Crosstalk attenuation. Figure 10.9. b) Circuit for measuring crosstalk attenuation at the near end. Frequency dependence of next and fext

The need to continuously increase the volume and speed of information transfer makes it necessary to improve the quality indicators of cable routes. However, the possibilities for reducing the attenuation of horizontal cables have already been practically exhausted and remain only for data centers with their small line lengths.

The natural desire to ensure the normal performance of the information and telecommunication system (ITS) stimulates the introduction of communication channels with a constantly increasing bandwidth.

The trend towards more and more high-speed technology of high categories is clearly visible at all levels of the information cable system. Its horizontal subsystem, which in the overwhelming majority of cases is implemented on an electrically conductive symmetric element base, was no exception. Standard symmetric SCS cable tracts are characterized by high Shannon bandwidth combined with a relatively small bandwidth. The need to make the most of the potential capabilities of this guiding system forces the developer of network interfaces to use complex multi-position linear signals that are demanding on the quality indicators of the communication channel. The slightest failure to comply with the norms for these parameters leads to a sharp decrease in throughput and, accordingly, to a drop in the consumer value of the ITS as a whole, which is unacceptable.

Features of ensuring the quality of the signal in balanced cable SCS

The technology of local area networks (LAN) assumes that when switching to the next fastest-performing equipment, the transmission rate in the overwhelming majority of cases increases by an order of magnitude. This is a prerequisite for ensuring significant economic benefits from the introduction of more advanced technology.

One of the key factors determining the quality of information transmission in any telecommunication system is the signal-to-noise ratio at the input of its receiver with sufficient bandwidth. The prevailing type of interference in electrically conductive symmetrical paths of the SCS is transient noise. Interfering influences of other types, which also, within certain limits, affect the transmission quality with an accuracy sufficient for performing engineering calculations, are considered minor. This is largely due to the high efficiency of their suppression by the network interface itself, with appropriate processing of the mixture of signal and noise at the reception and correction at the transmitting end.

As a numerical measure of the signal-to-noise ratio in the SCS, the ACR parameter is used - an indicator of immunity from transient interference. To take into account the features of the transmission and processing circuit of the line signal used in modern high-speed interfaces, it is additionally indicated for the usual, total and inter-element influence, as well as for the near and far ends of the path.

It is easy to show that the security does not depend on the level of the output signal of the transmitter and is numerically equal to the difference between the values of the corresponding transient and operating attenuation, i.e. determined solely by the cable path itself. For example, used in the first editions of the standards, the inter-pair security at the near end is found as

ACR = NEXT - IL, dB,

where NEXT is the crosstalk attenuation at the near end, IL is the operating attenuation.

Other types of protection are obtained by simply replacing NEXT with the appropriate crosstalk.

The limiting bandwidth of a symmetric path is determined by the well-known Shannon ratio and is used with a high degree of completeness for modern multi-gigabit technology (about 60% in 10-gigabit systems). Therefore, when switching to the next generation of network equipment in terms of speed, the ACR value should be increased by about 10 dB over the entire operating frequency band. This is necessary to ensure a bit error rate of no more than 10-12, fixed by the IEEE specifications.

From the above ratio it follows that ACR can be increased in two, at first glance, equivalent ways: by decreasing IL and increasing NEXT.

Methods for reducing operating attenuation

To reduce the amount of operating attenuation, a cable designer can use several basic techniques:

increase the diameter of the twisted pair wire;
use materials with lower resistivity for the manufacture of conductors;
apply better insulation with reduced dielectric losses;
to improve the degree of matching of the wave impedances of those individual components, the series resistance of which forms the SCS cable path;
increase the nominal value of the characteristic impedance over 100 ohms.

Increasing the diameter of the conductive wire of the twisted pair over 0.64 mm is impractical because of the risk of incompatibility with the IDC contacts of the cable part of the connectors of the existing switching equipment.

Electrical copper, used for the manufacture of twisted-pair wires, is an almost ideal material, second only to silver in its characteristics, the transition to which is impossible for economic reasons. In addition, the baseband transmission method employed in Ethernet network interfaces makes it technically highly inefficient to resort to significantly more economical bimetallic wires when a thin layer of silver is applied only to the surface of the copper wire.

Also, the reserves for improving the quality of insulation have been largely exhausted. Modern polymer materials used to form insulating coatings of copper conductors are characterized by extremely low losses. In addition, the relative dielectric constant has been brought to a value of about 1.5. This is achieved, among other things, through the use of hollow materials obtained by foaming or structuring (Fig. 1). Its further significant reduction is problematic due to the difficulties associated with ensuring the mechanical stability of the insulating coating itself.

Improving the degree of matching of individual components makes it possible to bring the operating attenuation closer to the characteristic (theoretical minimum). The current editions of the standards state that for modern components the permissible deviation of the characteristic impedance from the nominal does not exceed ± 15% in the entire operating frequency range. Consequently, the degree of approaching the optimum is quite high and no significant progress in this area can be expected.

Increasing the wave resistance as a technique that does not require a transition to other source materials, allows you to achieve serious results. For example, the use of 120-ohm cables, which were allowed for use in SCS by standards back in 1995, for the widespread category 5e at a frequency of 100 MHz with a 100-meter line length gives a gain of about 5 dB. However, at the same time, due to the loss of the backward compatibility property, the operation of the cable system is greatly complicated. The reason is that a significant increase in the level of reflections at a point with different characteristic impedances does not guarantee the performance of gigabit network equipment and its faster modifications when directly connected to a fixed line. Turning to matching elements, regardless of their version, is associated with a number of obvious inconveniences of the operational plan and is considered extremely undesirable.

It follows directly from the foregoing that the possibilities of the known methods of attenuation reduction are rather limited and a breakthrough in this area should not be expected. It is no coincidence that the specifications of cable paths of promising category 8, which are currently being developed, proceed from linear-logarithmic interpolation of the characteristics of the attenuation coefficients of the element base of categories 6a and 7a in the high-frequency part of the spectrum of the linear signal of 40-gigabit network interfaces (Fig. 2).

Increased crosstalk

In wide engineering practice, there are many ways to improve the characteristics of individual components and complex objects of the electrical conductive subsystem of SCS in terms of transient influences. To improve the in-cable crosstalk, the following are involved:

reduction of the pitch of twisted pairs up to values less than 10 mm;
introduction of twisted-pair separator into the core design;
application of individual shielding for each pair.

The crosstalk attenuation for products of category 6a and higher is increased to the required value by the following measures:

artificial increase in the effective outer diameter of unshielded structures in order to reduce intercable influences;
use of braided and film screens (in the latter case, their ungrounded design is possible).

From the above list it follows that the changes that form the basis for the correction of the cable design are purely mechanical in nature. Due to this, they do not require a radical restructuring of cable production and the introduction of new materials.

How to increase ACR?

Of course, there are no contraindications to improving the quality indicators of the SCS wiring lines due to the simultaneous decrease in the operating attenuation and the increase in the transient attenuation. First of all, this refers to a balanced cable as the most "noisy" component of the path.

From the data presented above, it follows that achieving the required ACR value by increasing the NEXT is much more efficient. Let us illustrate this position with a numerical example. When switching from category 5e equipment to category 6, the twisting step is reduced by several tens of percent. In the overwhelming majority of cases, a separator is additionally introduced into the design of the cable core. The complex of these measures, rather simple by modern standards, makes it possible to achieve an increase in NEXT by the above 10 dB. In addition, the NEXT build-up appears to be the same over the entire operating frequency range. The insertion loss IL is reduced by increasing the diameter of the conductor of the pair from 0.51 to 0.53 mm. The absolute amount of reduction according to the standards is approximately 2 dB at 100 MHz, i.e. the gain in this parameter from the transition to a higher quality element base turns out to be quite small. Moreover, as the frequency decreases, the gain decreases, which further reduces the efficiency of increasing the bandwidth of the cable path in this way.

Further analysis can be based on the fact that with the current state of the art, the practical need to increase the guaranteed minimum value of the ACR value currently exists only in the data center. A clear manifestation of this trend is those significant tightening of requirements for the main parameters of electrical conductive paths, which are fixed in the draft specifications of the equipment of the promising category 8. The focal area of application of this equipment is considered to be the hardware rooms of the data center.

SCS for a data center has a number of features, the totality of which led to the separation of this type of information cable systems into an independent class with its own regulatory framework. Along with noticeably higher frequencies of transmitted signals, such cable systems are distinguished by noticeably smaller average lengths of the organized paths.

Under these conditions, the technical and economic efficiency of the SCS can be significantly increased due to the rejection of the guaranteed provision of the classical 100-meter length of the channel. Turning to this approach is also advisable because it has a positive effect on the energy efficiency of the facility as a whole.

From a technical point of view, reducing the maximum allowable length of the tract to 30 m is beneficial in that it is accompanied by a sharp drop in the IL value. For example, for cable type UC1500 from Draka on the top cutoff frequency 1500 MHz gain reaches 45 dB. In this case (even taking into account the decrease in the gain as the frequency decreases), the contribution of IL to the ACR build-up and, through it, the Shannon bandwidth becomes comparable to that achieved by the NEXT improvement.

In addition, the reduction in total losses is also valuable in that it leads to a natural expansion of the bandwidth (the upper cutoff frequency of the path is determined by the ACR criterion) and significantly simplifies circuitry solutions when designing a network interface transceiver. The most significant possibility is to keep the bit width of the linear signal unchanged and to use a less complex receiver. 

To increase the throughput of a symmetric path up to 10 Gbit / s and higher, it is not enough to use the internal reserves of the existing element base and it is necessary to improve its basic parameters.

Improvement of the quality indicators of a symmetrical conductive path is achieved mainly by improving the characteristics of a horizontal cable in terms of influence parameters.

The reserves for minimizing the attenuation coefficient of horizontal cables within the limits set forth in the existing regulatory documents and the achieved level of technology are almost completely exhausted.

Reducing the total attenuation of a symmetric path is relevant exclusively for a data center and is ensured by reducing its maximum permissible length to the limit determined by the energy efficiency of the control room as a whole.

Maximum attenuation between two telephones on the city telephone network should be no more than 28 dBr (decibel difference). In this case, all attenuation values are shown from the level of the previous point. Wherein attenuation of subscriber lines(AL) should not exceed 4.5 dB for a cable with a core diameter of 0.32 and 3.5 dB for cores with a large diameter.

Attenuation of the station four-port network should not exceed 1 dB at RATS (regional automatic telephone exchanges) and 0.5 at nodal stations (outgoing UIS or incoming message - UMS).

With four-wire switching the attenuation of the station four-port network of the nodal stations is taken to be zero. When transition from two-wire connection to a four-wire path, the attenuation is 1 dB. When using electronic automatic telephone exchanges attenuation in areas with a PCM transmission system should be 7 dB. The distribution of attenuation in dB on the GTS is shown in Fig. 2.6.

Fig. 2.6.

Crosstalk attenuation

Crosstalk attenuation- a value that characterizes the relative amount of energy transferred from one circuit to another due to electromagnetic coupling; expressed in decibels. Just like normal attenuation, it is measured as the ratio of output power to input power. But in this case, the input is the power of the useful signal of one circuit, the output is the power of the same signal in the adjacent circuit. This effect necessarily takes place between adjacent circuits (cable cores, overhead line wires). It can be generated by signal transitions from the receiver to the transmitter, as well as by converting a four-wire line to a two-wire line and reverse conversion.

The crosstalk attenuation is different:

measured at the near end (NEXT - Near End Cornstalk)... This refers to the transfer of power from one pair to another, which is measured at the end closest to the transmitter of the affected pair;
measured at the far end (FEXT - Far End Cornstalk)... This refers to the transfer of power from one pair to another, which is measured at the end farthest from the transmitter of the affected pair. Measurements are carried out over the entire range of operating frequencies, i.e. for a speech signal - in the frequency range 300-3400 Hz.

Crosstalk mitigation measures.

Twisted pair cable

To reduce the effect of crosstalk attenuation, cables with twisted (twisted) pairs... These are multicore cables in which the conductors are twisted in pairs or fours. The principle of dealing with crosstalk interference is that when twisting the wires, affecting individual sections of the cable, induce electromagnetic energy equal in amplitude and opposite in direction, as shown in Figure 2.7. With a perfectly balanced twist (equal twist pitch, perfect symmetry of the wires), the crosstalk is zero.

Fig. 2.7. A method of eliminating interference by "crossing" wires for example, the use of electromechanical and electronic systems in the same room. IN modern systems using subscriber data transmission devices, the impulse noise ratio is of great importance

Impulse noise ratio serves for digital estimation of the line state, it indicates the number of errors per a certain number of transmitted bits. The error rate is considered normal - this means that one interference appears on the bits in the channel, which can lead to an error. The minimum acceptable value of the error rate (usually allowed when using a radio path) is. The value is considered good. It should be borne in mind that these indicators are conditional. They are measured over a certain time interval, for example, an hour. But in reality, during each interval, they are distributed unevenly and can come in concentrated (pack). Therefore, sometimes the coefficient of "patchiness" (concentration of errors) is introduced, which shows the ratio of the number of errors received in a given time interval to the expected average over all intervals. Various algorithms are used to overcome errors, which will be discussed below. Interference degrades the quality of voice reception, and during data transmission can lead to incorrect reception or delays that slow down the actual data exchange rate (modem speed). The greatest problems arise with the deterioration of this coefficient and with the control of the channel quality from the side of transmitting or receiving devices. If these devices are configured to disconnect the channel when the error is exceeded, then in case of random disturbances in the network, a complete shutdown of the station often occurs. Therefore, with automatic control of this parameter, it is necessary to leave the possibility of adjusting the threshold.

Determination of the magnitude of the currents of influence on the near and far ends of the cable line

Cable lines are assembled from separate cable sections (construction lengths) supplied by factories with twisted (crossed) conductors of circuits, and therefore the phases of the influence currents arriving at the near and far ends of the cable line are unknown. When determining the total current of influences, the quadratic law of addition of currents of individual building lengths is used. The situation under consideration differs from the case of influences between the circuits of overhead lines, where the phases of the currents coming from separate sections of the mutually influencing circuits to the near (far) end are known, since chain crossing schemes are installed during the construction of the overhead line.

Let us assume that there is a cable line of n cable segments of length S with circuits having the same parameters. To determine the crosstalk attenuation at the near end, we assume that the electromagnetic couplings between the circuits are constant along the entire length and the influence current of the first building length, then the influence current from the second building length will be, etc., and from the last building length.

Total influence current at the near end

In this case, the ratio of currents

Equating , for the transition damping we get

where is the crosstalk attenuation at the near end of the face-to-face length, usually determined by measurements.

All currents of influence on the far end pass through separate building lengths and their paths from the beginning of the influencing circuit to the end of the affected circuit are the same. Therefore, when summing them up according to the quadratic law, all terms under the square root are the same, and the total current

Passing to the ratio of currents and taking the logarithm, we obtain

where is the crosstalk attenuation at the far end of the headroom, determined by measurements,.

Far-end security

The building lengths of the cables are interconnected during the installation work; they form a cable line.

Balancing cables

Cable chains in building lengths of the same type of cable always have different electrical characteristics (within the limits allowed by the technical conditions), and their protection from mutual influences and influences of external sources depends on how they are connected. Therefore, when performing installation work with symmetrical cables, balancing is carried out - a set of measures aimed at reducing the effects.

Balancing methods... Mutual influences arise as a result of the presence of electromagnetic links between the circuits. At the same time, electrical connections prevail in low-frequency (up to 4 kHz) cables, and electromagnetic complex connections prevail in high-frequency cables. Proceeding from this, in LF cables, it is enough to carry out balancing of capacitive couplings; in HF cables, it is necessary to balance all components (active and reactive) of electrical and magnetic connections. For balancing LF cables, the crossing method and the capacitor method are used. Balancing of HF cables is carried out by the methods of crossing cores and concentrated balancing by counter-coupling circuits.

The essence of balancing by crossing the cores is to compensate for the electromagnetic connections between the circuits in one section of the cable line with the connections of another section. The compensation is explained by the fact that when crossing, the bonds change their sign.

When balancing the capacitor method, the latter are installed in an intermediate sleeve connecting two sections of the cable line, and are connected between the conductors of the circuits. Their capacity is chosen such that the sum of the partial capacities C 13 + C 24 (Fig. 1) is close to the sum of C 14 + C 23. In the case of equality of the sums, the equilibrium of the electric bridge is achieved, and the capacitive coupling is equal to zero.

Concentrated balancing by the counter-coupling circuits is that the interference currents caused by electromagnetic couplings between the circuits are compensated by the currents of the influence of the opposite phase created by the circuits connected between the cores of the interacting circuits.

Figure 2 shows the circuit for switching on the counter-coupling circuit F p, and the natural distributed coupling is shown in the form of an equivalent coupling F. Since the currents of influence I and I p on the far end of different sections of the approach of the circuits have the same phase, then to compensate these currents, it is enough with the help of the circuit create the same current, but the opposite phase. In practical balancing, the difficulty lies in the implementation of the required frequency dependence of the counter-coupling circuit, which reproduces the frequency dependence of the natural electromagnetic coupling, which is of a complex nature, and in the need to take into account the permutation effect.

Balancing is greatly simplified when using a set of instruments for visual measurement of complex couplings in active and reactive components, as well as transient attenuation in absolute value and phase, instead of measuring instruments frequency characteristics near-end crosstalk and far-end protection.

The influence currents from different sections come to the near end of the circuit with different phases, and it is difficult to compensate for them with counter-coupling currents, since the counter-coupling loops must be connected in places where electromagnetic communication is affected. Considering that in reality electromagnetic connections are distributed in nature, then to obtain compensation, it is necessary to connect a large number of counter-coupling loops between the circuits, which is practically unacceptable. Therefore, concentrated balancing by the counter-loops is used only to reduce the effect on the far end. The impact on the proximal end is reduced by crossing.

The balancing technique for high-frequency and low-frequency circuits is different. High-frequency circuits have a large attenuation at high frequencies, and currents of influence on the near end of sections located at a distance corresponding to an attenuation of 10-11 dB (at high frequencies transmitted spectrum) are negligible. This allows balancing across the entire gain section. Low frequency circuits have significantly less attenuation, and by reducing the far-end impact, the near-end impact can be increased and vice versa. Low-frequency cables are balanced in small sections, called balancing steps: sections of a cable line consisting of several construction lengths with a total length of up to 4 km. Usually, the length of the balancing step of low-frequency cables is taken equal to 2 km.

In long-distance railway cables, there are high and low frequency fours. When balancing such cables, both methods must be used.

3. Balancing low-frequency circuits... In star-twisted cables, the greatest influences are between the circuits within the fours. The influence between chains of adjacent quadruples is less due to the different steps of their twisting. However, with a long cable length, this influence can exceed the permissible value. The influence is reduced by mixing the fours, which consists in the fact that along the cable line the fours change places, then moving away from each other, then approaching. In railway cables, balancing is mainly used inside quadruples. Before starting the balancing of the circuit, all branches from the main cable to the automation and communication devices must be connected to it.

Low-frequency circuits of balanced cables, in contrast to high-frequency ones, have higher values of characteristic impedance. Therefore, when signals of the same power are transmitted through these circuits, the voltage in the low-frequency circuits will be higher, and the current is less than in the high-frequency circuits, and, therefore, the effects between the low-frequency circuits are more due to electrical connections than magnetic ones. Low-frequency circuits of main railway cables must be balanced in the same couplings as high-frequency ones. If the locations of the amplifying points of the LF and HF circuits coincide, the low-frequency circuits should be balanced simultaneously with the high-frequency circuits, and if they do not coincide, the high-frequency circuits are first symmetrical, and then the low-frequency circuits.

To balance the quadruples, first measure the capacitive couplings in the connected construction lengths of the cable: k 1 = (C 13 + C 24) - (C 14 + C 23) between the main circuits in the quad; k 2 = (C 13 + C 14) - (C 23 + C 24) between the first main and artificial; k 3 = (C 13 + C 23) - (C 14 + C 24) between the second main and artificial. Also measure the capacitive asymmetry e 1 = (C 10 -C 20) of the first pair of the four; e 2 = (C 30 -C 40) of the second pair of the four; e 3 = (C 10 + C 20) - (C 30 + C 40) of an artificial circuit, where C 13, C 23, C 14, C 24 are the capacitances between the strands of the circuits; C 10, C 20, C 30, C 40 - capacitances between the conductors and the ground (shell) (see Fig. 1).

Then balancing is performed in three stages: inside balancing steps; when connecting steps and on the mounted amplifying section.

Balancing within steps (the first stage) can be performed at one, three and seven points located at the same distance from each other and from the ends of the balancing step (Fig. 3). Couplings, in which crossover balancing is performed, are called baluns. Couplings in which balancing is performed by crossing and capacitors are called capacitor couplings. Couplings in which balancing is not performed and the conductors are connected directly are called straight couplings and are designated with a circle (see Fig. 3).

With a single-point scheme, first, direct couplings are mounted, and then a capacitor (K). In the case of a three-point circuit, straight couplings are first mounted, then balancing and only then capacitor ones. When balancing according to a seven-point scheme, first mount the balancing couplings A, then B and, lastly, the capacitor coupling K.

The circuits for crossing the veins of the chains when connecting fours in balancing couplings are selected according to the measurement data of capacitive couplings and asymmetry. For example, if in one section of the cable line there is a capacitive coupling between the circuits of one of the fours pF, and in the other section the capacitive coupling between the circuits is also inside one quadruple of pF, then when the cores of both quadruples are connected without crossing, the resulting connection is pF. If the cores of one of the circuits are crossed in the coupling, then the resulting bond is k 1 = 350-300 = 50 pF. In the case of crossing both chains, the value of the resulting bond will not change (650 pF).

When there is an artificial chain, 8 crossings are possible. These combinations of crosses and the corresponding signs of capacitive couplings and asymmetries are shown in Table 1.

The dashes next to the letters indicate cable sections. For convenience, introduced legend called operators. The cross corresponds to crossing, and the points to direct connection (color to color).

When performing balancing by crossing, try all possible schemes and choose the one at which the bonds and asymmetry have the lowest values. When it is impossible to simultaneously reduce ties and asymmetry, the operator is selected based on the problem of reducing ties.

Table 1

If crossing failed to reduce the bonds and asymmetry to acceptable values (k 1, k 2, k 3 ≤ 20 pF; e 1, e 2 ≤ 100 pF), then balancing with capacitors is used.

The capacitances of these capacitors are chosen as follows. Suppose it was found by measurements that k 1 = - 30 pF. This means that in the equation for k 1, the sum of capacities (C 13 + C 24) is less than (C 14 + C 23) by 30 pF. Therefore, in order to obtain the value k 1 = 0 and not change k 2 and k 3, it is necessary to include additional 15 pF capacitors between cores 1-3 and 2-4 of the quad. Similarly, you can reduce the connections k 2 and k 3. To reduce asymmetry, capacitors are selected in the same way, but they are connected between the corresponding cores and the shell (ground).

When the steps are connected to each other (the second stage), balancing is performed by the crossing method according to the results of measurements of the crosstalk attenuation between the circuits at a frequency of 800 Hz. Select the operators that give the greatest crosstalk. The steps are built up sequentially, starting from the ends of the amplifying section to its middle, according to the measurements of the crosstalk at the near and far ends, achieving their maximum value. At the same time, the working capacitances and resistances of the cores of the main circuits are aligned in the balancing step so that the asymmetry does not exceed 0.1 Ohm.

In areas where large external influences are possible, at the second stage of balancing, additional measures are taken to reduce the coefficient of sensitivity of the circuits to interference.

For this, when connecting balancing steps together in the direction from the end of the amplifying section to its middle, according to the results of measuring the crosstalk attenuation at the near end and the voltages U 1 and U 2 in the connected quadruples of the cable (Fig. 4).

The measuring generator G is switched on at the end of the incremental balancing step S (point C). At the head-end station (point A), a series of measurements is carried out at the terminals of the load resistances of cable circuits in a quadruple.Each group of two measurements refers to specific operator crossing of four veins in a mounted sleeve. The lowest measured voltage will correspond to the minimum gain of the circuit.

An acceptable operator (connection diagram of the conductors at point B) is chosen in a compromise based on the results of comparing the crosstalk attenuation values between the circuits in the cable quad and the measured voltages U 1 and U 2. In this case, the transient attenuation should not be less than permissible, and the measured voltages should be the smallest.

In the third stage, balancing on the mounted amplifying section is performed in a sleeve located approximately in the middle of the amplifying section. In this coupling, the wires are connected in a quadruple according to the results of measuring the security at the far end and the voltages U 1 and U 2, choosing the most compromise profitable operator... In quadruples that do not meet the crosstalk attenuation and protection standards, compensating circuits are included.

4. Balancing high-frequency circuits.

To reduce labor intensity and increase the efficiency of balancing at the stage of preparatory work, the cable construction lengths are grouped according to the average values of the working capacitance of the circuits and according to the magnitude of the transient attenuation at the near end. In this case, from the passport data for the building lengths, the minimum values of the crosstalk attenuation at the near end between all circuits are selected and a list of laying these cables on the site is drawn up. At the ends of the amplifying section, cables with the highest crosstalk attenuation are laid, which makes it possible to exclude or significantly facilitate the process of balancing to the near end of the circuit. For high-frequency circuits, balancing is performed within the amplifying sections of transmission systems with frequency division multiplexing (digital systems have greater noise immunity and do not require balancing of RF circuits). Balancing at the far end of the amplifying section is performed in two stages: at the first - systematic crossing of the first chain of the four when connecting the construction lengths of the cable (the operator of the connection in the sleeve of the cable cores x ..); on the second, crossing the chains at one, two or three points (couplings) (Fig. 5) with the selection of the best combination of crossing operators by experience based on the results of measuring the security of the chains at the far end of the reinforcing section. The efficiency of a two-stage crossing of RF circuits depends on the values of the so-called symmetry parameter of intra-quadruple influence combinations for the cable length. This parameter is determined by the minimum value of A l, which can be achieved by compensating for direct influences. The efficiency of two-stage crossover also depends on the frequency range and the length of the gain section.

The best combination of crossing operators for three-point or two-point balancing schemes is understood to be one that achieves the required security norm A s l over the entire frequency range. If this cannot be achieved, then the selected crossing operators must first of all eliminate the effect of permutation in order to be able to use symmetry using opposition contours. In the latter case, the balancing of the RF circuits is obtained in three stages.

In addition to the considered methods of reducing mutual influences between RF circuits, in some cases, other (additional) measures may be required, for example, to reduce the influences from the output of an intermediate amplifier (regenerator) to its input in combined railway communication cables and a compensatory method for weakening mutual influences on sections of the medzhu neighboring served amplifying points (OUP-OUP). This method serves to ensure interference immunity from mutual influences when organizing communication via a cable designed according to technical conditions for operation in a narrower frequency range than required by the equipment used.

The influence from the output of the intermediate amplifier on its input must be taken into account on cable lines in the presence of low-frequency circuits passing without break through a high-frequency amplifying point (UE). In this case, the indicated influences take place through the third low-frequency circuits (Fig. 6). Elimination of these influences can be ensured by the transition of RF circuits from one cable to another in each amplifier point (Fig. 7). The influences from the output to the input of the RF amplifiers through the third two-wire circuits can be reduced by the inclusion of low-pass filters in the latter.


Fig. 6	Fig. 7

To reduce these influences on overhead lines, the inputs to the amplifying points are arranged in different cables. To reduce the influence through the ground path, locking coils (ZK) are included in all circuits at the input and output to the amplifying points (Fig. 8). Each half-winding of the ZK coil is included in one of the wires of the two-wire circuit. As a result, the magnetic fields of the earth path currents (having the same direction) are added, which increases the inductive resistance of the wire-to-earth circuit. The magnetic fields of currents having different directions in the wires of a two-wire circuit are mutually compensated, and the attenuation introduced by the blocking coil for the transmitted signals is small. When inserted at the endpoints, the locking coils are only included in sealed circuits.

Compensation method for weakening mutual influences in the sections of the OUP-OUP. Line paths of railway trunk cable lines are in more difficult conditions in comparison with similar lines of the Ministry of Communications. This is due to the presence of third unsealed chains, a large number of paper-insulated and aluminum-sheathed cables, which are difficult to symmetrically in wide range frequencies, a large number of taps from the trunk cable. Therefore, in relation to cable lines of railway transport, this method of weakening mutual influences is most applicable.

The compensation method has great opportunities for weakening mutual influences in comparison with balancing methods. This is explained, firstly, by the fact that it takes into account the presence of the permutation effect, which arises due to the difference in the propagation constants of mutually influencing chains (the permutation effect is manifested in the fact that the complex connections for combinations of the influence of the first chain on the second and vice versa are different); secondly, the use of a wider element base (except for resistors and capacitors, which are used, as in the method of balancing by anti-coupling circuits within the amplifying section, adjustable delay lines and inductors are used, on the basis of which band-pass filters with the required characteristics are created). The disadvantage of the method under consideration is that it can only be applied on a highway with fully tuned line paths, and when using it, it is impossible to control the quality of construction by the most important parameter - transient attenuation and security.

Mutual influences in the sections of the OUP-OUP are suppressed by the inclusion of a counter-coupling circuit in the receiving OUP (Fig. 9). The counter-coupling circuit is selected so that the compensation current I to is the same in modulus and opposite in phase to the resulting interference current at the input of this OUP where is the interference current induced within v-th amplifying section. To ensure independence of suppression of mutual interference between different combinations of influences (taking into account the effect of permutation), a unidirectional device is used, which is installed at the input of the counter-coupling circuit.

The selection of the elements of the counter-coupling contours is possible in two main ways - by calculation and by instrumental-iterative. The latter is used on railway cable lines, since it is more visual and does not require the use of special equipment. The hardware-iterative method of synthesis of the counter-coupling circuits consists of three stages: the first is the measurement of the hodograph of complex bonds in the OUP-OUP section, the second is the selection of the elements of the counter-coupling loops based on the data obtained at the first stage; the third is the measurement of the difference hodograph after connecting the contour of the opposition between the mutually influencing circuits and the refinement of the elements of the latter. The selection of the elements of the contours of the countermeasures consists in the choice of the required typical scheme of the countermeasure or a combination of their inclusion. The average value of the efficiency of attenuation of mutual influences in the sections of the OUP-OUP is 10-12 dB.

Control questions

1.Explain the physical meaning of balancing using opposition contours and crossing method.

2. What are the features of balancing low-frequency (high-frequency) communication cables?

3. What is the peculiarity of balancing low-frequency circuits when exposed to external influences?

4. Explain the advantages and disadvantages of the compensatory method of weakening mutual influences in the sections of the PMO-PMO.

5.Explain the purpose and principle of operation of the locking coils.

A twisted pair is a copper-based cable that joins one or more pairs of conductors in a sheath. Each pair consists of two insulated copper wires twisted around each other. Cables of this type often differ greatly in the quality and transmission capabilities of information. The conformity of the characteristics of cables to a specific class or category is determined by generally recognized standards (ISO 11801 and TIA-568). The characteristics themselves directly depend on the structure of the cable and the materials used in it, which determine the physical processes that take place in the cable during signal transmission.

Balance pair

The balance of a pair is actually a defining characteristic of cable quality, as it affects most of its other properties. The fact is that an electromagnetic (Electro Magnetic - EM) field induces an electric current in conductors and is formed around a conductor when an electric current flows through it. Interaction between EM fields and live conductors can have a negative impact on the quality of signal transmission. In both conductors of a balanced pair, electromagnetic interference (em1 and em2) induces signals of the same amplitude, (S1 and S2) in antiphase. Due to this, the total radiation of the "ideal pair" tends to zero.

If there is more than one pair in the cable, then in order to exclude mutual interference of the pairs, which could violate the electromagnetic balance, the pairs are twisted with different steps.

Impedance

(Characteristic impedance)
Like any conductor, "Twisted pair" has a resistance to alternating electric current. However, this resistance can be different for different frequencies. Twisted pair has an impedance of typically 100 or 120 ohms. In particular, for Category 5 cable, the impedance is measured in the frequency range up to 100 MHz and should be 100 Ω ± 15%.
For an ideal pair, the impedance should be the same along the entire length of the cable, since the effect of signal reflection occurs at the points of inhomogeneity, which in turn can degrade the quality of information transmission. Most often, the homogeneity of the impedance is violated when the twist pitch, cable bend during laying, or other mechanical defect changes within one pair.

Signal propagation speed / delay

NVP (Nominal Velocity of Propagation) - signal propagation speed. Expressed as the ratio of the speed of propagation of the signal to the speed of light. However, often the derivative of NVP and cable length is the "delay" characteristic, which is expressed in nanoseconds per 100 meters of the pair. If more than one pair is present in the cable, then the concept of "delay skew" or delay difference is introduced. The point is that pairs cannot be perfectly the same, which gives rise to different propagation delays in different pairs. Ideal systems assume that such differences will be minimal.

Attenuation

In addition to impedance and signal propagation speed, other important characteristics of the "Twisted-pair" cable are also distinguished. One of these is attenuation, which characterizes the amount of signal power loss during transmission. The characteristic is calculated as the ratio of the power received at the end of the line to the power of the signal applied to the line. Since the amount of attenuation changes with increasing frequency, it must be measured over the entire frequency range used. The value itself is expressed in decibels per unit length.

The graph below shows the transmission loss of signal power as a function of both cable length and frequency used.

(Near End Crosstalk)
Others important parameter is NEXT (Near End Crosstalk), or crosstalk between pairs in a multi-pair cable, measured at the near end - that is, at the signal transmitter side, that characterizes the crosstalk between pairs. NEXT is numerically equal to the ratio of the applied signal on one pair to the received signal on the other pair and is expressed in decibels. NEXT matters the more the better balanced the pair. Measurements must be carried out on both sides, since this characteristic depends on the relative position of the measuring instruments and the locations of possible defects in the cable. Like the linear attenuation, NEXT must be measured for the full range of frequencies.

In a multi-pair cable, measurements are made for all pair combinations. However, nowadays, more and more in-depth tests are also being used, based on the identification of group interference at the near end between all pairs (Power Sum Crosstalk) present in the cable.

Power sum crosstalk

Another name for this characteristic is Power Sum NEXT or PS-NEXT. Like NEXT, Power Sum CrossTalk expresses the crosstalk between pairs in a multi-pair cable, measured at the near end - that is, at the transmitter side of the signal. However, simultaneous pickups from all pairs present in the cable are taken into account. Like NEXT, PS-NEXT is measured at both ends of the line for the full frequency range.

In addition to evaluating the mutual interference of pairs at the near end of the cable, the crosstalk is measured from the side of the signal receiver. This test is called FEXT (Far End Crosstalk).

FEXT

(Far End Crosstalk)
Far End Crosstalk, or far end crosstalk, characterizes the effect of a signal in one pair on another pair. Unlike NEXT, FEXT is measured by applying a test signal to a pair in a cable from one pair and measuring the induced signal in the other pair from the receiver side. The characteristic is numerically equal to the ratio of the test signal to that induced by the generated electric field. FEXT, like the entire family of crosstalk characteristics, is measured over the entire frequency range used and is expressed in decibels.

ACR

(Attenuation Crosstalk Ratio)
One of the most important characteristics reflecting the quality of a cable is the difference between linear and crosstalk attenuation, expressed in decibels. The lower the linear attenuation, the greater the amplitude of the useful signal at the end of the line. On the other hand, the greater the crosstalk, the less is the mutual pickup of the pairs. Thus, the difference between these two values reflects the real possibility of separating the useful signal by the receiving device against the background of interference. For confident reception signal, it is necessary that the Attenuation Crosstalk Ratio is not less than the specified value defined by the standards for the corresponding cable category. With equal linear and cross-section attenuation, it becomes theoretically impossible to isolate a useful signal. Since the response is not measured but is the result of calculations based on attenuation measurements, which in turn depend on the frequency used, the ACR must be calculated for the entire frequency range used.

ELFEXT

(Equal Far End Crosstalk)
ELFEXT is the reduced crosstalk. This characteristic is calculated from the far-end crosstalk (FEXT) and the coupled attenuation (Attenuation) measurements. In fact, ELFEXT is the ACR at the far end of the cable link, i.e. the difference between the FEXT parameters of the first pair and the Attenuation of the second. ELFEXT, like the entire family of crosstalk characteristics, is calculated over the entire frequency range used and is expressed in decibels.

PS-ELFEXT

(Power Sum Equal Far End Crosstalk)
PS-ELFEXT is the total reduced crosstalk. This characteristic is calculated for each individual pair by simply summing the values of its parameters elfext relative to all other pairs.

Return loss

(RL)
When a signal is transmitted, the so-called effect of signal reflection in the opposite direction occurs. The amount of return loss, or "reverse fade", is proportional to the attenuation of the return signal. This characteristic is especially important when building networks with support for the Gigabit Ethernet protocol, using transmission of signals over twisted pair in both directions (full duplex transmission). The reflected signal, which is large enough in amplitude, can distort the transmission of information in the opposite direction. Return Loss is expressed as the ratio of the power of the direct signal to the power of the reflected signal.

Bob kenny
CIO at Prestolite Wire Cop.

It seems that everyone knows about the cable with unshielded twisted pairs. However, one more detailed acquaintance with it will be useful, especially in connection with the appearance of its new varieties.

Unshielded twisted pair wiring has had a huge impact on network infrastructure. Thanks to it, users were able to use one type of cabling system for any local network application. However, in Lately solutions based on UTP have become much more diverse. At the moment, manufacturers offer numerous varieties of UTP wiring from basic Category 3 to non-standard Category 6 yet. As a result, it becomes more and more difficult for end users to understand how different types of wiring differ.

Many articles have been written on this topic. In some, the boom of new UTP wiring classes is considered nothing more than a marketing gimmick of manufacturers, in others, the proposed improvements are classified as a belated modernization of outdated technology. So who's right?

DEFINITION OF UTP WIRING

UTP wiring has undergone significant changes over the past decade. The growth of network needs has led to the emergence of a potential demand for UTP wiring more than High Quality... But before we move on to discussing the merits of UTP wiring, we must first understand the terms that define it.

Appointment of any network cable consists in transferring data from one device to another. These devices can be terminals, printers, servers, etc. They can connect to various types of cable media, including optical, coaxial, biaxial, and various combinations of shielded and unshielded pairs. Choosing the best fit for of this application The type of wiring depends on many factors, including the distance of the endpoints, age, noise level, security requirements, financial constraints, expandability, and transmission speed. Many end users view unshielded twisted pair cables as the standard transmission medium that solves many of these problems.

The most popular UTP is used as horizontal wiring, namely for connecting desktop systems to telecommunication closets (Telecommunication Closet, TC). As the name suggests, UTP consists of multiple unshielded twisted pairs surrounded by a common sheath. Despite the availability of two and 25-pair cables, four-pair wiring is the most popular. Although most LAN environments, such as 10 / 100BaseTX, use only two of the four pairs, the new protocols under consideration, in particular Gigabit Ethernet, will use all four pairs.

RUNNING ATTENUATION

Figure 1. Specific attenuation.

Signal fading is one of the biggest challenges for any cabling infrastructure. Unfortunately, when information is transmitted from device to device, the signal quality deteriorates. For example, when traveling a distance of 100 m over a UTP cable, a 100BaseT signal usually loses a significant part of its original power (see Figure 1). If these losses are too large, then the receiving device will not be able to recognize the transmitted data. To prevent this from happening, the standards committees impose limits on the amount of loss allowed.

Loss is characterized by the term "linear attenuation" or simply "attenuation". In the case of UTP, attenuation refers to the amount of transmission loss through a conductive medium and is expressed in decibels (dB). There are advantages to using decibels as a unit of measurement. For example, it is easy to remember that when the signal is attenuated by 3 dB, it loses 50% of its power. Table 1 shows how decibels compare to wasted signal power.

The amount of loss depends on the cable design, including the conductor size, composition, insulation and / or sheath material, operating frequency range, transmission speed and cable length. The influence of the first factor, the size of the conductor, is most obvious. Generally, the larger the conductor, the lower the loss. For this reason, many older UTP cables use 23 AWG instead of 24 AWG.

The conductor material (composition) is also important. For example, copper has lower losses than steel. Some materials, in particular silver, have more best performance than copper, but many of them are too expensive for mass use. Insulation material can also have an impact on signal attenuation. High quality UTP cables typically use low loss materials such as fluorinated ethylene propylene or polyethylene to insulate the conductor. These materials generally have lower losses than other compounds such as PVC. The sheath material is also reflected in the amount of attenuation. This is why many manufacturers separate the jacket from the insulated pairs using a non-rigid pipe design. In addition, the attenuation in UTP copper wiring is known to increase with increasing frequency. For example, at 100 MHz, the attenuation is greater than at 1 MHz (assuming the cables are the same length). Finally, signal loss depends on the length of the cable. All other things being equal, the longer the cable, the greater the loss. For this reason, attenuation is expressed in decibels per unit length.

Attenuation summary:

when passing through the cable, the signal loses its strength;
attenuation determines the amount of losses;
attenuation is expressed in decibels (dB);
cable attenuation depends on factors such as conductor size and composition, operating frequency (frequency range), speed and distance.

TRANSITIONAL ATTENUATION

Figure 2. Crosstalk.

A twisted pair is called active if a signal is transmitted over it. An active pair naturally creates an electromagnetic field. This field can affect other nearby active pairs (see Figure 2).

One of the more difficult things to understand about crosstalk has to do with units, namely decibels. In the case of linear attenuation, the higher the value in decibels, the higher the signal loss. In the case of crosstalk, the opposite is true - the higher the decibel value, the less interference. Table 2 will provide a better understanding of the situation.

Obviously, the appearance of noise in adjacent pairs is undesirable. As can be seen from the diagram, the greater the crosstalk attenuation in decibels, the lower the induced voltage (i.e., noise) in adjacent pairs.

The attenuation per unit length characterizes the signal loss. Therefore, the higher the decibel value, the higher the signal loss. However, crosstalk is the loss of noise. In this case, the larger the decibel value, the greater the noise loss. And of course, the more actively the noise is attenuated, the better.

TYPES OF TRANSITIONAL ATTENUATION

Crosstalk at the near end. Systems such as 10BaseT Ethernet use two pairs for communication, one for transmitting and one for receiving (see Figure 3). The signal has the highest power immediately after the moment of data transmission. Conversely, the signal has the lowest power just before the moment of receiving the data.

Most often, the term "crosstalk" is used in conjunction with the phrase "near end". The reason for this is that at the near end, where the signal has the greatest strength, it generates powerful electromagnetic radiation (electromagnetic interference). Next to the transmitter, along the adjacent pair, an attenuated signal goes to the receiver. This combination can have the most serious consequences for the received signal, as it is affected by a strong adjacent field. This phenomenon takes place at the near end, which is why it stands out.

Total crosstalk attenuation. As noted earlier, some systems use all four pairs. When considering the crosstalk at the near end, we assumed that only two pairs are used. However, if all four pairs are active, as in the Gigabit Ethernet standard, they generate significantly more noise.

Figure 4. Cumulative crosstalk.

This is where we need such a characteristic as the total crosstalk. It takes into account the influence of all active pairs (see Figure 4). For example, we took a cable with four pairs. In the case of 25-pair backbone wiring, this value is even more important, as six times as many pairs can be potentially active.

Far-end crosstalk. Typically, data is transmitted in one direction, namely from the transmitting device to the receiving device. However, in some systems, data is transmitted in two directions. These systems are called full duplex systems. In this case, data is injected into the cable at both the near end and the far end at the same time. Therefore, in the case of full duplex transmission, noise occurs at both the near and far end. Because of this, far end crosstalk has been introduced in many new specifications.

Far-end noise is not easy to measure because a significant portion of the noise is lost or attenuated on the way to the test device. Therefore, it is standard practice to subtract the specific attenuation and take into account only noise. The magnitude of "noise minus attenuation" is called the reduced crosstalk at the far end.

Figure 5. Third-party crosstalk.

Third-party crosstalk. This term is used to describe crosstalk between cables. This effect is most noticeable when multiple pairs are active in the cable. In this case, the energy radiated from a single cable can be quite significant. In the example in Figure 5, six cables with four active pairs each surround another four-pair cable. The total number of active pairs is 24. Together, they can seriously interfere with the signal in the center cable. In this case, knowledge of third-party attenuation will be essential for the efficient operation of the network.

Crosstalk Summary:

Near-end crosstalk is so important because the transmit signal has the highest power at the near end and the receive signal has the least power. As a result, the receiving pair is particularly susceptible to interference from the transmitting pair. The total crosstalk takes into account the influence of several active pairs;
Far end crosstalk describes the consequences of full duplex operations when signals are generated simultaneously at the near and far ends. Third-party crosstalk determines the effects of crosstalk from other cables. This effect is most pronounced when multiple pairs in the cable are active.

IMPEDANCE AND RETURN LOSSES

Figure 6. Impedance as a function of frequency.

Impedance characterizes the data path. For example, if the signal is transmitted with an impedance of 100 ohms, then the structured wiring must correspond to an impedance of 100 ohms. Any deviation from this value will cause part of the signal to be reflected back to the data source. A change in impedance can be caused by a variety of reasons. One of them is non-compliance with the technology during the manufacturing process: any deviation from the provided distance between the conductors or violation of the properties of the insulating material can lead to a change in impedance (see Figure 6).

Figure 7. Impedance.

Another common cause is component mismatch. For example, a mismatch occurs when a switch cord with one impedance is connected to a horizontal conductor with a different impedance (see Figure 7a).

Such a mismatch will inevitably cause energy reflection at the break point (see Figure 7b). If impedance causes the possibility of a mismatch, then return loss characterizes its consequences. Return loss (measured in dB) allows you to find out how much of the signal is lost due to reflections.

Summary on impedance and return loss:

impedance characterizes the data path. Any deviation in the magnitude of the impedance results in signal reflection;
reflection means that instead of continuing on its way further forward, in reality the energy is reflected back to the transmitter;
this ultimately leads to a weakening of the forward propagating signal.

DELAY SKEW

Figure 8. Delay skew.

Another parameter that attracts considerable attention is latency skew. Delay skew refers to the timing of the signal paths across the different pairs in the cable (see Figure 8).

When all four pairs are active, signals must arrive consistently. The delay skew, measured in nanoseconds, characterizes the difference in the time of arrival of signals along different pairs of the cable. If this difference is too large, the receiving device will not be able to recover the signal. This will eventually lead to errors and data loss.

WHY IMPROVE WIRING?

The first improved versions of Category 5 wiring appeared about five years ago. Many of the parameters discussed above have been improved through the use of unique cable designs, in particular tighter strands and intracable fillers. The purpose of these enhancements was to prepare users for the coming changes in LAN technology.

When Category 5 first appeared, few systems really needed the operating frequency range it provided. For example, 10 Mbps Ethernet and 4 Mbps Token Ring were designed for Category 3 wiring. However, with the advent of new systems such as 100BaseT and 155 Mbps ATM, the need for Category 5 became apparent. Recently, new protocols, in particular ATM at 622 Mbit / s and 1000BaseT, make many people think about the sufficiency of Category 5 for their implementation. Hence the trend towards the improvement of UTP.

What is so special about these networks that their emergence led to a similar trend?

Increased data transfer rates. Systems such as 100BaseT and ATM at 155 Mbps are widely used in modern networks. Due to their complexity compared to 10BaseT and its counterparts, the wiring for these systems should provide less signal attenuation, have better immunity to interference, and generally be more consistent.

Complex coding schemes. In order to optimally distribute energy over the frequency range, 100BaseT systems use multi-level coding schemes. They have many advantages, in particular low noise levels. Unfortunately, the more complex the encoding scheme, the more sensitive the system is. Therefore, the cable should not have impedance discontinuities while still having good insulation.

Full duplex operation. On systems like 10BaseT, only one pair is active at a time. Data is transmitted over one pair, and data is received over the other. This mode of operation is called half duplex. Thanks to advances in technological process and electrical engineering, new systems can operate in full duplex mode, that is, signals can be transmitted and received simultaneously. This virtually doubles the bandwidth of the UTP cable. However, to do this, the cable must have stable impedance characteristics with minimal reflection and good crosstalk isolation between the near / far end pairs.

Using multiple pairs. In conventional networks, only two of the four pairs are active. In the meantime, throughput can be dramatically increased by using all four pairs of Category 5 cable. With clever electronics, data can be transmitted simultaneously over multiple pairs and restored at the point of reception. For this to be possible, the cable must provide as little interference as possible between pairs when the signal travels when all four pairs are active. This prompted the certification of Category 5 cables to meet the total attenuation parameters.

BRIEF SUMMARY

The whole debate about the need for improved wiring can be boiled down to two very simple questions.

How will advanced UTP wiring benefit my existing network?
How will the improved UTP wiring benefit when upgrading my network?

If someone is trying to sell you a better wiring solution, ask them to answer these two simple questions. If he is unable to do this, then his statements are nothing more than a marketing ploy. After all, improvements alone don't make sense. Your choice should directly depend on the real benefits of upgrading your network. And the key word here is "your". Not all enhancements are needed for your network. It is also important that the promised benefits become realized benefits.

Figure 9. Readings from the LeCroy NEWSLine test device.

Therefore, simply installing the advanced UTP wiring does not guarantee an increase in system performance. The user needs to demonstrate that these enhancements will enhance the capabilities of the network and / or improve its performance. Figure 9 shows the signal measurements with a LeCroy 100BaseT test device called NEWSLine. The cable used was Category 5. The bottom graph corresponds to the original signal, and the top one shows what the signal becomes after passing 100 m.

However, the question remains, what are the general implications for the network? There is no doubt that UTP is capable of providing a connection. Much more important, however, is the ability of UTP to transfer data in a consistent manner and without errors.

Table 3 shows the impact of errors on the throughput of a 100BaseT Ethernet network. An increase in data transmission errors of up to one percent has been found to result in an 80 percent reduction in throughput. Therefore, if improvements to UTP wiring can prevent errors from occurring, then a move to higher grade wiring is justified. Improvements in parameters such as total crosstalk, extraneous crosstalk, and signal strength can reduce the likelihood of errors in existing and future networks. However, these characteristics need to be demonstrated and explained to end users.

Reliably functioning UTP wiring increases in importance as data rates increase. Systems like 1000BaseT are potentially four times more sensitive than 100BaseT. In both cases, error prevention is imperative for the successful functioning of the network. Using devices such as the aforementioned LeCroy tester, end users can see how UTP affects network performance. And in some cases, switching to improved wiring can increase throughput by avoiding data transmission errors.

CONCLUSION

While the improved UTP wiring has the potential to expand the capabilities of both existing and future networks, questions still remain: "How much are these enhancements needed for your system and how can they help to take it to the next level?" Only by answering these two questions will you be able to distinguish real from imaginary needs.