WPCL 2BJ|x ` Hx   x|@  6'6' @8'@8' Recommendation Q.552 $TRANSMISSION CHARACTERISTICS AT 2WIRE ANALOGUE INTERFACES 7OF A DIGITAL EXCHANGE 9Table of contentsă HHX X   1.General 2.Characteristics of interfaces 3.Characteristics of half connections  8 Annex A: Example of a longitudinal interference coupling network Recommendation Q.552 /TRANSMISSION CHARACTERISTICS AT 2WIRE ANALOGUE INTERFACES BOF A DIGITAL EXCHANGE 1.XHGeneral HThis Recommendation provides characteristics for: HH© 2wire analogue interfaces (Type C2 and Z),  `    HH© input and output connections with 2wire analogue interfaces, and  `  HH© halfconnections with 2wire analogue interfaces,  H in accordance with definitions given in Recommendation Q.551 particularly in Figure 1/Q.551. HThe characteristics of the input and output connections of a given interface are not necessarily the same. The characteristics of halfconnections are not necessarily identical for different types of interfaces. HThis Recommendation is valid for equipment that may terminate an international longdistance connection via 4wire circuits interconnected by 4wire exchanges. It also includes, in a separate category, characteristics for interfaces which cannot terminate an international connection and are therefore entirely national in application. 2.XHCharacteristics of interfaces Note For measuring 2wire analogue interface conditions it is necessary to apply a quiet code, i.e. a PCM signal corresponding to decoder output value 0(law) or output value 1 (Alaw), with the sign bit in a fixed state, to the exchange test point Ti, when no test signal is stipulated. 2.1HCharacteristics of interface C2 HThe recommended values of interfaces C2 are valid for digital exchanges including PABXs with transit functions and routing capabilities for originating and terminating traffic. Depending on the type of traffic to be handled, two different sets of relative levels are required. This suggests subdivision into C21 and C22 interface specifications. The interface C21 provides the termination of outgoing and incoming international long distance connections and possible national connections, with the exchange acting as transit switch. The interface C22 provides for the connection of a 2wire trunk line. A typical example is the interconnection of a Z interface with a C22 interface in a local exchange for routing through the 2wire analogue trunk network. A C22 interface cannot be part of the international 4wire chain (see Figure 2/Q.551). 2.1.1HExchange impedance HH2.1.1.1 Nominal value HNominal values of exchange impedance should be defined depending on national conditions. The definition shall include a test network for the exchange impedance. Administrations may want to adopt different test networks corresponding to the cable types used (e.g. unloaded and loaded). HH2.1.1.2 Return loss HThe return loss of the impedance presented by a C2 interface against the test network for the exchange impedance should comply with the limits given in Figure 1/Q.552. HFFrequency (log f) HN HGFIGURE 1/Q.552 HN H2Minimum value of return loss against the test network foră H8the exchange impedance at a 2wire interfaceă 2.1.2HImpedance unbalance about earth HThe longitudinal conversion loss (LCL), defined in RecommendationG.117, 4.1.3, should exceed the minimum values of Figure2/Q.552 with the equipment under test in the normal talking state, in accordance with Recommendation K.10. Note 1 An Administration may adopt other values and in some cases a wider bandwidth, depending on actual conditions in its telephone network.   Note 2 A limit may also be required for the transverse conversion loss (TCL), as defined in Recommendation G.117, 4.1.2, if the exchange termination is not w. reciprocal with respect to the transverse and longitudinal paths. A suitable limit would be 40 dB to ensure an adequate nearend crosstalk attenuation between interfaces. JFIGURE 2/Q.552 Q 9Minimum values of LCL measured in the arrangementă Fshown in Figure 3/Q.552ă Test method HLongitudinal conversion loss should be measured in accordance with the principles given in Recommendation O.121, 2.1 and 3. Figure 3/Q.552 shows an example of the basic measuring arrangement for digital exchanges. HMeasurements of the longitudinal and transverse voltages should preferably be done with a frequencyselective level meter. JFIGURE 3/Q.552 Q CArrangement for measuring LCLă 2.1.3HLongitudinal interference threshold level HUnder study. 2.1.4HRelative levels HH2.1.4.1 Nominal levels 2.1.4.1.1Interface C21 HC21 interfaces should meet the recommended values for Z interfaces in 2.2.4.1 if no loss compensation comparable to 2.2.4.3 is provided. 2.1.4.1.2Interface C22 HTo adjust the transmission loss of a digital transmission section to the values of national transmission planning for local or national traffic, depending on the relative levels given in 2.1.4.1.1 and 2.2.4.1, the following ranges encompass the requirements for C22 interfaces of a large number of administrations: HH© input level:Li = +3.0 to 7.0 dBr in 0.5dB steps;  `  HH© output level:Lo = +1.0 to 8.0 dBr in 0.5 dB steps. HAccording to Annex E of Recommendation G.121 (column 2 of TableE1/G.121), the range of transmission loss from 1.0 to 8.0 dB for the digital transmission section encompasses the requirements of a large number of administrations.   HIn order to compensate loss on long toll or junction lines, an administration may, to satisfy local conditions, choose values of relative levels derived from the basic values as follows: ' HHLi = Li + x dB,  `  ' HHLo = Lo x dB,   where x should take a negative value. The value of x is in national competence. Such compensation of loss requires careful selection and application of balance networks. HIt has been recognized that it is not necessary for a particular design of equipment to be capable of operating over the entire level range. HH2.1.4.2 Tolerances of relative levels HThe difference between the actual relative level and the nominal relative level should lie within the following values: HH© input relative level:0.3 to +0.7 dB;  `  HH© output relative level:0.7 to +0.3 dB.  ( HThese differences may arise, for example, from design tolerances, cabling between analogue ports and the (DF), and adjustment increments.   Note Level adjustment procedures are given in Recommendation G.715, 2.1. 2.2HCharacteristics of interface Z HThe recommended values of interface Z are valid for digital local exchanges, PABXs and digital remote units. For PABXs, see RecommendationQ.551, w. section 2.1.1. 2.2.1HExchange impedance HH2.2.1.1 Nominal value HThe principal criterion governing the choice of the nominal value of the exchange impedance is to ensure an adequate sidetone performance for telephone sets, particularly those operated on short lines. If this criterion is met, the impedance will also be suitable for subscriber lines fitted with voice band modems. HAs a general rule a complex exchange impedance with a capacitive reactance is necessary to achieve satisfactory values of stability, echo and sidetone. For additional information, see Supplement No. 10, FascicleVI.5 of the CCITT Blue Book and Recommendations G.111 and G.121. HThe use of the preferred configuration below will minimize the diversity of types of exchange impedances. At present no unique component values can be recommended. However, to provide guidance for administrations, examples of nominal values chosen by some administrations are given in Table 1/Q.552. M GTABLE 1/Q.552 M 2Test networks for exchange impedances being consideredă Note 1 The test network and the component values represent a configuration that exhibits the required exchange impedance. It need not necessarily correspond to any actual network provided in the exchange interface. Note 2 The range of component values reflects the fact that there are substantial differences in the sensitivity and sidetone performance of the various telephone instruments throughout the world. In general, the combination of short lines and sensitive telephone sets might be rather common in the future due to increased use of remote concentration. In order to control sidetone performance, administrations need to take into account telephone set parameters. Not only should the parameters of existing telephone sets be considered but also the parameters that may be desirable in the future to allow improvement in sidetone performance to be achieved. Note 3 It may be necessary to group the subscriber lines of a particular exchange into classes, each requiring a different exchange impedance of the Zinterface. HH2.2.1.2 Return loss HTolerances are needed for values of exchange impedance. For this purpose the return loss of the impedance presented by a 2wire port against the test network for the exchange impedance should comply with limits which depend on the particular conditions of the subscriber network considered. These are given in the template of Figure 1/Q.552. HSome administrations may want to specify higher values. Examples of limit values for the return loss, currently accepted by some administrations, are given in Table 2/Q.552 for guidance. GTABLE 2/Q.552 M 6Examples of limit values of return loss againstă Bthe exchange impedanceă FRGH14 dB at 300 Hz, rising (log scale) to 15 dB at 500 Hz remaining at 18dB to 2000 Hz and then falling (log scale) to 14 dB at 3400Hz. NTTH22 dB: 3003400 Hz. BTXH15 dB: 200800Hz: 20 dB: 8002000Hz: 20004000Hz. USAH20 dB: 200500Hz: 26 dB: 5003400Hz.  HHAustria 14.5 dB at 300 Hz, rising (log scale) to 15 dB at 500 Hz remaining at 18 dB to 2500Hz and then falling (log scale) to 14.5dB at 3400Hz. Note The 12 dB spread in values stems from the difference in telephone set sensitivities. 2.2.2HImpedance unbalance about earth HThe longitudinal conversion loss (LCL) of the Zinterface should meet the values given in 2.1.2 and Figure 2/Q.552, measured in accordance with the test method given in Figure 3/Q.552. 2.2.3HLongitudinal interference threshold levels HThe signalling and transmission performance of the Zinterface can be degraded when the subscriber line is exposed to an electromagnetic field of sufficiently high intensity. The value of induced interference energy causing performance degradation may be below a level which would cause permanent damage or operate protective devices. Longitudinal interference may come from power or traction lines or radio frequency sources. HRadio frequency interference tests at the Z interface should be in accordance with Recommendations of the KSeries (intended by Study GroupV). HLongitudinal interference tests relative to power and traction line sources should be performed according to Figure 4/Q.552. HInterference up to the interference threshold level should not affect signalling and transmission more than the limits stated below. Measurements should be performed using quiet code at the exchange test point Ti. HThere are two groups of parameters to be observed while performing the tests: HHi)h  signalling related parameters;  `  w.Ԍ  HHii)  transmission related parameters, i.e. noise parameters.  `    HFor group i) the performance of the signalling parameters mentioned in Recommendation Q.543 should be tested in a go no go procedure under normal operating conditions. HFor group ii) two test steps should be performed under normal operating conditions, the first step without and the second one with the longitudinal test generator connected to the coupling network. The additional noise in the second test step should not contribute more than:   HHLEN = Y1 pWp using sinusoidal longitudinal test signal with X1 volts HHX rms;  `   H HHLEN = Y2 pWp using longitudinal EMF test signal with defined harmonic HH X  content (e.g. triangular waveform with X2 volts zero to peak).  `    HThe values Y1 and Y2 of the noise power must be specified depending on the interface the noise measuring set is connected to, i.e. the analogue interface at the termination T representing subscriber apparatus or the digital interface at the exchange test point To. The noise measuring set should be provided with a notch filter to exclude the activating signal at the nominal reference frequency. HThe associated noise level limit results from the use of the equations given in 3.3.2.1 and 3.3.3 of this Recommendation. Note 1 The values of X1 and X2 need further study. (Some administrations reported an X1 value of 15 volts and an X2 value of 25 volts.) Note 2 The value of the induced noise power LEN needs further study. (Attention is drawn to 3.1.6.2 of this Recommendation and to section1 of Recommendation G.123.) Test method HHFIGURE 4/Q.552 HO H6Arrangement for measuring longitudinal interferenceă HHthreshold levelă HThe longitudinal interference test generator should provide the longitudinal interference EMF with the fundamental frequency of the interference source (as appropriate to national conditions, i.e. 16 2/3 Hz, 50 Hz or 60 Hz) with a sinusoidal waveshape, and additionally with a waveshape having a certain amount of harmonic content(*), e.g. a triangular waveshape. HThe coupling network CN(*) should represent a typical subscriber line (length, type of cable) exposed to power or traction line interference. The impedance of the coupling path within the network should be primarily capacitive. (One RPOA reported an impedance of j 1.17 kOhm at 60 Hz for each capacitor indicated in Figure 4/Q.552.) HThe termination T representing subscriber apparatus should provide for an appropriate loop current and the requested internal impedance of the reference frequency signal generator. Note 1 Annex A gives an example of a CN applicable to the measuring arrangement of Figure 4/Q.552, the application of which needs further study. Note 2 The measuring arrangement in Figure 4/Q.552 covers the general use of subscriber equipment, as recommended in Recommendation K.4, without low impedance to earth, especially without signalling using earth return. National deviations from this general case need to be considered for each special type of subscriber circuit.    h (*) The exact definition of the harmonic content and the coupling network is for further study. 2.2.4HRelative levels HOperation of the Z interface in the ranges of relative levels given below is recommended when the interface terminates an entirely 4wire international longdistance connection. Pairs of input and output levels can be chosen for internal, local, or national longdistance traffic in a wider range if these connections can be discriminated from international ones for correct level switching. If digital pads are used, the additional distortion must be considered (see Recommendation G.113, Table 1/G.113). HIn assigning the relative levels for international longdistance connections to the interface it should be noted that:  H HH© The limiting of "difference in transmission loss between the two directions of transmission" in Recommendation G.121, 6.4 must be taken into account. For the national extension this is the value "loss (tb)loss(at)". (See the text in the cited Recommendation for guidance.) This difference is limited to +4dB. However, to allow for additional asymmetry of loss in the rest of the national network, only part of this difference can be used by the digital exchange.  `   8 HH© If within the ranges of Li and Lo given under 2.2.4.1.1 and 2.2.4.1.2, the values are chosen such that Li Lo > 6 dB, and if adequate balance networks are used (e.g. 3.1.8 and Figure11/Q.552), the requirements of Recommendation G.121, 6 (Incorporation of PCM digital processes in national extensions) as well as for Recommendation G.122 (Stability and echo loss) will be satisfied.  `  w.Ԍ HH2.2.4.1 Nominal levels 2.2.4.1.1Input relative level  8 HAccording to Annex C to Recommendation G.121 (columns 1, 2 and 3 of Table C1/G.121), the following range of input relative level for all types of connections (internal, local, national and international) encompasses the requirements of a large number of administrations. HHLi = 0 to +2.0 dBr.  `   X Note 1 Recommendation G.101, 5.3.2.3 indicates that if the minimum nominal send loudness rating (SLR) of the local system under the same conditions is not less than 1.5dB, then the peak power of the speech will be suitably controlled. It follows that, for instance, the value Li = 0 dBr (lower limit of the range for Li) is suited to a send loudness rating > 1.5 dB. Note 2 The values given above are in conformity with current national practices and with the existing text of Recommendation G.101. However, the latter is itself partly based on a very old investigation (which Study GroupXII has been asked to review) of the relationship between loudness ratings and speech levels. This may, in the near future, lead to amending the basis of objectives, so that it may be useful to allow wider design margins. 2.2.4.1.2Output relative level HAccording to Annex C to Recommendation G.121 (column 3 of Table C1/G.121), the following range of output relative level for international longdistance connections encompasses the requirements of a large number of administrations. HHLo = 5.0 to 8.0 dBr.  `   X HThe chosen value may be used for connections entirely within a national network as well. HIf the connection type can always be detected, the nominal output relative levels for local or national connections can take other values in accordance with national transmission planning. According to Annex C to Recommendation G.121 (columns 1 and 2 of Table C1/G.121) the following range encompasses the requirements of a large number of administrations: HHLo = 0 to 8.0 dBr.  `   X HIt has been recognized that it is not necessary for a particular design of equipment to be capable of operating over the entire range. HH2.2.4.2 Tolerances of relative levels HThe difference between the actual relative level and the nominal relative level should lie within the following limits: HH© input relative level:0.3 to +0.7 dB,  `  HH© output relative level:0.7 to +0.3 dB.  ( HThese differences may arise, for example, from design tolerances, cabling (between analogue ports and the DF) and adjustment increments. Short term variation of loss with time as discussed in 3.1.1.3 is not included. Note Procedures for adjusting relative level are given in RecommendationG.715, 2.1. HH2.2.4.3 Consideration of short and long subscriber lines   HIn order to compensate for the loss of short or long subscriber lines, an administration may choose values of the relative levels derived from the basic values as follows: HH ' HHLi = Li + x dB  `  HH ' HHLo = Lo x dB.  H HThe value of x is within national competence (e.g. x = 3 dB for short subscriber lines). HH 'Ġ'  8 HIf values of Li and Lo are chosen as indicated, the loss difference with respect to the conditions given in 2.2.4.1 will be left unchanged. HThe use of values of x < 0 requires careful selection of balance networks; values of x < 3 dB are not recommended. 3.XHCharacteristics of half connections HFor interfaces C2 this Recommendation is valid for digital local and transit exchanges and for C21 interfaces of PABXs connected to the digital local exchange by a digital transmission system. HFor interface Z this Recommendation is valid for digital local and combined local/transit exchanges, for PABXs and for digital remote units, each connected to the digital local exchange by a digital transmission system. For further information concerning PABXs, see Recommendation Q.551, section2.1.1.  X Note In measuring an input connection it is necessary to apply a quiet code, i.e. a PCM signal corresponding to decoder output value 0 (law) or output value 1 (Alaw) with the sign bit in a fixed state to the exchange test point Ti. (See Recommendation Q.551, section 1.2.3.1.) 3.1HCharacteristics common to all 2wire analogue interfaces 3.1.1HTransmission loss HH3.1.1.1 Nominal value HThe nominal transmission loss according to Recommendation Q.551, 1.2.4.1 is defined in 3.2.1 and 3.3.1 for input and output connections of half connections with a 2wire analogue interface. HH3.1.1.2 Tolerances of transmission loss HThe difference between the actual transmission loss and the nominal transmission loss of an input or output connection, according to 2.1.4.2 and 2.2.4.2 should lie within the following range: HH©0.3 to +0.7 dB.  `   ( HThese differences may arise, for example, from design tolerances, cabling (between analogue equipment ports and the DF) and adjustment increments. Shortterm variation of loss with time as discussed in 3.1.1.3 is not included. w.Ԍ HH3.1.1.3 Shortterm variation of loss with time  X HWhen a sinewave test signal at the reference frequency of 1020 Hz and at a level of 10 dBmO is applied to the 2wire analogue interface of any input connection, or a digitally simulated sinewave signal of the same characteristic is applied to the exchange test point Ti of any output connection, the level at the corresponding exchange test point To and the 2wire analogue interface respectively should not vary by more than +0.2 dB during any 10minute interval of typical operation under the steady state condition permitted variations in the power supply voltage and temperature. HH3.1.1.4 Variation of gain with input level HWith a sinewave test signal at the reference frequency 1 020 Hz and at a level between 55 dBmO and +3 dBmO applied to the 2wire analogue interface of any input connection, or with a digitally simulated sinewave signal of the same characteristic applied to the exchange test point Ti of any output connection, the gain variation of that connection, relative to the gain at an input level of 10dBmO, should lie within the limits given in Figure 5/Q.552. HThe measurement should be made with a frequencyselective level meter to reduce the effect of the exchange noise. This requires a sinusoidal test signal. IFIGURE 5/Q.552 P ?Variation of gain with input levelă HH3.1.1.5 Loss distortion with frequency HThe loss distortion with frequency of any input or output connection according to Recommendation Q.551, 1.2.5 should lie within the limits shown in the mask of Figure 6a or 6b/Q.552 respectively using an input level of 10dBmO. Note The limits of this clause shall not apply to Z halfconnections which include equalization for the distortion in the subscriber line. IFIGURE 6a/Q.552 ALoss distortion with frequencyă HInput connectionă P *In the marked frequency ranges relaxed limits are shown which apply if the maximum length of instation cabling is used. The more stringent limits shown apply if no such cabling is present.  w.Ԍ IFIGURE 6b/Q.552 P ALoss distortion with frequencyă HOutput connectionă * In the marked frequency ranges relaxed limits are shown which apply if the maximum lengths of instation cabling is used. The more stringent limits shown apply if no such cabling is present. 3.1.2HGroup delay H"Group delay" is defined in the Yellow Book, Fascicle X.1. 3.1.2.1 Absolute group delay HSee Recommendation Q.551, section 3.3.1. 3.1.2.2 Group delay distortion with frequency HTaking as the reference the minimum group delay, in the frequency range between 500Hz and 2500Hz, of the input or output connection, the group delay distortion of that connection should lie within the limits shown in the template of Figure 7/Q.552. Group delay distortion is measured in accordance with Recommendation O.81. IFIGURE 7/Q.552 P :Group delay distortion limits with frequencyă HThese requirements should be met at an input level of 10 dBmO. 3.1.3HSingle frequency noise HThe level of any single frequency (in particular the sampling frequency and its multiples), measured selectively at the interface of an output connection, should not exceed 50 dBmO. Note See Recommendation Q.551,  1.2.3.1. 3.1.4HCrosstalk HFor crosstalk measurements, auxiliary signals are injected as indicated in Figures 8 and 9/Q.552. These signals are: H the quiet code, (see Recommendation Q.551, section1.2.3.1); H a low level activating signal. Suitable activating signals are, for example, a band limited noise signal (see RecommendationO.131), at a level in the range 50 to 60 dBmO or a sinewave signal at a level in the range from 33 to 40 dBmO. Care must be taken in the choice of frequency and the filtering characteristics of the measuring apparatus in order that the activating signal does not significantly affect the accuracy of the crosstalk measurement. 3.1.4.1 Input crosstalk HA sinewave test signal at the reference frequency of 1020 Hz and at a level of O dBmO, applied to an analogue 2wire interface, should not produce a level in any other half connection exceeding 73 dBmO for nearend crosstalk (NEXT) and 70dBmO for farend crosstalk (FEXT) (see Figure 8/Q.552).  w.Ԍ IFIGURE 8/Q.552 P 0Measurement with analogue test signal between different equipmentă 3.1.4.2 Output crosstalk HA digitally simulated sinewave test signal at the reference frequency of 1020Hz applied at a level of 0 dBmO to an exchange test point Ti, should not produce a level in any other half connection exceeding 70dBmO for nearend crosstalk (NEXT) and 73dBmO for farend crosstalk (FEXT) (see Figure9/Q.552). IFIGURE 9/Q.552 P 0Measurement with digital test signal between different equipmentă 3.1.5HTotal distortion including quantizing distortion HWith a sinewave test signal at the reference frequency of 1020Hz (see Recommendation O.132) applied to the 2wire interface of an input connection, or with a digitally simulated sinewave signal of the same characteristic applied to the exchange test point Ti of an output connection, the signaltototal distortion ratio, measured at the corresponding outputs of the half connection with a proper noise weighting (see Table 4 of RecommendationG.223) should lie above the limits given in 3.2.3, Figures13 and 14/Q.552 for interface C2 and 3.3.3, Figure15/Q.552 for interface Z. Note The sinusoidal test signal is chosen to obtain results independent of the spectral content of the exchange noise. 3.1.6HDiscrimination against outofband signals applied to the input interface  `  H(Only applicable to input connections.) 3.1.6.1 Input signals above 4.6 kHz   HWith sinewave signal in the range from 4.6kHz to 72kHz applied to the 2wire interface of an input connection at a level of 25dBmO, the level of any image frequency produced in the time slot corresponding to the input connection should be at least 25dB below the level of the test signal. This value may need to be more stringent to meet the overall requirement. 3.1.6.2 Overall requirement HUnder the most adverse conditions encountered in a national network, the half connection should not contribute more than 100pWOp of additional  h noise in the band 10Hz4kHz at the output of the input connection, as a result of the presence of outofband signals at the 2wire interface of the input connection. 3.1.7HSpurious outofband signals received at the output interface H(Only applicable to an output connection.) 3.1.7.1 Level of individual components HWith a digitally simulated sinewave signal in the frequency range 3003400Hz and at a level of 0dBmO applied to the exchange test point Ti of a half connection, the level of spurious outofband image signals measured selectively at the 2wire interface of the output connection should be lower than 25 dBmO. This value may need to be more stringent to meet the overall requirement. 3.1.7.2 Overall requirement HSpurious outofband signals should not give rise to unacceptable interference in equipment connected to the digital exchange. In particular, the intelligible and unintelligible crosstalk in a connected FDM channel should not exceed a level of 65dBmO as a consequence of spurious outofband signals at the half connections. 3.1.8HEcho and stability HTerminal Balance Return Loss (TBRL) as defined in 3.1.8.1 is introduced in order to characterize the exchange performance required to comply with the w. network performance objective of Recommendation G.122 with respect to echo. The TBRL of an equipment port is measured in the talking state as in an established connection through a digital exchange. HThe parameter "Stability Loss", as defined in RecommendationG.122, applies to the worst terminating conditions encountered at a 2wire interface in normal operation. 3.1.8.1 Terminal Balance Return Loss (TBRL) HThe term TBRL is used to characterize an impedance balancing property of the 2wire analogue equipment port. HThe expression for TBRL is: where HZo = exchange impedance of a 2wire equipment port HZb = impedance of the balance network presented at a 2wire equipment M$N port HZt = impedance of the balance test network HSome administrations have found that it is advantageous to choose Zo = Zb in order to optimize TBRL. In this case the expression reduced to and the balance test network will be identical to the test network for the exchange impedance. HThe balance test network should be representative of the impedance conditions to be expected from a population of terminated lines connected to 2wire interfaces, as determined by the national transmission planning. HThe TBRL is related to the loss aio between the exchange test point Ti and To of a half connection as follows: where ao and ai are the losses between the exchange test point Ti and the 2wire port and between the 2wire equipment port and the exchange test point To, respectively. HTBRL can thus be determined by measurement of aio provided the sum (ao + ai) is known. This can be derived in several ways:   HHa)h  ao and ai are assigned their nominal values NLo and NLi as defined in 3.2.1 and 3.3.1. Then:  `    Hb)h  ao is measured with the load matched to the exchange impedance as actual transmission loss ALo and ALi (see 3.1.1.2). Then: Hc)h  the loss aio is measured with the 2wire equipment port open and shortcircuited, giving losses a'io and a"io, respectively. Then: HMethod b) provides the most accurate results. HHFIGURE 10/Q.552 HO H<Arrangement for measuring the loss aioă HUsing the arrangement of Figure 10/Q.552 and sinusoidal test signals, the measured TBRL should exceed the limits shown in Figure 11/Q.552. HHFIGURE 11/Q.552 HO HHLimits for TBRLă HFigure 12/Q.552 gives examples of balance test networks adopted by some administrations for unloaded subscriber lines. These examples may provide guidance for other administrations in order to minimize the diversity of types of test networks.  w. Ԍ HHFIGURE 12/Q.552 HO H1Examples of test networks to be used by some administrationsă H;(applicable to unloaded subscriber lines) Note Some administrations may need to adopt several balance test networks to cover the various types of unloaded and loaded cables. 3.1.8.2 Stability loss HThe stability loss should be measured between the exchange test points Ti and To of a half connection (Figure 10/Q.552) by terminating the 2wire interface with stability test networks representing the "worst terminating condition encountered in normal operation". Some administrations may find that open and shortcircuit terminations are sufficiently representative of worst case conditions. Other administrations may need to specify, for example, an inductive termination to represent the worst case condition. HWith worst case terminating conditions on the 2wire interface of a half connection, the stability loss Ti to To measured as aio should be: HHStability Loss = aio > x;  `    where x is under study for sinusoidal signals at all frequencies between 200Hz and 3600Hz. This frequency band is determined by the filters used in the interface designs. HThe need for requirements outside this frequency band is also under study. HWhere the digital exchange is connected to the international chain using only 4wire switching and transmission, the half connection of the digital exchange may provide the total stability loss of the national extension. The value of stability loss (SL) that is required for a 2wire interface is a matter of national control provided that the requirements of Recommendation G.122 are met. A SL value of 6 dB at all frequencies between 200 Hz and 3600Hz will ensure that the G.122 requirements are met. However, SL values of between 6 dB and 0 dB will formally comply with the present requirements of G.122 (Red Book 1984) but further study is required to provide guidance in this area. One administration has found that a value of 3 dB is satisfactory in its environment. Note It is suggested that the half connection of a digital PABX, or of a digital remote unit, when connected to the digital local exchange by a digital transmission system, should also meet the requirements of 3.1.8. 3.2HCharacteristics of interface C2 3.2.1HNominal value of transmission loss HAccording to the relative levels defined in 2.1.4.1, the nominal transmission losses of input or output connections NLi and NLo of a half connection with C2 interfaces are in the following ranges: HC21 interfaces HNLi = 0 to 2.0 dB for all types of connections HNLo = 5.0 to 8.0 dB for international connections H 0 to 8.0 dB for local or national connections HC22 interfaces HNLi = 3.0 to 7.0 dB H for all types of connections HNLo = 8.0 to 1.0 dB HIt has been recognized that it is not necessary for a particular design of equipment to be capable of operating over the entire range of nominal transmission losses. HIf a loss compensation is applied the nominal loss NLi and NLo should be corrected by the value of x dB chosen in connection with sections2.1.4.1.2 or 2.2.4.3. 3.2.2HNoise 3.2.2.1 Weighted noise HFor the calculation of noise, worst case conditions at the C2 interface are assumed. The band limiting effect of the encoder on the noise was not taken into account. For a more exact calculation further study is necessary. 3.2.2.1.1 Output connection HTwo components of noise must be considered. One of these arises from the quiet decoder, the other from analogue sources, such as signalling equipment. The first component is limited by RecommendationG.714, 10 as receiving equipment noise to 75dBmOp; the other component by RecommendationG.123, 3 to (67+3) dBmOp = 70dBmOp for one 2wire analogue interface. This results in the maximum value for the overall weighted noise in the talking state at the C2 interface of a digital exchange of: H68.8 dBmOp for equipment with signalling on the speech wires, H75.0 dBmOp for equipment with signalling on separate wires. 3.2.2.1.2 Input connection HTwo components of noise must be considered. One of these arises from the encoding process, the other from analogue sources, e.g. signalling equipment. The first component is limited by RecommendationG.714, 9 as idle channel noise to 66dBmOp; the other component by Recommendation G.123, 3 to (67+3)dBmOp = 70dBmOp for one 2wire analogue interface. This results in the maximum value for the overall weighted noise in the talking state at the exchange test point To of a digital exchange of: H64.5 dBmOp for equipment with signalling on the speech wires, H66.0 dBmOp for equipment with signalling on separate wires.  w. Ԍ3.2.2.2 Unweighted noise HThis noise will be more dependent on the noise on the power supply and on the rejection ratio. Note The need for and value of this parameter are both under study. RecommendationsQ.45bis, 2.5.2 and G.123, 3 must also be considered. 3.2.2.3 Impulsive noise HIt will be necessary to place limits on impulsive noise arising from sources within the exchange; these limits are under study. Pending the results of this study, RecommendationQ.45bis, 2.5.3 may give some guidance on the subject of controlling impulsive noise with low frequency content. Note 1 The sources of impulsive noise are often associated with signalling functions (or in some cases the power supply) and may produce either transverse or longitudinal voltage at C2 interfaces. Note 2 The disturbances to be considered are those to speech or modem data at audio frequencies, and also those causing bit errors on parallel digital lines carried in the same cable. This latter case, involving impulsive noise with high frequency content, is not presently covered by the measurement procedure of RecommendationQ.45bis. 3.2.3HValues of total distortion HThe total distortion including quantizing distortion of a half connection with a C2 interface is measured in accordance with 3.1.5. HThe signaltototal distortion ratio for a half connection at interface C2 should lie above the limits shown in Figure13/Q.552 for equipment with signalling on separate wires, and in Figure14/Q.552 for equipment with signalling on the speech wires both measured in the talking state. JFIGURE 13/Q.552 Q 0Limits for signaltototal distortion ratio as a function of inpută 0level; input or output connection with signalling on separate wiresă Q JFIGURE 14/Q.552 Q 0Limits for signaltototal distortion ratio as a function of inpută /level; input or output connection with signalling on the speech wiresă Q HThe values of Figure 14/Q.552 include the limits for the encoding process given in Figure 4/G.714 and the allowance for the noise contributed via signalling circuits from the exchange power supply and other analogue sources (e.g. analogue coupling), which is limited to (67+3)dBmOp = 70dBmOp for one C2 analogue interface by RecommendationG.123, 3. 3.3HCharacteristics of interface Z 3.3.1HNominal value of transmission loss HAccording to the relative levels defined in 2.2.4.1, the nominal transmission losses of input or output connections NLi and NLo of a half connection with Z interfaces are in the following ranges: HNLi = 0 to 2.0 dB for all types of connections HNLo = 5.0 to 8.0 dB for international connections H 0 to 8.0 dB for internal, local or national connections. HIf a compensation for the loss of short or long subscriber lines is applied, the nominal loss NLi and NLo should be corrected by the value of xdB chosen in connection with 2.2.4.3. 3.3.2HNoise 3.3.2.1 Weighted noise HFor the calculation of noise, worst case conditions at the CZ interface are assumed. The band limiting effect of the encoder on the noise has not been taken into account. For a more exact calculation further study is necessary. 3.3.2.1.1 Output connection HTwo components of noise must be considered. One of these, e.g. noise arising from the decoding process, is dependent upon the output relative level. The other, e.g. power supply noise from the feeding bridge, is independent of the output relative level. The first component is limited by RecommendationG.714, 10 as receiving equipment noise to 75dBmOp; the other component is assumed by RecommendationG.123, AnnexA to be 200pWp (67dBmp). This can be caused by the main DC power supply and auxiliary DCDC converters.  w. ԌHInformation about the subject of noise on the DC power supply is given in SupplementNo.13 to the GSeries Recommendations (Orange Book, VolumeIII3). HThe total psophometric power allowed at a Z interface with a relative output level of Lo dB may be approximated by the formula: HThe total noise level is given by: where  x HHX HPTNo: total weighted noise power for the output connection of the local digital exchange;  `    HHX HPAN: weighted noise power caused by analogue functions according to Recommendation G.123, Annex A for local exchanges, i.e. 200pWp;  `    HHX HLINo: receiving equipment noise (weighted) for PCM translating equipment according to Recommendation G.714, 10, i.e. 75dBmOp;  `    HHLo:  output relative level of a half channel of a local digital exchange according to 2.2.4.1.2, e.g. 0 to 8.0dBr;  `   x HHX HHLTNo: total weighted noise level for the output connection of the local digital exchange. HFor the range of output relative levels according to 2.2.4.1.2 the resulting total psophometric powers and the total noise levels for the output connection are: HLoĠ=05.06.07.08.0dBr HPTNo =231210208206205pWp HLTNo =66.466.866.866.966.9dBmp 3.3.2.1.2 Input connection HTwo components of noise must be considered. One of these, e.g. noise arising from the encoding process, is dependent upon the output relative level. The other, e.g. power supply noise from the feeding bridge, must be corrected by the input relative level for calculation at the exchange test point To. The first component is limited by RecommendationG.714, 9 as idle channel noise to 66dBmOp; the other component is assumed by RecommendationG.123, AnnexA to be 200pWp (67 dBmp) which results in 67dBmp Li at the exchange test point To. HThe total psophometric power allowed at the exchange test point To with a relative input level of Li dB may be approximated by the formula: LiĠ90 + LINi  (à ) HPTNi = PAN . 1010 + 1010pwOp and the total noise level by PTNi HLTNi = 10 log (  ) 90 dBmOp 1 pW where HHX HHPTNi: total weighted noise power for the input connection of the local digital exchange; HHX HHPAN: weighted noise power caused by analogue functions according to Recommendation G.123, Annex A for local exchanges, i.e. 200pWp; HHX HHLINi: idle channel noise (weighted) for the input connection of a digital local exchange according to RecommendationG.714, 9 i.e. 66dBmOp; HLi:  input relative level of a half channel of a local digital exchange according to 2.2.4.1.1, e.g. 0 and +1dBr; HHX HHLTNi: total weighted noise level for the input connection of the local exchange. HFor the relative levels according to 2.2.4.1.1, the resulting psophometric power and the total noise levels for the input connection are: HLiĠ=0+1.0+2.0dBr HPTNi =451410377pWOp HLTNi =63.563.964.2dBmOp Note The calculation above is intended to account for the worst case. No band limiting effect of the encoder on the noise was taken into account. 3.3.2.2 Unweighted noise HThis noise will be more dependent on the noise on the power supply and on the rejection ratio. Note The need for and value of this parameter are both under study. Recommendation G.123, 3 must also be considered. 3.3.2.3 Impulsive noise HIt will be necessary to place limits on impulsive noise arising from sources within the exchange; these limits are under study. Note 1 The sources of impulsive noise are often associated with signalling functions (or in some cases the power supply and the ringing voltage) and may produce either transverse or longitudinal voltages at Z interfaces. Note 2 The disturbances to be considered are those to speech or modem data at audio frequencies, and also those causing bit errors on parallel digital subscriber lines carried in the same cable. This latter case, involving impulsive noise with high frequency content, is not presently covered by the measurement procedure of Recommendation Q.45bis. 3.3.3HValues of total distortion HThe total distortion including quantizing distortion on half connections with Z interfaces is measured in accordance with 3.1.5. HThe signaltototal distortion ratio required for a half connection may be w.  approximated by the formula: where HS:  resulting signaltototal distortion ratio for input or output K#LNTN%Oconnections in digital local exchanges; HLs:  signal level of the measuring signal in dBmO; HLr:  for input connections, input relative level Li in dBr H for output connections, output relative level Lo in dBr; HHX HHS/N: signaltototal distortion ratio for PCM translating equipment in Recommendation G.714; HLN:  weighted noise caused by analogue functions according to Recommendation G.123, Annex A for local exchanges, i.e. 67dBmp at the Z interface. HOne resulting template for the signaltototal distortion ratio of input and output connections in a local exchange is shown in Figures15a and 15b/Q.552 as an example. HThe values of Figures 15a and 15b/Q.552 include the limits for the coding process given in Figure5/G.714 and the allowance for the noise contributed via signalling circuits from the exchange power supply and other analogue sources, which is limited to 67dBmp for a Z interface (with feeding) by Recommendation G.123, Annex A. As an example, the mean relative levels according to 2.2.4.1 are assumed to be Li = 0dBr and Lo = 7dBr. Note For an input connection the calculation above is assumed to be the worst case. No band limiting effect of the encoder on the noise was taken into account. LFIGURE 15a/Q.552 T FInput connection; Li = 0 dBră T T T T T T T T T T T T T T T T T T T T T T T T T T T LFIGURE 15b/Q.552 T EOutput connection; Lo = 7 dBră T <Limits for signaltototal distortion ratio as aă <function of input level including analogue noiseă QAnnex A T H(to Recommendation Q.552) T 9Example of a longitudinal interference coupling networkă T MTABLE A1/Q.552 T MComponents listă HThe component should be chosen with small absolute tolerances (at least resistors and capacitors with 1% and the inductance with less than 5%) and matched to pairs where relevant to achieve a longitudinal conversion loss better than 60 dB at 1000 Hz. For this LCL measurement a terminating resistance of 600 Ohms symetrically applied to each port should be used.