BS 5049:1987 download.British Standard Methods for Measurement of radio interference ch a ra cte risti cs of overhead power lines and highvoltage equipment.
BS 5049 gives methods for measurement of radio noise from overhead power lines and equipment for operation at I kV and above which may cause interference to radio reception, excluding the fields from power line carrier signals.
The frequency range covered is 0.15 MHz to 300 MHz.
Radio noise limits and measurement voltages are not specified. These are specified in the appropriate apparatus standard or agreed between the purchaser and the manufacturer.
2. MeasurIng Instruments
2.1 Response of a siandard C.1.S.P.R. measuring set to a.c. generated corona noise
C.1.S.P.R. Publication 16 specifics the response characteristic of a measuring set to periodically repeated pulses, according to their repetition frequency, for a number of measuring sets of differing frequency range and bandwidth including the range 0. 15 MHz to 30MHz and a bandwidth of 9kHz.
Figure I indicates the form these pulses take as they progress through the various stages of the measunrig set. However, in the special case of corona pulses generated by high-voltage a.c. power systems, the individual pulses are not equally spaced throughout a cycle but occur in closely packed groups or bursts around the peaks of the voltage waveform. A burst has a duration not exceeding 2 ms to 3 ms and this is followed by a quiescent no-corona period.
Owing to its inherent time constants, a C.l.S.P.R. measuring set is unable to respond to individual pulses within a burst, which is seen as a single pulse whose amplitude is discussed below.
Hence, the pulse repetition frequency, in the meaning of the C.1.S.P.R. definition, is constant a 2f (where! is the power system frequency) for single phase and 6f for three-phase single or multi-circuit systems, provided that the individual circuits are part of the same system.
Figure 2 indicates the usual case wherc individual corona pulses generated around the positive peaks of the voltage waveform are much greater in amplitude than those generated around the negative peaks. Hence in a three-phase power line there are three bursts of higher amplitude and three bursts of lower amplitude noise during each period of 1/f.
To eiwmine the response of the CJ.S. P.R. measuring set to a given burst of pulses, it should be borne in mind that each individual pulse becomes, at the output of the amplifier of Figure 1 of pass-band f, a damped oscillation whose duration can be taken as approximately 2/B, or 0.22 ms for 9kHz. When there is a large number of pulses distributed at random within a burst, the resulting oscillations will overlap randomly and the overall quasi-peak signal will be approximately equal to the quadratic sum of the individual quasi-peak values. This statement, which is difficult to prove mathematically, has been well proven by experience and justifies the usc, in quasi-peak detection, of the quadratic summation law which would moreover be rigorous if the noise levels were expressed in r.m.s. values.
2.2 Other measuring instruments
Measuring instruments differing from standard C.1.S.P.R. instruments are referred to in Appendix A although measuring apparatus having detectors other than quasi-peak are referred to in C.L.S.P.R. Publication 16.
3. C.I.S.P.R site measurements — 0.15 MHz to 30 MHz range
3.1 Measurement frequency
The reference measurement frequency is 0.5 MHz. It is recommended that measurements are made at a frequency of 0.5 MHz ± 10% but other frequencies, for example 1 MHz, may be used. The frequency of 0.5 MHz (or 1 Ml{z) is preferred because, usually, the level of radio noise at this part of the spectrum is representative of the higher levels and also because 0.5 MHz lies between the low and medium frequency broadcast bands.
Because of the possibility of error due to the presence of standing waves, it is inadvisable to rely on the measured value of the radio noise field at a single frequency but to draw a mean curve through the results of a number of readings throughout the noise spectrum. Measurements should be made at, or near, the following frequencies: 0.15,0.25, 0.5, 1.0, 1.5, 3.0,6.0, 10, 15 and 30 MHz although, clearly, frequencies at which interference to the wanted noise is received, should be avoided.
Interference may also result from broadcast stations which may be overcome by selecting a measurement frequency, within the specific toicrance, which is clear Of interference. The use of a resonant circuit L1C1, which is correctly tuned, as the rejection filter F, can often be most effective in reducing background noise,
4.12 Calibration of the test circuit
The test circuit shown in Figure 5, together with the circuit shown in either Figure 6 or Figure 7 shall be calibrated to obtain the value of the correction Ictor that shall be applied to the measuring set readings. This factor is the sum of the circuit attentuation and the resistance network flictor, both expressed in decibels (dB). Such calibration is required where the test assembly is being used for the first time, or has been re-arranged, or where the test object has been changed to one of a significantly different capacitance. The power supply to the high-voltage transformer should be disconnected during calibration,
a) Circuit attenuation A
Before starting the calibration, the rejection filter F shall be tuned, if applicable, as described in Item c) of Sub-clause 4.7, to the particular measurement frequency. A signal generator with an output impedance of at least 20 O00Ll shall then be connected in parallel with the test object, the test circuit being complete, as shown in Figure 5 together with the circuit shown in either Figure 6 or Figure 7. (Such a generator is easily arranged by connecting a 20 OO0 resistor in series with the output of a standard signal generator.) The generator shall be set to deliver a sine wave output of I V, at the measurement frequency, which will inject a current of about 50 iiA into the test circuit. This current will ensure that, with a C.I.S.P.R. measuring set, its reading will be well in excess of the usual background noise level. This reading, in decibels, of the measuring set shall be noted.
With the settings of the generator unchanged, the test object shall be disconnected from the high-voltage part of the test circuit and connected as shown in Figure 9. The new reading, in decibels, of the measuring set shall also be noted and the difference between the two readings is the circuit attenuation A.
Notes 1. — To avoid removing R1 and R2 from the test circuit during the calibration procedure, other high-stability. non-inductive resistors of the same value may be used.
2. – In Figure 9 the test object may be replaced by an equivalent capacaance, Ii this is known.