’re concentrating on giving a growing number of people access to the network.

Nominal Velocity of Propagation (NVP)

Because electricity travels at different speeds in different types of cable, the testing device needs to know the NVP for the cable before it can make accurate distance determinations. NVP is expressed as a fraction, or sometimes a percentage of the speed of light in a vacuum, and varies from .60 to .90. For example, Level 3 UTP cable has an NVP of .62, Level 5 UTP an NVP of .72, and token-ring Type 1 an NVP of .78.

Attenuation

The decrease in a signal’s strength over the length of the cable. Longer cables, and higher cable frequencies experience greater attenuation (loss). The cable tester measures attenuation in decibels (dB), with 0 dB signifying no signal loss. As an example, the maximum attenuation allowed for 10Base-T unshielded twisted pair cabling is 11.5 dB.

Impedance

Expressed in ohms, impedance of a cable is the opposition to AC current flow that it would have if the wire was infinitely long. The thickness (gauge) of the copper conductors, the distance between the conductors, and the properties of the insulation (the dielectric material) in the cable all influence impedance. A change in the impedance of a wire somewhere along its length causes reflections inside the cable. Such a reflection is called a discontinuity. Broken wires, short circuits and mixed wire types are examples of discontinuities.

Crosstalk

Electrical interference between wire pairs in twisted-pair cable. NEXT is interference that occurs in the cable adjacent to a connector at either end. The cable tester measures NEXT by transmitting a signal through one wire pair and detecting the resulting spillover of current into the other wire pair. A high degree of signal loss (attenuation) between the wire pairs denotes low NEXT. Low attenuation signifies a problem.

Far End Crosstalk (FEXT) is similar to NEXT, except that a signal is sent from the local end, and any crosstalk is measured at the far end. Equal Level Far End Crosstalk (ELFEXT) is a calculated result, rather than measured, and is derived by subtracting the attenuation of one wire pair from the FEXT that this pair induces in an adjacent pair. Power Sum NEXT (PSNEXT) is also a calculation, derived from an algebraic summation of the individual NEXT effects on each pair by the other three pairs. PSNEXT and ELFEXT are relevant on four-pair transmission schemes such as Gigabit Ethernet.