Cable Testing

Copper and Fiber Cable Tests

These notes guide you through the understanding of cable testing standards.

Introduction
Cable Testing Standards
        Attenuation - Insertion Loss
        Cable Length
        Propagation Delay
        Propagation Delay Skew
        Return Loss
        Wire Map
        NEXT
        PSNEXT
        FEXT, ELFEXT and PSELFEXT
        PSELFEXT
        Optical Link Loss Budget
Summary

Introduction

On completion of these notes you should...

• understand that various factors such as noise, attenuation, impedance discontinuities, crosstalk and other factors can affect signal quality on cable
• understand that cables have to meet testing standards
• be aware that to meet the TIA/EIA-568-B standard copper cable has to pass ten tests
• be aware that fiber cable must also undergo quality tests

Cable Testing Standards

All cables are subject to some degree of interference from factors such as noise, attenuation, impedance discontinuities, crosstalk, EMI and RFI. These factors can interfere with the quality of a signal. Therefore, all cables need to be tested to ensure that interference is within acceptable bounds and signal quality is maintained to an acceptable degree.

The TIA/EIA-568-B standard specifies that copper cable has to pass ten tests to be acceptable for use on modern networks. These tests are the subject of these notes.

Attenuation - Insertion Loss

Attenuation is a measurement of the loss in strength of a signal at the receiving end. All electromagnetic signals lose strength as they travel away from their source and LAN signals are no exception. This loss of signal energy arises from resistance to electrical signals from the transmission medium and from some of the energy leaking through the cable material.

When a signal is transmitted down a cable the amplitude of the signal slowly decreases as the resistance of the medium saps the signals energy. If the cable is too long between the source and target then the signal attenuates too far and the receiving end may not be able to read the signal. Repeaters, hubs and other devices can regenerate a signal.

Temperature and frequency can influence the degree to which attenuation occurs. The higher the temperature or frequency the more a signal attenuates. A signal transmitted on 10BASE-T cables may pass an attenuation test but may fail if transmitted on a cable of the same length on a 100BASE-Tx network, simply due to an increase in attenuation from the higher transmission frequency. Attenuation must therefore be measured over the applicable frequency range

Stated another way, the longer the cable or the higher the transmission frequency, the more a signal attenuates.

Signal energy loss can also arise from poorly terminated connectors. Insertion loss is the term used for the loss in signal energy from the resistance of the cable material, from impedance of poorly terminated connectors and loss through the cable material.

If only one or two wires in a cable show high insertion loss then this may be due to poor termination. If all wires in a cable show high insertion loss then this may be due to the cable being too long.

Cable Length

With twisted pair cable, physical length is the length of the outer cable but the true length - as measured by a tester - is the length of the conductor. This is always longer due to the twisting of the wires.

On CSMA/CD networks it is important to limit cable length. One of most important reasons for this is control. A sending device needs to know if a collision has occurred while it is signalling. It can only know a collision has occurred if it detects collision fragments returning along the wire before it has stopped transmitting. Since the time for a signal to be sent, collide and return depends on cable length, then a cable run must not be too long or collisions would be undetectable.

Propagation Delay

Propagation delay is a measurement of the time taken for a signal to propagate from one end of a cable circuit to another. This is measured in nanoseconds (ns). Most structured wiring standards accept a maximum horizontal delay of 570 ns.

As mentioned above, a cable that is too long will cause problems on CSMA/CD networks because collisions will be undetectable. Simply put, if the cable length is too long, the prorogation delay measurement will be unacceptable and your cabling arrangement will contravene structured wiring standards.

Propagation Delay Skew

Propagation Delay Skew is a measurement of the difference in propagation delay between the fastest and slowest wire pairs in a cable. In twisted pair cables, it is unlikely the wires inside the cable are twisted by the exact same amount. The wires will have slightly different lengths due to this difference in twisting and so will have slightly different prorogation delay times.

Most structured wiring standards accept a maximum delay skew of 50 ns over 100 metres of cable. In other words, the difference between the propagation delay of the fastest and slowest wire in the cable can be up to 50 ns over 100 metres.

You may wonder why the concern over this. On high-speed networks like Gigabit LANs, signals are transmitted down all four wire pairs at the same time. If the signals reach the other end at different times, the receiver may have difficulty recombining the signals.

Return Loss

Return Loss is the measurement of signal echo and reflections caused by impedance discontinuities along the cable circuit. Return loss is measured in decibels (DB).

Impudence discontinuities along the cable circuit may cause part of the signal to be reflected back toward the transmitter, like an echo. This may cause jitter at the transmitter end. Part of the signal may also be reflected in the same direction as the original signal. This may cause difficulties at the receiving end because the original signal may be corrupted by the reflected signal. Impedance discontinuities are often caused by poorly terminated ends where wire pairs are incorrectly attached to the connector.

You should note that impedance discontinuities also cause some signal attenuation since the original signal will lose energy if any part of the signal is reflected back.

Wire Map

Each wire in a cable must be properly connected to the correct pin at both ends. The diagram below shows the correct wire arrangement for a T568-B cable.

One wire connects to pin1 at both ends, the next wire connects to pin2 at both ends and so on.

This is the pin arrangement for a straight through cable wired to the T568-B standard. There are other pin arrangements for other types of cables, such crossover and rollover cables.

The wire map test is used to identify cable wiring errors such as...

• Open Circuits
• Short Circuits
• Reversed pairs
• Split pairs
• Transposed pairs

Open and Short Circuits

Open circuits occur when a wire is not attached to a pin correctly. A short circuit occurs when two wires are attached to each other.

Reversed Pair

Reversed pair wiring occurs when the wires of one pair are attached to the correct pins at one end but are reversed at the other end. The diagram below show the orange and orange/white wires of the first pair attached to pin1 and pin2 at the left end. At the other end the orange and orange/white wires are reversed and are attached to pin2 and pin1 instead of pin1 and pin2.

Reversed Pair Wiring Fault

Split Pair

A split pair fault occurs when a wire in one pair is connected to the wrong pin at both ends. This means the wire in another pair is also connected to the wrong pins at both ends. The diagram below shows the orange/white wire incorrectly attached to pin3 at both ends instead of pin2. The green wire is incorrectly attached to pin 2 at both ends instead of pin3.

Split Pair Wiring Fault

Transposed Pair

A transposed pair fault occurs when both wires in a pair are attached to the incorrect pins at one end. This means another wire pair are also attached to the incorrect pins at that end. The diagram below shows the green - green/white wire pair attached to pin1 and pin2 at the left end instead of the orange - orange/white pair.

Transposed Pair Wiring Fault

Some testers can only detect open or short circuit faults. More advanced high quality testers can detect faults such as reversed pair etc.

NEXT

NEXT stands for Near End Crosstalk and is a measure of the degree to which a signal from one wire pair is picked up by another wire pair. NEXT is measured in decibels (DB). When electric current flows through a wire, an electromagnetic field is generated. This field can cause a signal to flow in adjacent wires. This effect is called crosstalk.

When measuring NEXT you measures the difference in the signal strength between the transmitting wire pair and an adjacent wire pair. Since the ideal would be for the signal in the adjacent pair to be zero, the higher the decibel value for NEXT the better.

NEXT needs to measured by transmitting a signal down each wire pair in turn and measuring the crosstalk on the other wire pairs. This needs to be carried out on both ends of the link.

Crosstalk is minimized by the twisting of wire pairs. This is the reason that wire pairs in UTP and STP are twisted around each other. When a signal is sent down a wire pair, one signal is sent as a mirror image of the other signal which produces opposing electromagnetic fields. Twisting the wires cancels out the opposing electromagnetic fields and this protects adjacent wire pairs from crosstalk.

Crosstalk varies with frequency; the higher the frequency the more crosstalk and so NEXT measurements are taken across a range of frequencies. Higher category cables that carry higher frequency signals have wire pairs that are twisted even tighter to offset the higher crosstalk effect.

Excessive crosstalk can be caused by poorly terminated cables. At the connector point, the wire pairs must not be untwisted more than is necessary.

PSNEXT

PSNEXT stands for Power Sum Near End Crosstalk and is calculated by transmitting a signal down each wire pair in turn and summing all the NEXT values recorded on a particular wire pair.

Four PSNEXT values need to be calculated at each end of the link. Like NEXT, the higher the decibel value of PSNEXT the better, since this indicates less crosstalk.

On Gigabit 1000BASE-T LANs where all four wire pairs are used to transmit data, PSNEXT is an important measurement, although it is not required in the IEEE 802.3ae specification. However TIA/EIA-568-B certification now requires the PSNEXT test.

FEXT and ELFEXT

FEXT stands for Far End Crosstalk and is similar to NEXT in that it is a measure of the degree to which a signal from one wire pair is picked up by another wire pair. The difference is you send the signal from one end and measure the crosstalk at the opposite end.

Since a signal attenuates as it travels away from the source, FEXT is always weaker than NEXT. In fact, the longer the cable length the weaker FEXT becomes, so FEXT measured on longer cables will be different to FEXT measured on shorter cables, even if the NEXT measurements are similar.

This attenuation means that FEXT results are not really reliable, since they are dependent on cable length. A better measure of crosstalk disturbance at the far end is ELFEXT - Equal Level Far End Crosstalk.

ELFEXT is calculated by subtracting the attenuation measured on the wire pair the signal is sent on from the FEXT measured on an affected wire pair. Both these measurements are taken at the far end of the link. This removes the affect of cable length on the results.

PSELFEXT

PSELFEXT - Power Sum Equal Level Far End Crosstalk is calculated by summing each of the three ELFEXT effects from adjacent wire pairs on one wire pair.

Altogether, you would need to take twelve ELFEXT measurements on each end of a link and calculate four PSELFEXT values, one for each wire pair.

Optical Link Loss Budget

Data is transmitted through a fiber cable using light. The fiber jacket prevents the light signal from escaping and so crosstalk is not a problem on fiber cable. However, like copper cable, fiber is subject to some signal attenuation, although to a much lesser degree. Moreover, optical discontinuities in the cable can cause some of the signal to be reflected back toward the transmitter, causing loss of energy in the original signal. Optical discontinuities are primarily caused by improperly terminated ends. The main concern is that the light signal that reaches the receiver is of a sufficiently acceptable strength so that data is not lost.

The idea of the Optical Link Loss Budget is to ensure the amount of signal energy loss is acceptable and has not dropped below the requirements of the receiver. To test this, a light is sent down the link and the light energy is measured at both ends. The amount of signal energy loss can then be calculated.

Summary

On completing these notes you should have learned the following key points:-

• copper cable has to pass ten tests
• An optical link loss budget calculation is carried out on fiber

~~Activity~~