Copyright © 1996

Tutorial Index
by Lee Chae

Lesson 93: Cable Testing

Understanding the electrical impulses of UTP
cabling sheds light on cable performance.

With the exception of wireless systems, networks rely on cable

to conduct data from one point to another. In the case of
copper-wire-based, unshielded twisted pair (UTP) cabling,
data is conveyed in the form of electric, digital signals.
Because these signals are essentially bursts of electricity, the
electrical characteristics of the cable itself greatly affect the
integrity of the signal being transmitted. A bad length of cable
or a poor cable installation can result in signal loss or
distortion, and consequently, network failure.

To minimize such occurrences, cable vendors test their cables

to guarantee performance. However, this doesn't make their
products fault-proof; bad cabling does exist. In some cases,
the error lies in improper cable installation. Network managers
can use cable testers to ensure that a cable can conduct
signals correctly. They can also use cable testers to verify if a
cable is properly installed and to troubleshoot faulty cable.

A solid grounding in the electrical properties of UTP is a good

way to learn how cable can affect the performance of a
network.

THE COMPUTER CIRCUIT

A network can be broken down in simplistic fashion into an

electrical circuit metaphor. In this case, a network essentially
comprises energy sources, conductors, and loads. An energy
source is a network device that transmits an electrical signal
(data). The conductors are the wires that the signal travels
over to reach its destination, which is usually another network
device. The receiving device is known as the load. In its
entirety, the connected network is a completed circuit.

When an energy source transmits a signal, it is outputting an

1 of 5
electric charge onto the conductor by applying voltage to the
completed circuit. Voltage is measured in volts. The voltage
propels the charge across the cable, and the flow of the charge
is known as a current, which is expressed in amperes, or
amps.

In the computer world, the electric signal transmitted by an

energy source is a digital signal known as a pulse. Pulses-in
the form of a series of voltages and no voltages-can be used to
represent a series of ones and zeros. Digital pulses form bits,
and a series of eight bits creates the almighty byte.

The key to a successful signal transmission is that when a load

receives an electrical signal, the signal must have a voltage
level and configuration consistent with what had been
originally transmitted by the energy source. If the signal has
undergone too much corruption, the load won't be able to
interpret it accurately.

In short, a good cable will transfer a signal without too much

fudging of the signal, while a bad cable will render a signal
meaningless.

PROPERTY LIMITS

Due to the electrical properties of copper wiring, the signal will

undergo some corruption during its transit. Obviously, signal
corruption within certain limits is acceptable. Once the
electrical properties exceed the limits prescribed to a certain
cable type, the cable is no longer reliable and must be
replaced or repaired.

As a signal propagates down a length of cable, it loses some

of its energy. So, a signal that starts out with a certain input
voltage, will arrive at the load with a reduced voltage level. The
amount of signal loss is known as attenuation, which is
measured in decibels, or dB. If the voltage drops too much, the
signal may no longer be useful.

Attenuation has a direct relationship with frequency and cable

length. The higher the frequency used by the network, the
greater the attenuation. Also, the longer the cable, the more
energy a signal loses by the time it reaches the load.

A signal loses energy during its travel because of electrical

properties at work in the cable. For example, every conductor
offers some resistance to a current. Resistance, which is
measured in ohms, acts as a drag on the signal, restricting the
flow of electrons through the circuit and causing some of the

2 of 5
signal to be absorbed by the cable. The longer the cable, the
more resistance it offers.

Due to its electrical properties, a cable not only resists the

initial flow of the current, it opposes any change in the current.
The property that forces this reaction is called reactance, of
which there are two relevant kinds: inductive reactance and
capacitive reactance.

In an inductive reaction, a current's movement through a cable

creates a magnetic field. This field will induce a voltage that
will work against any change in the original current.

Capacitance is a property that is exhibited by two wires when

they are placed close together. The electrons on the wires act
upon each other, creating an electrostatic charge that exists
between the two wires. This charge will oppose change in a
circuit's voltage. Capacitance is measured in farads or
picofarads (see table).

Reactance can distort the changes in voltage that signify the

ones and zeros in a digital signal. For example, if the signal
calls for a one followed by a zero, reactance will resist the
switch from voltage to no voltage, possibly causing the load to
misidentify what the voltage represents.

IMPEDING PROGRESS

When you combine the effects of resistance, inductance, and

capacitance, the result is the total opposition to the flow of the
current, which is known as impedance and is measured in
ohms.

It's important for components of a circuit to have matching

impedance. If not, a load with one impedance value will reflect
or echo part of a signal being carried by a cable with a
different impedance level, causing signal failures. For this
reason, cable vendors test their cables to verify that
impedance values, as well as resistance and capacitance
levels, comply to standard cable specifications.

It's also important for the impedance of a cable to be uniform

throughout the cable's length. Cable faults change the
impedance of the cable at the point where the fault lies,
resulting in reflected signals.

Cable testers use this trait to find cable faults. For example, a
break in a wire creates an "open circuit," or infinitely high
impedance at that point. When a high frequency signal emitted

3 of 5
from a cable tester encounters this high impedance, it will
reflect back towards the tester like an ocean wave bouncing off
a seawall. Similarly, a short circuit represents zero impedance,
which will also reflect a high frequency signal, but with an
inverted polarity.

The cable testing device can then tell you approximately how
far down the cable the fault lies. The formula for this feature
uses a cable value known as nominal velocity of propagation
(NVP), which is the rate at which a current can flow through the
cable, expressed as a percentage of light speed. The cable
tester multiplies the speed of light by the cable's NVP and by
the total time it takes the pulse to reach the fault and reflect
back to the tester, and divides it by two, for the one-way
distance.

The same concept is used to check the electrical length of a

cable installation. In this case, you must make sure not to
terminate one end of the cable. The open end will register as
infinite impedance and reflect a pulse back to the tester. Again,
this response time is plugged into the formula to estimate the
overall electrical length of the wire.

As an aside, cable testers can't check the first 20 feet or so of

a cable. The reason for this blind spot is that a pulse
transmitted by the tester will be reflected back to the device
before it is entirely transmitted. Thus, the tester can't get an
accurate reading.

MIXING SIGNALS

Finally, the successful transmission of a signal can be

jeopardized by noise, which can introduce false signals, or
noise spikes, at different frequencies on a wire. A load may
interpret a noise spike as part of a digital signal, distorting the
original content of the signal. Common sources of noise spikes
include AC lines, telephones, and devices such as radios,
microwave ovens, and motors. Some cable testers test for
noise, running tests at different frequencies.

Another type of interference is called crosstalk, or more

specifically, near-end crosstalk (NEXT). As mentioned, when a
current moves through a wire, it creates an electromagnetic
field. This field can interfere with signals traveling on an
adjacent wire. To reduce the effect of NEXT, wires are
twisted-thus the name twisted pair. The twisting allows the
wires to cancel each other's noise.

The risks of NEXT are highest at the ends of a cable because

4 of 5
wire pairs generally don't have twists at their ends, where they
enter connectors. If the untwisted end length is too long, NEXT
levels can rise to distortional levels.

Also, due to attenuation, signals are strongest when they are

transmitted, and weakest when they arrive at their destination.
So, the magnetic field of a signal being transmitted from a
device on one wire may overwhelm a signal arriving at the
same device on the wire's pair.

NEXT is measured in decibels, which represent a ratio of a

signal's strength to the noise generated by crosstalk (see
table). The stronger the signal and weaker the noise, the
higher the NEXT value. For this reason, a high NEXT reading
is good. Low NEXT readings, which indicate high crosstalk
interference, can mean the cable is terminated improperly.

THE NOT-SO-FINAL WORD

In the past, the topic of cable testing and performance has

been a contested one, due mostly to the absence of accepted
testing and performance standards. Recently, the EIA/TIA
finalized TSB-67, which defines what it calls "transmission
performance specifications for field testing of unshielded
twisted pair cabling systems." These specifications cover
Category 3, 4, and 5 UTP. TSB-67 also details specifications
for cable tester performance-something that has been
noticeably absent.

Although the document is doing much to calm some pretty

turbulent waters, other questions concerning cable
performance and testing loom on the horizon. For example,
TSB-67 addresses specifications for Category 5 cabling up to
100MHz. How does this relate to 155MHz ATM, which is
supposed to run over Category 5?

While questions such as these may continue to cause

confusion, the only surety may be the fundamental electrical
properties at work in the cable.

5 of 5