Basic Cable Testing

This section provides the following information:

A description of LAN cable construction

A description of channel and basic link configurations

Explanations of LAN cable tests

Explanations of the TDR and TDXTM analyser tests and plots

Basic troubleshooting procedures for LAN cable.

LAN Cable Construction
LAN cables have a number of characteristics in common with other types of
electrical cables. All electrical cables have continuity, which means that they serve
as a complete path for electrical current flow. Each end of a cable has some type of
connector for connecting the cable to the appropriate electrical device. Cables with
multiple wires usually have a pin assignment that describes how the wires are
arranged in the connectors.

Cables are designed to perform best in specific applications. For example, power
cables are designed to minimise power losses at frequencies of 50 Hz or 60 Hz.

LAN cables are designed to minimise signal distortion at higher frequencies.
Two types of cables are designed for use with LAN systems: twisted pair cable
and coaxial cable.

Twisted Pair Cable
Twisted pair cable consists of wire pairs that are twisted together, as shown in
Figure 7-1. The wires are twisted to minimise crosstalk between the cable pairs.

Figure 7-1. Twisted Pair Cable Construction

Each cable pair forms a complete electrical path for signal transmission. The
currents flowing through the wires in each pair are equal, but flow in opposite
directions. These currents produce electromagnetic fields that could transmit
electrical noise to nearby wires. However, the fields surrounding the two wires
have opposite polarities. Twisting the wires together causes the fields to cancel
out, which minimises the electrical noise, or crosstalk, generated by each cable
pair.

There are two types of twisted pair cable: shielded twisted pair (STP) and
unshielded twisted pair (UTP). STP cable, which may also be referred to as ScTP
(screened twisted pair) or FTP (foil-screened twisted pair), contains a conductive
shield that is electrically grounded to protect the wires inside from electrical noise.
STP cable is more expensive and more difficult to install than UTP,

Figure 7-2 shows the pin connections and wire colours for correct wiring to TIA
568A and 568B standards.




Figure 7-2 EIA/TIA RJ45 Connectors

The 568A and 568B wiring standard have the same electrical performance, 
the two standards should not be mixed in the same network because of the
risk of wiring errors.

Coaxial Cable

Coaxial cable consists of a conductor surrounded first by an insulating material,
then by a braided conductive sheath, as shown in Figure 7-3. In LAN applications,
the sheath is electrically grounded and serves as a shield that protects the inner
conductor from electrical noise. The shield also helps eliminate signal loss by
keeping the transmitted sign, I confined to cable.

Coaxial cable can carry a wider range of frequencies and can be used in greater
lengths than twisted pair cable. However, coaxial cable is more expensive than
twisted pair.



Figure 7-3 Coaxial Cable Construction

Basic Link and Channel Connections

The cable links you test may or may not include equipment patch cables and extra
transition connections in the telecommunications closet and work area, For
example, cable installers are often responsible only for the permanent cable
installed between the closet and the first wall outlet in the work area. This segment
of cable is the basic link, shown in Figure 7-4. As defined in TSB-67, the basic
link consists of up to 90 m of horizontal cable, one transition connector at each
end, and two test equipment patch cables of no more than 2 m each,




Figure 7-4 Basic Link Test Connections

A channel includes the transition connectors and equipment patch cables added to
a basic link segment. The channel should be tested from end to end to verify the
performance of all the components. In this case, you use the equipment patch
cables to connect your test tool to the channel, as shown in Figure 7-5. TSB-67
defines the channel as a basic link plus one extra transition connector at each end
and up to 10 meters of equipment patch cables. Because of the extra connectors
and patch cables, the test limits for a channel are looser than those for the basic
link,
A channel with just one connector at each end resembles a basic link: however,
you would use a channel test standard if you are using the network equipment' s
patch cables to connect to your test tool,



Figure 7-5 Channel Test Connections


Attenuation

Attenuation is a decrease in signal strength over the length of a cable, as shown 
in Figure 7-6.




Attenuation is caused by a loss of electrical energy in the resistance of the cable
wire and by leakage of energy through the cable's insulating material. This loss of
energy is expressed in decibels. Lower attenuation values correspond to better
cable performance. For example, when comparing the performance of two cables
at a particular frequency, a cable with an attenuation of 10 dB performs better than
a cable with an attenuation of 20 dB.

Cable attenuation is determined by the cable's construction, length, and the
frequencies of the signals sent through the cable. At higher frequencies, the skin
effect and the cable's inductance and capacitance cause attenuation to increase.

Noise

Electrical noise is unwanted electrical signals that alter the shape of the signals
transmitted on a LAN cable. Figure 7-7 shows an example of how noise affects the
shape of an electrical signal known as a sine wave. Signals that are severely
distorted by noise can cause communication errors in a LAN.



Figure 7-7. Sources of Electrical Noise

Electrical noise is generated by any device that uses or generates voltages that vary
over time. Varying voltage generates a varying electromagnetic field, which
transmits noise to nearby devices in the same way that a radio transmitter transmits
signals to your radio. For example, fluorescent lights, which use 50 Hz or 60 Hz ac
(alternating current) power, continuously radiate a 50 Hz or 60 Hz signal that can
be received by nearby devices as electrical noise.

LAN cables act as antennas that can pick up noise from fluorescent lights, electric
motors, electric heaters, photocopiers, refrigerators, elevators, and other electronic
devices. Coaxial cable is far less susceptible to noise than twisted pair cable
because it is shielded by a conductive sheath. The sheath is electrically grounded
to prevent noise from reaching the inner conductor.

The test tool measures impulse noise on the cable under test. Impulse noise is
abrupt "spikes" of interference. This noise is caused by electronic devices that run
intermittently, such as elevators, photocopiers, and microwave ovens. You can
monitor impulse noise in the test tool's MONITOR mode. The test tool counts
noise spikes that have an amplitude greater than the selected impulse noise


Characteristic impedance

Characteristic impedance is the impedance that a cable would have if the cable
were infinitely long. Impedance is a type of resistance that opposes the flow of
alternating current (ac). A cable's characteristic impedance is a complex property
resulting from the combined effects of the cable's inductive, capacitive, and
resistive values. These values are determined by physical parameters such as the
size of the conductors, distance between conductors, and the properties of the
cable's insulation material.

Proper network operation depends on a constant characteristic impedance
throughout the system's cables and connectors. Abrupt changes in characteristic
impedance, called impedance discontinuities or impedance anomalies, cause
signals reflections, which can distort signals transmitted through LAN cables and
cause network faults.

Minimising Impedance Discontinuities

Characteristic impedance is usually altered slightly by cable connections and
terminations. Sharp bends or kinks in LAN cable can also alter the cable's
characteristic impedance. Networks can operate with small discontinuities because
the resulting signal reflections are small and are attenuated in the cable. Larger
impedance discontinuities can interfere with data transmission. Such
discontinuities are caused by poor electrical contacts, improper cable terminations,
mismatched cable or connector types, and by disturbances in the twisting pattern
of twisted pair cable.

You can avoid problems with impedance discontinuities by observing the
following precautions during installation:

Never mix cables with different characteristic impedance's (unless you use
special impedance-matching circuitry).

· Always terminate coaxial cables with a resistance equal to the cable's
characteristic impedance. The terminating resistance prevents signal
reflections by absorbing the signal's energy.

· When untwisting cable pairs to install connectors or make connections at
punch-down blocks, make the untwisted sections as short as possible.

· Do not make sharp bends or kinks in the cable. Check the cable manufacturer's
specifications for the minimum bend radius.

· Handle LAN cable carefully during installation. Do not step on the cable or
pinch it with tight cable ties.

Crosstalk

Crosstalk is undesirable signal transmission from one cable pair to another nearby
pair. Like electrical noise from outside sources, crosstalk can cause
communication problems in networks. Of all the characteristics of LAN cable
operation, crosstalk has the greatest effect on network performance.
The test tool makes two crosstalk measurements: NEXT and ELFEXT.

NEXT

The test tool measures NEXT (near-end crosstalk) by applying a test signal to one
cable pair and measuring the amplitude of the crosstalk signals received by the
other cable pairs. The NEXT value, expressed in decibels, is computed as the
difference in amplitude between the test signal and the crosstalk signal measured at
the same end of the cable. Higher NEXT values correspond to less crosstalk and
better cable performance.

All signals transmitted through a cable are affected by attenuation. Because of
attenuation, crosstalk occurring at the far end of a cable contributes less to NEXT
than crosstalk occurring closer to the signal source. To verify proper cable
performance, NEXT is measured from both ends of the cable.

FEXT and ELFEXT

FEXT (far-end crosstalk) is the difference between the amplitude of a far end test
signal applied to one wire pair and the resulting crosstalk at the near end on a
different wire pair. Like NEXT, FEXT is expressed in decibels, and higher FEXT
values correspond to better cable performance.
ELFEXT (equal level far-end crosstalk) is the difference between the FEXT and
the attenuation of the disturbed wire pair. Like ACR, ELFEXT is a type of signal-
to-noise ratio that indicates the transmission quality of a cable link. The name
"Equal Level FEXT" refers to the fact that all FEXT signals are attenuated equally.
Figure 7-8 illustrates this important difference between NEXT and FEXT.
Crosstalk signals received far from a transmitter contribute little to NEXT because
they are attenuated over a long distance. But all FEXT signals are attenuated by
the same amount, so they all contribute equally to crosstalk at the far end.
Because all FEXT signals travel the same distance, they tend to add up in phase.
This means that the signals always add up to a worst-case FEXT value.
There can be a difference between the NEXT and FEXT of a link, particularly in
the connecting hardware. This difference is due to the nature of the capacitive and
inductive currents that cause crosstalk. At a signal source (the near end) these
currents can subtract. If the currents subtracted at the near end, they add up at the
far end. Thus, a connector that attains high NEXT by balancing the two currents
may have very poor FEXT performance.
In cable, the inductive crosstalk component is very low. Most of the crosstalk is
due to the capacitive component. Because the difference between the two
components is nearly equal to the capacitive component, NEXT and FEXT are
nearly the same in a cable.
Because of how FEXT signals add up along a cable link, good ELFEXT
performance is critical for systems that transmit data over multiple wire pairs.
Examples of these systems are 100BASE-T4, 100VG-AnyLAN, and 1000BASE-T
(Gigabit Ethernet).





Figure 7-8. How FEXT Signals are All Equally Attenuated

Locating NEXT and ELFEXT Problems

If the test tool reports ,NEXT or ELFEXT failure on a cable pair, you can use the
TDX analyser to locate the source of the crosstalk problem.

Like the TDR results, the TDX analyser results are presented in both a list and a
plot format. The list format shows the cable pairs tested, the peak magnitude of
crosstalk detected on the pairs, and the distance to the peak magnitude.

The TDX analyser plot shows the Location and magnitude of all crosstalk sources
detected on the cable. An example of a TDX analyser plot from a test on a good
twisted pair cable is shown in Figure 7-9.

The plot's horizontal scale represents the distance along the cable under test. In the
example plot, the cursor is placed at a small crosstalk source caused by a connector
77 ft (23.5 m) from the test tool.

Figure 7-9. A TDX Analyser Plot

The vertical scale represents the magnitude of the crosstalk detected. The crosstalk
levels shown on the plot are adjusted to compensate for cable attenuation. Without
this adjustment, the peak on the right side of the plot (farther from the test tool)
would appear much smaller. The adjusted plot helps you identify crosstalk sources
because you can use the vertical scale to measure crosstalk magnitudes plotted at
any distance from the test tool. You can also compare the relative magnitudes of
crosstalk peaks to determine the largest sources of crosstalk on the cable.

The vertical scale increases logarithmically. The scale units are arbitrary. A level
of 50 represents a crosstalk magnitude that is close to causing a cable failure. The
level of 100 is approximately 20 times greater than the level of 50. A level of 100
represents an extremely high level of crosstalk that is typically due to split pairs.
Cables or other hardware that cause crosstalk levels of 100 or above are considered
unusable. Crosstalk levels near O are considered to be inconsequential.

The TDX plot from a cable that failed a NEXT test may show one or more peaks
of crosstalk greater than a level of 50. A failure can also be caused by a crosstalk
level less than 50 if the level is sustained over a substantial distance of the cable.

Split Pairs and NEXT

A split pair occurs when one wire from a cable pair is twisted together with a wire
from a different cable pair. Split pairs most frequently result from miswires at
punchdown blocks and cable connectors. Figure 7-10 shows an example of split
pair wiring. Notice that the pin-to-pin connections across the cable are correct, but
the pairs twisted together do not form a complete circuit.




Figure 7-10. Split Pair Wiring

Split pairs cause severe crosstalk because the signals in the twisted pairs come
from different circuits. The high crosstalk levels produced by split pairs cause low
NEXT values during cable tests. If the NEXT value is low enough, the test tool
reports a split pair during the wire map test.

The tool may also report split pairs if you test untwisted cable, such as ribbon
cable or untwisted telephone line.

If the tool reports split pairs when you are testing a cable made of multiple
segments, you can determine which segment has the split pair by running the TDX
analyser. The TDX analyser plot will show a large value of crosstalk starting at a
distance that corresponds to the beginning of the segment with the split pair.


Minimising Crosstalk

Crosstalk problems are minimised by twisting together the two wires in each cable
pair. Twisting the two wires together causes the electromagnetic fields around the
wires to cancel out, leaving virtually no external field to transmit signals to nearby
cable pairs.
You can avoid problems with crosstalk by observing the following precautions
during installation:
When untwisting cable pairs to install connectors or make connections at
punch-down blocks, make the untwisted sections as short as possible.
Be attentive when making wiring connections. Wiring errors that cause split
pairs create severe crosstalk problems.
Do not make sharp bends or kinks in the cable. Check the cable manufacturer' s
specifications for the minimum bend radius.
Handle LAN cable carefully during installation. Do not step on the cable or
pinch it with tight cable ties.

Power Sum Values

Power sum values, such as PSNEXT, PSELFEXT, and PSACR, show how much a
wire pair is affected by the combined interference from the other pairs in the cable.
Power sum NEXT values show how much a pair is affected by the combined
NEXT from the other pairs. PSELFEXT values show the effects of the combined
FEXT. PSACR values show the ratio of a pair' s attenuation to the combined
NEXT of the other pairs.

Good power sum performance is important for higher speed networks, such as
1000BASE-T, that transmit data in parallel over multiple wire pairs. Though a
cable link may not be installed for use with parallel transmission systems,
verifying a link's power sum performances helps ensure successful upgrades in the
future.

Propagation Delay and Delay Skew

Propagation delay is the time it takes an electrical signal to travel the length of a
wire. A wire pair's delay depends on the pair's length, twist rate, and electrical
properties, such as the type of insulating material used around the copper
conductors. Delays typically measure in the hundreds of nano seconds. (One nano
second is one-billionth of a second, or 0.000000001 second.)

The propagation delays of wire pairs in a link can differ slightly because of
differences in the number of twists and in the electrical properties of each pair. The
delay differences between pairs is called delay skew. Delay skew is a critical
parameter for high-speed networks that use parallel data transmission, which
involves transmitting data bits simultaneously over multiple wire pairs. If the delay
skew between the pairs is too great, the bits arrives out of synchronisation and the
data cannot be properly reassembled. Systems that use parallel transmission
include 100BASE-T4, IOOVG-AnyLAN, and 1000BASE-T (Gigabit Ethernet).

Though a cable link may not be intended for parallel transmission, testing for
delay skew helps ensure that the link will support upgrades to high-speed
networks.

Nominal Velocity of Propagation (NVP)

NVP is the speed of a signal through a cable relative to the speed of light. In a
vacuum, electrical signals travel at the speed of light. In a cable, signals travel
slower than the speed of light. Typically, the speed of an electrical signal in a
cable is between 60 % and 80 % of the speed of light. Figure 7-11 shows how the
NVP percentage is calculated.




Figure 7-11. How NVP is Calculated

NVP values affect the limits on cable length for Ethernet systems because Ethernet
operation depends on the system's ability to detect collisions in a specified amount
of time. If a cable's NVP is too low or the cable is too long, signals are delayed
and the system cannot detect collisions soon enough to prevent serious problems in

NVP and Length Measurements

Length measurements depend directly on the NVP value entered for the selected
cable type. To measure length, the test tool first measures the time it takes for a
test pulse to travel the length of the cable. The test tool then calculates cable length
by multiplying the travel time by the signal speed in the cable.
Because the test tool uses the length measurement to determine cable resistance
limits, the NVP value also affects the accuracy of resistance measurements.

NVP Calibration

The NVP values specified for standard cables are included in the cable
specifications stored in the test tool. These values are accurate enough for most
length measurements. However, the actual NVP for one cable type can vary up to
20% between batches because of variations in the manufacturing process.
Therefore, if accurate length measurements are critical to your installation or
testing process, you should determine the actual NVP value for each spool of
cable. Determining the NVP value involves measuring a known length of cable
and adjusting the test tool's length measurement to match the known length. As
you adjust the length measurement, the NVP value changes accordingly. The
calibration procedure is explained in "NVP Calibration" in Chapter 6.

Time Domain Reflectometry (TDR)

TDR is a measurement technique used to determine a cable's length and
characteristic impedance and to locate faults along the cable. TDR is sometimes
referred to as cable radar because it involves analysing signal reflections in the
cable.
If a signal travelling through a cable encounters an abrupt change in the cable's
impedance, some or all of the signal is reflected back to the source. The timing,
size, and polarity of the reflected signals indicate the location and nature of
impedance discontinuities in the cable.

Reflections from Opens

An open, or break, in the cable represents an abrupt increase in the cable's
impedance. The impedance of an open is nearly infinite. In an open cable, a
signal's energy is not dissipated by a terminating impedance, so the signal bounces
back towards the source. This reflection appears at the source with the same
amplitude and polarity as the original signal, as shown in Figure 7-12. By
measuring the amount of time taken for the reflected pulse to return, the test tool
can determine the location of the open in the cable.




Figure 7-12. Signals Reflected from an Open, Shorted, and Terminated Cable

Reflections from Shorts

A short represents an abrupt decrease in the impedance between the two
conductors in a cable. A short is caused when the insulation surrounding a cable's
wires is damaged, allowing the wires to touch each other. The result is a near-zero
impedance connection between the conductors.

A short also causes signal reflections, but in a manner opposite of an open. In a
shorted cable, the signal's energy is not dissipated because the short's impedance
is close to zero. The signal is reflected back to the source, where it appears with
the same amplitude but a polarity opposite of the original signal, as shown in
Figure 7-12.

Reflections from Other Discontinuities

Reflections are also caused by impedance discontinuities that measure somewhere
between infinite and zero impedance. These discontinuities can be caused by
mechanical stress that damages cable wire or insulation without causing a
complete open or short. They can also be caused by cable mismatches and faulty
contacts at connectors or punch-down blocks.

A cable fault that has an impedance higher than the cable's characteristic
impedance reflects a signal that has the same polarity as the original signal. If the
fault is not a complete open, the amplitude of the reflected signal will be less than
the original signal.

If the fault' a impedance is lower than the cable's characteristic impedance, but is
not a complete short, the reflected signal will have the opposite polarity and less
amplitude than the original signal.

Cable Termination

Because signal reflections can distort the shape of communication signals, the
unused ends of cable segments must be terminated to prevent reflections. The
terminating device is a resistor with a value equal to the cable's characteristic
impedance. A signal reaching the terminator is neither reflected nor passed: the
signal is absorbed and dissipated by the terminating resistance.

Because the test tool relies on signal reflections to determine cable length, the tool
cannot measure the length of properly terminated cables.

Interpreting the TDR Plot

The TDR plot has a horizontal scale that represents distance and a vertical scale
that represents the percentage of reflection relative to the original signal, as shown
in Figure 7-13.




Figure 7-13. A TDR Plot

Notice that the reflection percentages can be positive or negative. A positive value
indicates that the polarity of the reflection is the same as the polarity of the
original signal. As discussed earlier, positive reflections are caused by abrupt
increases in the cable's impedance, such as those caused by mismatches in cable
types, poor connections, or breaks in the cable.

A negative reflection percentage indicates that the polarity of the reflection is the
opposite of the original signal. Negative reflections are caused by abrupt decreases
in the cable's impedance, such as those caused by mismatches in cable types or
shorts on the cable.

The results plotted in Figure 7-13 are the results from a TDR test on pair 4,5 in a
good twisted pair cable. The results show a positive anomaly caused by the open
end of the cable at about 219 feet. Note that the left edge of a TDR reflection
represents the location of the anomaly on the cable, while the peak of the reflection
represents the size of the anomaly.

The test tool's TDR plot includes a movable cursor with a readout that displays the
cursor' a position and the reflection percentage of the anomaly at the cursor' s
position. You can move the cursor left and right by pressing the O O keys on the
test tool. Figure 7-13 shows the cursor positioned near the beginning of an
anomaly.

ACR

ACR (attenuation to crosstalk ratio) is the difference between NEXT in dB and
attenuation in dB. The ACR value indicates how the amplitude of signals received
from a far-end transmitter compares to the amplitude of crosstalk produced by
near-end transmissions. A high ACR value means that the received signals are
much larger than the crosstalk. In terms of NEXT and attenuation values, a high
ACR value corresponds to high NEXT and low attenuation.

Figure 7-14 shows a plot of NEXT and attenuation limits, along with the resulting
ACR plot. Notice that the ACR is lower where NEXT and attenuation values are
near each other.



RL

RL (return loss) is the difference between the power of a transmitted signal and the
power of the signal reflections caused by variations in the cable's impedance. A
RL plot indicates how well a cable's impedance matches its rated impedance over
a range of frequencies. High RL values mean a close impedance match, which
results in a large difference between the powers of transmitted and reflected
signals. Cables with high RL values are more efficient at transmitting LAN signals
because little of the signal is lost in reflections.

Good return loss is especially important for high-speed systems, such as Gigabit
Ethernet (ZEEE 802.3x), that transmit full-duplex (bi-directional) data over
individual pairs. Full-duplex transceivers use directional couplers to distinguish
between incoming and outgoing signals. If a cable has poor return loss, the
couplers might interpret the reflected signals as incoming data, resulting in data
errors.

Troubleshooting Basics

Troubleshooting LAN cable installations is most often required during cable
installation or modification. When cable is handled carefully and installed
correctly, it usually operates trouble-free for years.

Finding Cable Faults

A general rule for finding cable faults is as follows: With very few exceptions,
faults occur at cable connections. Cable connections include telecommunication
outlets, patch panels, punchdown blocks, and transition connectors.
Connections are the most likely places for faults for at least three reasons: (1)
Connections always alter the impedance of the transmission path, (2) connections
are likely places for faults caused by wiring errors and faulty or incompatible
hardware installation, and (3) connections always cause some crosstalk due to the
untwisting of cable pairs.
When cable is handled carelessly, faults can occur in the middle of the cable.
These faults can happen when the cable is stepped on, sharply bent, pinched by
cable ties or other hardware, or otherwise stressed.
A general procedure for finding cable faults (excluding sources of noise and traffic
faults) involves the following steps:
1. Run an Autotest on the cable.
2. If the Autotest fails, press ~j for specific information on the fault.
3. Inspect the cable for the suggested fault at the location indicated by the
diagnostics display.
4. Repair any faults you find. To quickly check your repair, run the test that
failed as a Single Test.
5. Run the Autotest again to verify the cable's performance.

Reproduced by kind Permission - Fluke Networks, Inc