GB2029660A - Improvements in or relating to phase demodulation systems - Google Patents

Abstract

A phase demodulation system includes a filter which attenuates signals below a threshold frequency and having a transfer characteristic of -6 dB per octave above the threshold. This enables phase modulation to be measured by a frequency modulation measurement circuit and the specification describes a number of filters which enable a required characteristic <IMAGE> to be provided. <IMAGE>

Description

SPECIFICATION Improvements in or relating to phase demodulation systems This invention relates to phase demodulation systems. It is possible to derive modulation from a phase-modulated signal by means of a frequency discriminator, and to obtain a direct indication of the phase modulation by passing the derived modulation through a filter having a response which is inversely proportional to frequency so that it exhibits a transfer characteristic having a slope of -20 dB per decade with increasing frequency. The nature of the transfer characteristic emphasies low frequency components, but usually the wanted phase modulation is present only in frequencies above a minimum threshold frequency value.Significant response to frequencies below the threshold frequency value can give rise to excessive low frequency noise, which can cause serious difficulties of particular significance in test instruments which measure phase modulation characteristics over a broad band, and which seek to make accurate measurements of low level phase modulation.

The present invention seeks to provide a phased modulation system in which this difficulty is reduced.

According to this invention a phase demodulation system includes a filter for receiving modulation recovered from a phase modulated signal the filter having an amplitude-frequency response of the form

the filter being arranged to derive from the applied phase modulation a corresponding frequency modulation from the amplitude of which the applied phase modulation can be directly obtained.

The filter may consist of a plurality of stages, one stage being a low pass filter, and other stages being active high pass filters. In this case, preferably n is as large as it conveniently possible. A value of n = 5 is satisfactory in some circumstances. Alternatively the filter may consist of a single active stage in which n = 2.

The invention is further described by way of example with reference to the accompanying drawings in which, Figure 1 illustrates a number of different amplitude-frequency response characteristics obtainable with different filters, Figure 2 illustrates one form of a filter which can be used in a phase demodulation system in accordance with the present invention, Figure 3 illustrates another form of filter in which component values can be selected to give a performance which is the same as for Figure 2, Figure 4 shows an alternative filter and Figure 5 shows a filter which is modified slightly with respect to that shown in Figure 4.

Phase modulation can be measured by means of circuitry designed to make frequency modulation measurements provided that the frequency modulation measurement circuits are preceded by a filter circuit having a transfer characteristic arranged such that its output voltage is inversely proportional to the frequency of the input voltage so that a slope of - 6 dB per octave (i.e. - 20 dB per decade) with increasing frequency is obtaind. Requirements for phase modulation measurements are often concerned with the voice frequency band, a typical band of which extends from about 300 Hz to 3.4 kHz and in this case a test modulation frequency of about 1 kHz is satisfactory.A very simple CR circuit arranged as a low pass filter can be used, and the transfer characteristic of a circuit of this kind is shown in curve 1 of Figure 1, and it can be seen that above about 300 Hz the curve closely approaches the required - 6 dB per octave line. At 300 Hz the curve is arranged to give a 1% error. However, for frequencies below 300 Hz the response rises by a factor of 7 (17 dB) 50 that the presence of low frequency noise is likely to degradethedemodulation system and result in poor overall performance.

The addition of a simple CR high pass section reduces the low frequency response somewhat and one example of a transfer characteristic obtained by means of such a combination is illustrated in curve 2 of Figure 1. Curve 2 shows the response obtained when the time constants of the low pass and the high pass filter sections are equal and the values are chosen to give 1% error at 300 Hz. At 300 Hz the input signal is attenuated, by a factor of 10 (20 dB) and the response represented by curve 2 below 300 Hz is still undesirably high. Although the rejection of low frequencies has been improved this has been at the expense of reduced output voltage at 300 Hz and above.

The present invention provides filter circuits having improved characteristics, which are of-the form represented by curves 3 and 4. Filters which are cable of producing the characteristic shown in curve 4 are illustrated in Figures 2 and 3.

Referring to Figure 2, a low pass filter stage 1 consisting of resistor 2, capacitor 3 and amplifier 4 is followed by a number of high pass filter stages 5 and 6. Although only two additional stages 5 and 6 are shown more stages may be provided as required to give a better approximation to the ideal slope of- 6 dB per octave above a given threshold frequency value, in this case 300 Hz. Each stage 5 or 6 consists of a non-inverting amplifier 7 preceded by a series capacitor 8 and a shunt resistor 9 and having a feedback path coupled to the capacitor 8 by means of a resistor 10. Each stage 5 and 6 is coupled to the previous stage by means of an input capacitor 11. The transfer characteristic represented by curve 4 is a special case of the general equation

and is of the form

in which n = 5, n being the number of reactive components in the filter.

The component values of the circuit shown in Figure 2 are chosen to give a 1% amplitude error at 300 Hz and to give a transfer characteristic in accordance with equation (B). Curve 4 shows that the low frequency response is much reduced and the peak in the response curve is only 1.16 times (1.3 dB), the 300 Hz response. Attenuation at 300 Hz is 0.67 (3.5 dB) and n can be increased at the cost of greater circuit complexity, but with the benefit of increased steepness of the characteristic below 300 Hz and reduced attenuation at 300 Hz.

For the ideal -20 dB per decade characteristic, A ~ A (f)O) f/fo

For an amplitude error of 1% at frequency f,

For an amplitude error of 1% at 300 Hz, 300 1.477 The relationship between component values may be expressed by the following five equations

By using the above value of fro in these equations values of C1 R1, C2 R2, C4 R4, R3/R2 and Rs/R4 are obtained, and by choosing suitable values for C1, C2 and C4, corresponding values of R1, R2, R3, R4 and R5 are obtained.

Although in the circuit of Figure 2 operational amplifiers with 100% negative feedback are assumed, giving unity gain, it is possible to design the circuit with amplifier gains greater than unity, athough design formulae and component value would differ. The three sections 1,5 and 6 may be connected in any sequence, and if required other parts of the phase demodulation system can be inserted between them.

Figure 3 shows another form of filter circuit in which component values can be selected to give a transfer characteristic which is identical to that of Figure 2, but which uses one fewer operational amplifier, and which may be more convenient for some applications. In Figure 3 the low pass filter stage is integrally incorporated in the second active high pass filter stage. Two operational amplifiers 30 and 3 are provided.

Amplifier 30 has tw6 serially connected capacitors 32 and 33 connected to its input, and a feedback having a resistor 34 is connected from its output to the junction point between capacitors 32 and 33. The input to amplifier 30 is connected to ground via a resistor 35. The output of the amplifier 30 is connected to the input of amplifier 31 via a serial path comprising a capacitor 36, a resistor 37, a resistor 38 and a further capacitor 39. The output of the amplifier 31 is connected via a feedback resistor 40 to the junction point between resistors 37 and 38, and the input of the amplifier 31 is connected to ground via a resistor 41 and a capacitor 42 connected in shunt.

It will be seen that this circuit is generally similar to that of Figure 2. in Figure 3 the capacitor 42, and resistors 37 and 38 provide the low pass part of the filter characteristic, and these are incorporated in the second high pass filter section. The operational amplifiers 30 and 31 are non-inverting, and voltage gains of unity and suitably selected component values provide the characteristic shown in curve 4 of Figure 1. In Figure 3, n = 5 but the value of n can be increased with consequent improvement in performance by the addition of further sections similar to the one shown on the left of the broken line 43. Such sections are connected in cascade with the sections shown in Figure 3 and will have the same configuration as the section on the left of line 43. But the components will have different values.

Figure 4 shows an alternative and rather simpler form of filter circuit, which results in the transfer characteristic shown as curve 3 in Figure 1.

Referring to the drawing, the filter consists of a differential amplifier 20 in which the inverting input is coupled via a capacitor 21 and a resistor 22 to an input terminal. The output of the amplifier 20 feeds the output terminals of the filter and is also coupled via a further capacitor 23 and a resistor 24 to the input of the amplifier as shown. The transfer characteristic of such a filter is given in general terms by the equation

For the special case in which Q = +gT this characteristic reduces to

and it will be seen that this form is a special case of equation A, for which n = 2.

The filter shown in Figure 4may be modified slightly if desired in the form shown in Figure Sand it will be noticed that the only difference is that capacitor and resistor positions have been interchanged, but the values can be chosen so as to give a very similar overall transfer characteristic. In both Figures 4 and 5 the gain of the amplifiers is assumed to be very large in relation to the maximum value of the transfer equation.

For the ideal -20 dB per decade characteristic, A(f) ~ 4 A(fo) f/fo

For Figure 3,

For Figure 4,

By substituting into equation (C) a frequency f, the gain A(f) required at that frequency and the value of fro calculated above, the value of A(fo) may be found. Thus in the equations (D) and (E) only the values of C1, CPI R1 and R2 are unknown. If the value of one of these components is assumed, the other three may be calculated. As large a value as possible would normally be chosen for the resistors so that the capacitors are as small and inexpensive as possible. An upper limit to one or both resistor values is usually set by the bias and offset currents of the operational amplifier in use.Where an operational amplifier having low values of input current is used, a lower limit to the value of the capacitors is usually set by stray capacitance, so that the choice of one component value is not completely arbitrary.

The filters of Figures 4 and 5 introduce an inversion of the signal, and for some applications this may be undesired.

In the phase demodulation system the filter is preceded by a demodulation circuit, which produces from a received signal a frequency modulation signal. However, the phase modulation present in the frequency modulated signal cannot directly be obtained at this stage, since the amplitude of the phase modulation is dependent on the carrier frequency value. Once the signal has been passed through a filter having the required transfer characteristic with a slope of - 6 dB per octave, the phase modulation can be derived directly and unambiguously from the frequency modulation signal.

Claims (5)

1. A demodulation system including.a filter for receiving modulation recovered from a phase modulated signal, the filter having an amplitude4requency response of the form

the filter being arranged to derive from the applied phase modulation a corresponding frequency modulation from the amplitude of which the applied phase modulation can be directly obtained.

2. A demodulation system as claimed in Claim 1 and wherein the filter consists of a plurality of stages, one stage being a low pass filter, and other stages being active high pass filters.

3. A demodulation system as claimed in Claim 2 and wherein n = 5.

4. A demodulation system as claimed in Claim 2 and wherein the filter consists of a single active stage in which n = 2.

5. A demodulation system substantially as illustrated in and described with reference to Figure 2, 3, 4 or 5 of the accompanying drawings.