GB2281400A - Cable testing - Google Patents

Abstract

A major characteristic which determines the suitability of a local network for digital transmission is the attenuation at 40KHz which should not exceed 37dB. Testing which can be carried out at the exchange enables the attenuation to be determined by measuring: (a) the capacitance as seen between the two wires of the twisted pair and between each wire of the twisted pair and earth and hence calculating the actual capacitance C between the two wires of the twisted pair; (b) the resistance of the loop with the subscribers terminal equipment off hook and subtracting therefrom a nominal value for the resistance R of the terminal equipment; (c) the leakage between the two wires of the twisted pair and between each wire of the pair and ground and hence determining the total leakage G; and estimating the inductance L as seen between the two wires of the twisted pair and hence calculating a value for the attenuation q using the formula:- <IMAGE>

Description

CABLE TESTING A major characteristic which determines the suitability of a given local network loop for the transmission of digital information at the basic rate is it's attenuation at 40 KHz. The maximum value specified in the presence of noise is 37dB. If the length, gauge, and type of conductor were known then the attenuation could be estimated. However due to the poor state of local loop records there is no reliable information regarding the make up of a particular subscriber's connection.

Extensive work has been carried out in defining the ISDN Basic Rate Access specification for digital access in the local loop which will be suitable for the vast majority of customer connections.

However there is no guarantee that any single line which currently provides problem free analogue access will be suitable for digital.

A major factor which determines the suitablility of a given local network loop for the transmission of digital information at the basic rate is it's attenuation at 40 KHz. ANSI National Standard T1-601-1988 Integrated Service Digitial Network basic access interface for use on metallic loops for application on the network side of the NT layer 1 specification; and Specification RC7355 British Telecommunications plc 2B1Q Electrical Interface Specification give a maximum value of loop attenuation, in the presence of noise, of 37 dB.

If the attenuation of a particular local loop is to be determined some tests are required on the customer's line. A method is proposed which will enable determination of the attentuation of a given subscriber's line. These tests can be carried out at the exchange end, hence avoiding a visit to the customer's premises.

According to the present invention there is provided a method of establishing the suitability of a local loop twisted pair for Integrated Services Digital Network (ISDN) Basic Access by determining the attenuation of the loop comprising measuring: (a) the capacitance as seen between the two wires of the twisted pair and between each wire of the twisted pair and earth and hence calculating the actual capacitance C between the two wires of the twisted pair; (b) the resistance of the loop with the subscribers terminal equipment off hook and subtracting therefrom a nominal value for the resistance R of the terminal equipment; (c) the leakage between the two wires of the twisted pair and between each wire of the pair and ground and hence determining the total leakage G; and calculating a value for the attenuation using the formula

The present invention will now be described, by way of example, with reference to the accompanying drawings, in which: Figure 1 shows examples of ISDN reference circuits for local loops; Figure 2 shows a schematic diagram of the interconnection of an analogue subscriber's handset and a local loop; Figure 3 shows the British Telecom Subscriber Terminal Apparatus DC Characteristic; Figure 4 illustrates the equivalent capacitance circuit of a local network loop; Figure 5 shows an equivalent local loop circuit with an NT5 socket attached; Figure 6 shows a simplified diagram of a subscriber's line for capacitance measurements; Figure 7 shows a typical discharge curve for a local network loop with an NT5 socket attached; Figure 8 is an equivalent circuit showing leakage in a local network loop; Figures 9a and 9b shows examples of loops used to compare estimated and measured attenuation; Figure A.1 shows an equivalent circuit for a section of a local network loop; and Figure A.2 shows a simplified version of the circuit of Figure A.1.

Firstly as discussed the feasibility of determining the attentuation of a local loop by obtaining values for its primary characteristics, i.e. resistance, capacitance, inductance and leakage.

Appendix A shows that the voltage Ex at any point along a transmission line can be determined by the following equation: Ex = E5 e Where: E5 = sending end voltage x = distance from sending end = = propagation constant for the line Considering a sinusoidal signal of radial frequency i) , being transmitted along a transmission line, the primary characteristics of which line are R = Resistance Per Unit Length.

G = Leakage Per Unit Length.

L = Inductance per unit length.

C = Capacitance per unit length.

The propagation constant of such a line is given by:

The attenuation can now be determined by multiplying the propagation coefficient by x, the distance from the sending end.

Considering a second transmission line whose length is one unit length, but whose primary characteristics are a factor x greater than those of the first transmission line, i.e: Rx = Resistance Per Unit Length.

Gx = Leakage Per Unit Length.

Lx = Inductance Per Unit Length.

Cx = Capacitance Per Unit Length.

Let gx be the propagation constant of such a line.

i.e. equations 1.1 and 1.2 are the same.

The above shows that for a single gauge pair, if the total resistance, capacitance, inductance and leakage can be determined, then the attenuation of that line can be determined, and the lack of knowledge of the length of the line is no longer important.

Local loops do not consist of a single section of conductor with a constant gauge, but will often consist of sections of conductor of varying length and type. A comparision needs to be carried out to see how close the idea if a continuous section approximates to those used for real pairs in the local network.

A loss which is totally ignored by the continuous section model is that due to reflections when the electrical signal changes from one type of conductor gauge to the next. The ISDN reference circuits in Figure 1 where chosen to given typical boundaries. The reflection coefficient equations are defined in Appendix A.

Boundary Attenuation at 40KHz 0.40mm Copper/ O.50mm Copper 0.5 dB 0.32mm Copper/ 0.40mm Copper 0.7 dB 0.50mm Copper/ 0.63mm Copper 0.3 db 0.40mm Copper/ 0.63mm Copper 0.9 dB 0.63mm Copper/ 0 .90mm Copper 0.3 dB 0.40mm Copper/ 0.90mm Copper 1.2 dB Table 1.1 Attenuation at 40 KHz due to reflections at conductor boundaries.

It can be seen from Table 1.1 that the attenuation due to reflections, compared to the attenuation of the signal within a section is small, and at the maximum attenuation limit of 37dB are negligible, so can be ignored. The estimated attenuation of various lines, which consist of separate sections, now needs to be compared to the estimated attentuation of the line when modelled by a continuous section of a single gauge conductor. The loops chosen were again the ISDN reference circuits which consisted of multiple section links. Table 1.2 makes this comparison.

Loop (Section X = 1.5 KM) Estimated Estimated Total Sectional Single Attenuation Conductor Attentuation LOOP 2 22 dB 22 dB LOOP 3 26 dB 27 dB LOOP 4 32 dB 35 dB LOOP 5 32 dB 33 dB LOOP 6 24 dB 27 dB Table 1.2 Comparison of estimated attenuation when loop is considered to consist of sections, and when loop is a continuous single section conductor.

From the above table it can be seen that the estimated attenuation of a sectional line is close to the estimated attenuation if the line were a continuous single section of conductor with constant gauge. Therefore this model is adequate for use when determining the attenuation of a multiple section subscriber loop.

Testing techniques which can be applied to the line which will determine the total primary characteristics of that line now need defining. However prior to this the method of connecting an analogue subscriber's handset to the local loop needs to be discussed.

Due to the multitude of types of subscriber's equipment that exist, and the numbers of extensions which can hang off a single telephone line some assumptions have to be made. In what follows a single telephone connected to the local loop via a 'white' master jackbox, which is known as an NT5 is assumed. Figure 2 gives a schematic diagram of such an interconnect.

When the telephone is 'on hook' the hook switch is open and the 1.8 tF capacitor appears as an open circuit to DC. However it provides an AC path, via the bell, when ringing current is applied.

When the phone is 'off hook', the hook switch is closed and a DC path is provided via the telephone's speech circuitry.

In an ideal world the telecommunications administration would be able to remotely short circuit the line at the customer's premises. However in the UK, and in most places throughout the world this is not possible. However Development of Centralised Loop Testing System For Subscriber Loops. Tamio Motomitsu, Kazao Hamazato, Satoshi Kimura, Tsunetaka Ema. NTT Review Vol 3 No 1 January 1991 gives details of the development of a system which will provide this functionaltiy to an administration. This means that the electrical characteristics of the termination need to be considered.

Figure 3 gives the DC characteristic of a subscriber's handset. This graph has been obtained from the British Telecom Network Requirement, BTNR, 315. This describes the physical and electrical specifications that a subscriber's handset should meet before it can be deemed suitable for use in the UK. As can be seen from Figure 3, the DC resistance must lie within an allowed range.

However for testing purposes a nominal value needs to be defined. A value of 300 Ohms will be chosen as being typical.

The loop resistance can now be estimated by doing the following: a) Take subscriber's handset off hook.

b) Measure the DC resistance of the line at the main distribution frame at the local exchange.

c) The resistance of the subscriber's loop is approximately 300 Ohms less than the measurement in step b).

It is customary to give each conductor in the pair the name of either the 'A' or 'B' leg. Capacitance measurement equipment at the main distribution frame will see the line capacitance as shown in Figure 4. There are always two capacitors in series in parallel with the one being measured. This implies an 'as seen' capacitance, hence to deduce the true capacitance, three separate measurements must be made and are described below.

1. The capacitance measurement between the 'A' and 'B' leg.

This will provide the following: C1 = CAB + CAE.CBE .... Eqn 1.3 CAE+CBE 2. The capacitance between the 'A' leg and earth, with the 'B' leg earthed. This provides: C2 = CAE + CAB.CBE ..... Eqn 1.4 CAB+CBE 3. The capacitance between the 'B' leg and earth, with the 'A' leg earthed. This provides: C3 = CBE + CABCAE ..... Eqn 1.5 CAB+CAE Solving the above simultaneous equations yields the required values for the three capacitances.

The capacitance measurement is achieved by initially charging up the line to a known voltage, then discharging it through a known resistor value. Again in the ideal world, an administration would be able to open circuit the local loop at the subscriber's premises, so isolating the customer's equipment from any measurements made. Unfortunately, as with the short circuit requirement when measuring resistance, this is not available so an alternative technique is required.

Consider the model of a local loop as shown in Figure 5, note that the loop is terminated with an NT5, without the subscriber's handset and the line has been charged up to 50 volts.

The line is now discharged through a 100 KOhm resistor. This value was chosen so that the line resistance is negligible when compared with it, so can be ignored. The rate of discharge is assumed to be small enough so the effect of line inductance can be ignored.

Figure 5 can now be simplified, and is shown-in Figure 6. The line capacitance must discharge through the line's leakage in parallel with the 100 KOhm resistor, while the capacitance of the NT5 must also discharge through the 470 KOhm resistor. Given that the line capacitance is much smaller than that of the NT5, it will discharge much faster. Therefore the discharge of the line can be distinguished from that of the NT5. Note that these capacitance measurements must be performed with the subscriber's handset unplugged from the NT5. This is to prevent the bell circuitry reducing the effect of the 470 KOhm resistor in distinguishing the discharge from the NT5 from that of the line.

A single discharge of the line is sketched out in Figure 7.

When the discharge reaches the 10 volt threshold, then the line can be thought of as being discharged. From the above graph the RC time constant can be determined, so therefore the capacitance of the line.

Note that the leakage of the line needs to be determined prior to the capacitance measurements. Once the three individual capacitances have been determined, then the total capacitance of the line can be found as below: CT = CAB + CAE . CBE CAE + CBE Which is the measurement 1 above.

Leakage in a local network pair can occur in three ways.

These are: 1. Leakage between the conductor pairs.

2. Leakage from the 'A' leg to ground.

3. Leakage from the 'B' leg to ground.

These are shown in Figure 8.

As with the capacitance measurements, these should be carried out with the subscriber's equipment unplugged from the NT5.

Three measurements are required, these are: 1. Resistance between the 'A' leg and 'B' leg. This gives: R1 = GAE + GBE .... Eqn 1.6 (GAE + GBE )xGAB + GAEXGBE 2. Resistance between 'A' leg and earth, with the 'B' leg earthed.

R2 = GAB + GBE .... Eqn 1.7 (GAB + GBE )xGAE + GABXGBE 3. Resistance between 'B' leg and earth, with the 'A' leg earthed.

R3 = GAB + GAE .... Eqn 1.8 (GAB + GAE )XGgE + GABXGAE Solving these three simultaneous equations will yield the three leakages indicated in Figure 8. Once these have been determined then the total leakage can be found. It is in fact that measured in test 1 above.

No test could be determined that would enable a measurement of the inductance of the line to be made. Fortunately inspection of the primary parameters in table 1.3 shows that the capacitance and inductance are approximately constant across most gauge types used in the UK. Therefore if the capacitance of the line is measured then an estimate of the lines inductance can be made. The exception to this is overhead cable, which is seldom used except in the rural network.

Considering urban installations, the majority of the cabling will be underground in copper, with possibly a short length (3-10 metres) of dropwire. Division of the measured capacitance by 50 nF and multiplication by the resulting equivalent length by 600JUH should provide a good estimate of the loop inductance.

Cable type R/Km | C/Kn 0 L/Km ohms | nF uH 0.32mm Cu Underground 435 50 600 0.40mm Cu Underground 273 50 600 0.50mm Cu Underground 168 50 600 0.63mm Cu Underground 109 50 600 0.90mm Cu Underground 55 50 600 0.50mm Al Underground 281 62 600 0.80mm Al Underground 111 64 600 1.2mm Overhead 32 4.7 2500 1.7mm Overhead 19 4.7 2500 2.5mm Overhead 9 5.1 2500 Dropwire No 1 65 58 600 Dropwire No 2 42 90 600 Dropwire No 3 290 52 600 Dropwire No 4 120 49 600 Dropwire No 5 225 38 600 Table 1.3 Primary parameters of conductor types used in the UK local network.

The System X telephone exchange system, which is used by many administrations in the UK, has local network testing equipment which provides the above measurements of a line's leakage, resistance, and capacitance. Having discussed the principles of the testing involved, a comparison needs to be made to see how close the estimated attenuation at 40 KHz, using the above, compares with the actual measured attenuation.

Due to the lack of suitable gauge wire to cover the ISDN reference circuits testing was limited to 0.40-, 0.50mm, and O.90mm gauge copper. The loops chosen are shown in Figures 9a and 9b.

Table 1.4 compares the measured attenuation of the line with the estimated attenuation.

LOOP | Estimated Attentuation | Measured Attenuation 1 9 dB 10 dB 2 18 dB 19 dB 3 26 dB 28 dB 4 35 dB 38 dB 5 6 dB 7 dB 6 9 dB 11 dB 7 4 dB 4 dB 8 8 dB 8 dB 9 13 dB 16 dB 10 18 dB 20 dB 11 22 dB 26 dB 12 36 dB 38 dB 13 4 dB 5 dB 14 33 dB 37 dB 15 37 dB 41 dB 16 50 dB 55 dB Table 1.4 Comparison of Measured Attenuation With Estimated Attenuation For Local Loop Copper Gauge Pairs.

As can be seen, the predicted attenuation of the subscriber's loop is within a few dBs of the measured attenuation.

It must be noted however these tests used cable pairs which were wrapped around wire drums and were not resident in multipair cables that are typical of the UK local network. This method will require verification using real pairs in the local loop.

This method can use existing equipment that is already widely available to administrations in the UK. Therefore there will be no need for them to make further investments in their local loop testing plant.

Appendix A: The voltage E along any point of a transmission line can be defined a: Ex = Ese .... Eqn Al.

Where: E5 = sending end voltage.

x = distance from sending end.

d is known as the propogation coefficient and is a complex term, and can be written as follows:

is is known as the attenuation coefficient and has units of nepers/unit length. This can be converted to dB/unit length by multiplying by 8.686.

is is known as the phase coeffient, and has units of radians per unit length.

A single local network line can be considered to consist of sections, a single section is shown in Figure A.1. This can be further simplified, and is shown in Figure A.2. The components shown represent the primary characteristics of the line.

Where: L = Inductance per unit length, and is measured in henries.

R = Resistance per unit length, and is measured in ohms.

C = Capacitance per unit length, and is measured in farads.

G = Leakage per unit length, and is measured in mohs.

The propogation coefficient can now be defined in terms of these primary parameters as:

where O is the radial frequency of the electrical signal.

Determining the modulus of Eqn A3 gives:

Equating equations A4 and A5 gives:

Also Eqn A3 gives:

Equating the real parts of Eqn A7 gives: &alpha; - = RG -# LC .... Eqn A8.

Adding Eqns A8 and A6 and rearranging gives:

Reflection Coefficient Current Reflection Coefficient.

Claims (2)

1. A method of establishing the suitability of a local loop twisted pair for Integrated Services Digital Network (ISDN) Basic Access by determining the attenuation of the loop comprising measuring: (a) the capacitance as seen between the two wires of the twisted pair and between each wire of the twisted pair and earth and hence calculating the actual capacitance C between the two wires of the twisted pair; (b) the resistance of the loop with the subscribers terminal equipment off hook and subtracting therefrom a nominal value for the resistance R of the terminal equipment; (c) the leakage between the two wires of the twisted pair and between each wire of the pair and ground and hence determining the total leakage G; and estimating the inductance L as seen between the two wires of the twisted pair and hence calculating a value for the attenuation using the formula

2. A method of establishing the suitability of a local loop twisted pair for ISDN Basic Access, substantially as hereinbefore described, with reference to and as illustrated in the accompanying drawings.