1、BRITISH STANDARD Methods of Measurement for radio equipment used in satellite earth stations Part 3. Methods of measurement on combinations of sub-systems Section 3.4 Measurements for frequency division multiplex (f.d.m.) transmission The European Standard EN 6051034 : 1994 has the status of a Briti
2、sh Standard ICs 33.060.30 NO COPYING WITHOUT ES1 PERMISSION EXCEPT AS PERMITTED BY COPYRIGHT LAW 5s EN !996 t992 10510-3-4 : EC 510-3-4 : CENELEC EN*h0*53O-3- 4 94 3404583 0375983 hT8 = Amd. No. The following BI references relate to the work on this standard: Committee reference EPU12/5 Draft announ
3、ced in BSZ Naos Up from these results the intermodulation noise may be obtained. The noise performance may be expressed as a noise power ratio (n.p.r.), a signal-to-noise ratio, in units of noise power or noise power level referred to the system zero relative level point. The units used may be picow
4、atts, decibels above 1 pW or decibels below 1 mW, and they may be specified as a weighted or unweighted psophometric value. CENELEC EN*bO*5LO-3- 4 94 3404583 0375989 Olb Page 4 EN 60610-3-4 : 1994 Noise power ratio is defined as the ratio of the noise power in a measuring channel when the baseband i
5、s fully loaded with the white noise load, to the power in that channel either with all the baseband loaded .except the measuring channel (.e. total noise) or with all the baseband unloaded (.e. basic noise); n.p.r. is always expressed as a positive number of decibels. Signal-to-noise ratio is define
6、d as the ratio of the power of the standard test tone (O dBm0) to the noise power, in a specified bandwidth within the noise-measuring channel, both being referred to the same point in the circuit. Signal-to-noise ratio may be measured weighted or unweighted and is expressed as a positive number in
7、decibels. Conversion between commonly encountered noise-loading measurement units may be made by reference to appendix A. 2.1 .l Conventional load The conventional loading level, which is defined by the CCITT (reference 1, see clause 6) and recommended by the CCIR (reference 2, see clause 6), is sho
8、wn in table 1 for some typical channel capacities. For other channel capacities the mean power level i, of the conventional load may be calculated from the following expressions: L, = -15 + 10 log, N dBmO for N2 240 fc = -1 + 4 log, N dBmO for 12 s N 240 where N is the system channel capacity. Notes
9、 1 These levels simulate the mean power of speech plus signalling currents. etc., transmitted over the system during the busy hour. Where a significant proportion of the baseband is used for v.f. telegraphy or data transmission, these expressions do not apply. 2 Equations 2-1 and 2-2 give a good app
10、roximation to actual signals when N 2 60. For smaller channel capacities, however, tests with white noise are less realistic owing to the differing nature of actual signals and test signals. CENELEC EN*bO*510-3- Y 94 340Y583 0175990 838 Page 6 EN 60510-3-4 : 1994 Table 1 - Level of the conventional
11、load Number of telephone channels 24 36 60 72 96 132 192 252 312 372 432 492 552 61 2 792 972 1 092 i a72 Level of the conventional load (dBmO) +4,5 +5.2 +6,1 +6.4 +6.9 +7.5 +9,1 +9,9 +10,7 +11,4 +11,9 +12,4 +12,9 +14.0 +14.9 +15.4 +17,7 +a, i A conventionally loaded system is one which is loaded at
12、 the conventional loading level with a uniform spectrum random noise signal which is band-limited to correspond with the total bandwidth of the f.d.m. signal. The test signal level, in most cases, is chosen to equal the conventional load. 2.1.2 Noise components The total noise measured within the ba
13、seband of a simulated satellite system includes the following three components: a) Residual noise which is independent of path attenuation and loading. This is normally referred to as path-loss-independent basic noise. b) Thermal noise which varies with path attenuation. This is normally referred to
14、 as pat h-loss-dependent basic noise. c) Intermodulation noise which is dependent upon the baseband noise loading level. Basic noise a) + b) is measured without noise loading as described below in 2.3.4. Total noise a) + b) + c) is measured with noise loading as described below in 2.3.2 or 2.3.3. 2.
15、2 Measuring equipment 2.2.1 General considerations Equipment for the measurement of noise-loading performance is commercially available and known either as “white noise test sets“ or “noise-loading test sets“. A white noise test set comprises a noise generator and a noise receiver; a typical circuit
16、 arrangement is shown in figure 1. CENELEC EN*bO*510-3- 4 74 D 3404583 0375993 774 Page 6 EN 60610-3-4 : 1994 To ensure test equipment compatibility and to achieve good measurement accuracy, the relevant characteristics of white noise test sets are closely specified by both CCIR (reference 2, see cl
17、ause 6) and the CCITT (reference 3, see clause 6). Commercial white noise test sets are normally sufficiently accurate for measurements on simulated satellite systems without making allowance for test equipment errors. However, where the required accuracy of measurement is comparable with the intrin
18、sic accuracy of the test equipment, due allowance for measurement error should be made in the presentation of results. Measurement accuracy depends upon many factors, including the following: - generator and receiver attenuator and monitor accuracies; - number of band-stop filters inserted and the e
19、ffective bandwidths of the noise-measuring channels; - region of the loading curve at which the measurement is being made (.e. whether basic or intermodulation noise predominates); - order of distortion predominant in the system under test. These factors are discussed in references 3 and 4 (see 6) a
20、nd in the publications listed in the bibliography (see 7). 2.2.2 Noise generator 2.2.2.1 Output characteristics The r.m.s. voltage of the noise source, when measured in a bandwidth of about 2 kHr, shall not vary by more than i 0,s dB within the bandwidth corresponding to the baseband of the system u
21、nder test. The test signal should have a Gaussian amplitude distribution up to a peak-to-r.m.s. ratio of at least 12 dB. The density of the noise power at the generator output shall have a maximum value of not less than -40 dBm/kHz to enable loading levels up to at least 10 dB above the conventional
22、 loading level to be used. The transmit level should be adjustable continuously or in small steps (e.g. 0.1 dB), by means of an output attenuator, to the specified value. The attenuator may typically have a range in excess of 50 dB. 2.2.2.2 Band-limiting and band-stop filters High-pass and low-pass
23、filters are required to define the baseband frequency limits appropriate to the simulated system under test and a series of band-stop filters are required to determine the noise-measuring channels. A wide variety of filters is available with current white noise test sets so enabling tests to be carr
24、ied out for all commonly encountered telephone channel capacities. The recommended filter frequencies are listed in table 2 and detailed filter specifications are given in reference 2 (see clause 6). Capacity (channels) Effective cut-off frequencies of band-limiting filters EN 606103-4 : 1994 Freque
25、ncies of recommended Table 2 - Recommended filter frequencies (from reference 2, see clause 6) Limits of band occupied by telep hone channels (kHz) measuring channels (kHz1 High pass 12 24 36 48 60 72 96 1 32 192 252 312 432 612 792 972 1092 1332 1872 372 492 552 12- 60 12- 108 12- 156 12- 204 12- 2
26、52 12- 300 12- 408 12- 552 12- 804 12-1 052 12-1 300 12-1 796 12-2 540 12-3 284 12-4 O28 12-4 892 12-5 884 12-8 120 12-1 548 12-2 o44 12-2 292 Page 7 12 f0,5 12 fO.5 12 t0.5 12 f0.5 12 f0.5 12 f0.5 12 f0,5 12 f0.5 12 f0.5 12 fO,5 12 f0.5 12 i0,5 12 i 0,5 12 iO,5 12 iO,5 12 f0.5 12 kO.5 12 f0.5 12 i0
27、.5 12 IO,5 12 i0,5 I Low pass 60 i0.5 108 i 1.0 156 f 1.0 204 f 1,5 252 f2.0 300 f2,O 408 I3.0 552 f4.0 804 f6.0 1 052 I8.0 1 296 f8.0 1796f12 2600 i 20 3 284 i 25 4100f30 4892 i 40 5884f50 8 160 f 75 1548i10 2044 f 14 2292f 17 16 56 16 98 16 140 16 185 16 240 16 270 16 240 16 240 16 394 16 534 16 5
28、34 16 534 16 770 16 1002 16 1002 70 1002 70 1002 70 1002 16 534 16 534 16 770 394 534 770 1002 1 248 1002 1730 1730 2438 2438 3150 2438 3886 2438 4650 3150 4650 5340 3150 5340 7600 1002 1490 1248 1940 1730 2150 2.2.2.3 Band-stop filter insertion loss An indication of the wideband noise power output
29、from the noise generator is required and it is normal practice to provide a power monitor at the end of the filter chain (see figure 1). The pass-band attenuation of crystal band-stop filters usually varies as a function of frequency and equalizers are often introduced to compensate for this variati
30、on. When equalizers are used, the total pass-band insertion loss is of the order of several decibels. After inserting band-stop filters the output level should be restored to its initial value to compensate for this insertion loss. Modern noise generators are fitted with automatic level control whic
31、h provides the correction automatically. However, some generators, which monitor wideband power prior to the band-stop filters, require correction of the output level by reference to an insertion loss table given in the instrument handbook. NOTE - Restoring the output power modifies the power densit
32、y of the signal, but the band eliminated by the band-stop filter is generally sufficiently narrow to make this effect negligible. CENELEC EN*b0*510-3- 4 94 3404583 0375993 547 lU Page 8 EN 60610-3-4 : 1994 2.2.3 Noise receiver Two distinct kinds of noise receiver are in common use. The first kind is
33、 suitable for the measurement of noise power ratio and contains a single attenuator of adequate range, for example O to 80 dB, connected directly to the input terminals of the receiver (see 2.3.2). The second kind is suitable also for the measurement of noise power referred to the system zero relati
34、ve point, in units of pWOp or dBm0p. Two attenuators are normally incorporated: the first is calibrated in transmission level (dBr) and the second acts as a 10 dB step range attenuator to be read in conjunction with a meter (see 2.3.3). In either kind, the design of the amplifier, mixer and attenuat
35、or should be chosen to avoid saturation or excessive non-linear effects when white noise loading levels are applied at up to 10 dB above the conventional loading 1evel;and at relative levels of up to -15 dBr. This corresponds to receiver input levels of up to about +10 dBm for a 972 channel system.
36、The intrinsic noise of the receiver shall be below -125 dBmp in order to measure noise powers on high capacity systems at loading levels down to 10 dB below the conventional loading level. The effective bandwidth of the receiver shall be not less than 1,7 kHz. It shall not exceed about 2,5 kHz in or
37、der for it to be narrower than the 70 di3 bandwidth of the band-stop fi It ers . Band-pass filters are required with centre frequencies coincident with the noise generator band-stop filters. The selectivity of these filters shall be sufficient to prevent spurious response or overloading of the recei
38、ver amplif ier(s) or mixer(s). 2.2.4 Inherent intermodulation of the white noise test set With the noise generator connected directly to the noise receiver and with the generator output noise power level equal to the conventional load (see table l), the total noise appearing within any noise measuri
39、ng channel shall be equivalent to an n.p.r. of at least 67 dB. The corresponding value of noise power level is -859 dBrnOp for N = 240. 2.3 Methods of measurement 2.3.1 Input-noise level The noise generator is connected to the baseband input port of the system under test. A high-pass and a low-pass
40、filter are selected to limit the bandwidth of the noise to that of the system baseband. The level of the conventional load is calculated from equations 2-1 or 2-2, or taken from table 1. The noise power to be applied to the baseband input port K is found by adding the level of the conventional load
41、to the relative power level at K. For example, for a system capacity of 1872 channels, the baseband input relative level is -37 dBr so the noise generator output level (Lout) will be: Lout = -15 + 10 log, (1872) (dBmO) -37 dBr = -19.3 dBm (2-3) CENELEC EN*bO*530-3- 4 74 3404583 0375774 483 W Page 9
42、EN 60510-3-4 : 1994 2.3.2 Method for noise receivers indicating in units of noise power ratio The noise generator and noise receiver are connected to the baseband input and output ports of the system under test and a noise-measuring channel is selected. The system input noise level is set to the con
43、ventional level or other specified level with the generator band-stop filter switched out of circuit. The receiver attenuator is set to give a reference reading on the receiver meter. The appropriate band-stop filter is then switched into circuit and the generator levei restored, if necessary (see 2
44、.2.2.3). Receiver attenuation is reduced until the reference reading is again obtained. The n.p.r. is the difference between the two settings of the receiver attenuator. Conversion of n.p.r. to other noise-loading units is described in appendix A. 2.3.3 Method for noise receivers indicating in units
45、 of noise power or signal-to-noise ratio The noise generator and noise receiver are connected to the baseband input and output ports of the system under test. A noise measuring channel is selected and the appropriate band-stop filter is inserted. The system input noise level is set to the convention
46、al level or other specified level, and allowance for filter insertion loss should be made if necessary (see 2.2.2.3). The receiver transmission levei attenuator is set to a value appropriate to the relative level of the system baseband output port. The range attenuator is then operated to increase s
47、ensitivity until a reading is obtained on the receiver meter; this reading should, if possible, be within the top 10 dB of the meter scale. The sum of the range attenuator and meter readings gives directly the noise power in units referred to the system zero relative level point. 2.3.4 Basic noise T
48、o measure the basic noise, a reading is taken without noise loading. This measurement is facilitated on modern noise-generators by a noise ON/OFF switch which suppresses the noise output whilst maintaining the generator output impedance constant. Receivers calibrated in noise power or level provide
49、a direct measurement of basic noise. Receivers calibrated in n.p.r. require the measurement to be made relative to the reference level as described in 2.3.2. The n.p.r. obtained is either expressed as the basic n.p.r. or converted into units of noise power level as indicated in appendix A. CENELEC EN*b0*510-3- 4 74 m 3404583 0375995 3LT m Page 10 EN 60610-3-4 : 1994 2.3.5 Total noise as a function of noise loading level and received r.f. carrier level The total noise may be measured over a range of noise-loading levels (e.g. -10 d6 to + dB) relative to the conventional loading level i
