Digital FM demodulating apparatus.

0001The present invention relates to a digital FM demodulating apparatus for demodulating a frequency-­modulated (FM) signal to be used for a signal recording/­reproducing system such as a video cassette recorder (VCR), video disc player, and the like.

0002In recent years, for improving the home use VCRs to have higher performance, higher function, and more rationalized circuit configuration, attempts have been made to perform the signal processing for recording and reproducing by means of digital signal processing.

0003With respect to the FM demodulation of the reproduced FM signal of VCR, a digital FM demodulating apparatus is disclosed in the paper, "DIGITAL SIGNAL PROCESSING IN VIDEO TAPE RECORDERS", by Sonke Mehrgardt, IEEE Trans. on Consumer Electronics, Vol.CE-31, pp374, August 1985.

0004By the way, the FM signal reproduced from a magnetic tape has suppressed upper sideband signals and emphasized lower sideband signals. This would frequently cause the reproduced FM signal to be over-modulated. Therefore, the demodulated signal by the abovementioned digital FM demodulating apparatus is damaged at the over-modulated portions. As a result, the image produced by the demodu­lated signal inversely drops into a black level in the portion where the signal should rise up from a black level to a white level, thereby producing a degraded image.

0005Another problem is that the dynamic range of the demodulated signal becomes extremely large when, even if no over-modulation phenomenon occurs, the lower sideband signal of the reproduced FM signal is relatively large to the upper sideband signal. Therefore, the number of bits for expressing the signal must be large, involving a problem of enlargement in the succeeding circuit scale. On the other hand, if amplitude limitation is made by an ordinary amplitude limiter to reduce the dynamic range, the level of the direct current component of the signal would be changed.

0006An object of the present invention is to provide a novel digital FM demodulating apparatus which demodulates a FM signal without degradation even if the FM signal has suppressed upper sideband signals and emphasized lower sideband signals.

0007Another object of the present invention is to provide a novel digital FM demodulating apparatus which demodulates a FM signal without enlargement of the dynamic range of the demodulated signal.

0008These objects can be accomplished by a digital FM demodulating apparatus comprising: phase detection means for detecting a phase of an input FM signal; differentiation means for time differentiating an output signal from said phase detection means thereby to obtain a demodulated signal in proportion to an instantaneous frequency of said input FM signal; over-modulation detection means for detecting an occurrence of an over-modulation of said input FM signal according to an amplitude value of said demodulated signal from said differentiation means to generate a detection signal; and correction means responsive to the detection signal for correcting the demodulated signal from the differentiation means to obtain a demodulated output signal which is corrected of deterioration due to said over-modulation.

0009The above and other objects, features and advantages will be apparent from consideration of the following description given hereinbelow.

0010The present invention will be better understood from the detailed description given hereinbelow in reference to the appended drawings which are given by way of illustration only and are not limitative of the invention, wherein: <ul id="ul0001" list-style="none"><li>Fig. 1 is a chart for illustrating the over-modulation phenomenon;</li><li>Fig. 2 is a block diagram of a digital FM demodulating apparatus as an embodiment of the present invention;</li><li>Fig. 3 is a block diagram of an over-modulation detection circuit used in the embodiment of the present invention;</li><li>Fig. 4 is a block diagram of an example of a correction circuit used in the embodiment of the present invention;</li><li>Fig. 5 is a block diagram of another example of a correction circuit used in the embodiment of the present invention;</li></ul>

0011Prior to the description of a preferred embodiment of the present invention, description will be made on the over-modulation phenomenon in VCR.

0012The over-modulation phenomenon occurs from the result that, in the FM signal reproduced from the magnetic tape, the upper sideband signal is suppressed and the lower sideband signal is emphasized. In examining the relation between this over-modulation phenomenon and the instan­taneous frequency of the FM signal reproduced by VCR, for simplification it is assumed that in the reproduced FM signal the upper sideband signal is fully attenuated, and for the lower sideband signal, only the first lower sideband signal is considered. Then, the reproduced FM signal v(t) is assumed to be expressed by the following equation: v(t) = cos ω_c t - ℓ cos(ω_c - ω_m)t      (1) where, ω_c is a carrier angle frequency of the FM signal, and ω_m is an angle frequency of the demodulated signal. In equation (1), the first term, cos ω_c t , shows a carrier component wherein the amplitude is normalized to 1, and the second term, ℓ cos (ω_c - ω_m)t, shows a lower sideband component. The lower sideband component has an amplitude of ℓ times that of the carrier component.

0013The equation (1) can be further modified as follows:<maths id="math0001"><img file="EP0343892A1_D0001.tif" /></maths>

0014Now, the instantaneous angle frequency ω (=2π f) of v(t) represented by equation (2) is obtained by time differentiating the phase ϑ to become as follows:<maths id="math0002"><img file="EP0343892A1_D0002.tif" /></maths> The instantaneous angle frequency ω obtained by varying ω_m t from -π to +π using the equation (3) is shown in Fig. 1 with ℓ as a parameter. As apparent from these data, as ℓ approaches 1 from 0, the amount of the instantaneous angle frequency ω at ω_m t = 0 becomes larger. When ℓ exceeds 1, i.e., when the amplitude of the lower sideband signal exceeds the amplitude of the carrier component, an over-modulation phenomenon occurs, at which time the amount of the instantaneous angle frequency ω at ω_m t = 0 becomes smaller than the carrier angle frequency ω_c. Especially, when ℓ is in the range to satisfy the following equation, the instantaneous angle frequency ω at ω_m t = 0 becomes a negative value.<maths id="math0003"><img file="EP0343892A1_D0003.tif" /></maths>

0015From the above, it is seen that, by detecting that the instantaneous frequency is smaller than the predeter­mined value, the over-modulation phenomenon can be detected. Alternatively, in the reproduced signal of VCR, in order to satisfy the equation (4) in general, it may be so arranged as to detect that the instantaneous frequency is negative.

0016Next, explanation is given as to how the over-­modulation phenomenon when detected as above should be corrected. When, in equation (3), the values of t in ω_m t are averaged from -π to +π, the relation becomes as follows:<maths id="math0004"><img file="EP0343892A1_D0004.tif" /></maths>

0017According to equation (5), in the case of ℓ &lt; 1, i.e., when there is no over-modulation phenomenon, the average value over one cycle of the modulated wave is equal to ω_c Accordingly, even if a waveform distortion is occurred as in the case of ℓ = 0.9 in Fig. 1, by eliminating a higher order harmonic distortion component by band limiting, a distortion-free demodulated output is obtainable, without causing variation of the direct current component. However, in the case of ℓ &lt; 1, i.e., when an over-modulation phenomenon is involved, not only the value of the instantaneous angle frequency ω at ω_m t = 0 becomes smaller than the carrier angle frequency ω but also the average value over one cycle of the modulated wave is smaller by 2π than ω_c. Accordingly, a variation takes place in the direct current component in the demodulation output. To prevent this, it is seen that, when an over-modulation phenomenon has been detected, a value corresponding to 2π may be added to the one-time over-modulation phenomenon.

0018As described above, by detecting that the instantaneous frequency is smaller than a predetermined value, the over-modulation phenomenon can be detected. Further, when an over-modulation phenomenon has been detected, it can be corrected by adding a value corresponding to 2π to the demodulation output on one-time over modulation phenomenon.

0019Based on the abovementioned consideration on the over-­modulation phenomenon, description will be made hereinbelow on an embodiment of the FM demodulation system of the present invention with reference to the drawings.

0020Fig. 2 shows a block diagram of a preferred embodiment of the invention which is a digital FM demodulating apparatus.

0021In Fig. 2, a digitized FM signal is inputted to an input terminal 1. The FM signal may be, for example, reproduced from a recording medium such as a tape or a disk by an appropreate known reproducing means.

0022In a phase detection circuit 2, by means of a π/2 phase shifter circuit 10 which is implemented by a Hilbert transformer or the like, the input FM signal is converted into two signals 22 and 23 which are different in phase by π/2 from each other. One of the two signals 22, 23 is divided by the other in a divider 12 to obtain a signal 24. Now, assuming that the signal 22 represents sinϑ and the signal 23 cosϑ, the signal 24 becomes a signal to represent tanϑ . Then, the signal 24 is inputted to a read-only-memory (ROM) 13 having stored therein values of arc-tangent function as a table to give a signal 25 representing the phase ϑ as an output signal. The signal representing the phase ϑ detected by the phase detection circuit 2 as above is inputted to a differentiation circuit 3.

0023The differentiation circuit 3 comprises a delay circuit 14 for delaying the input signal 25 by one sampling period to obtain a delayed signal 26 and a subtraction circuit 15 for subtracting the signal 26 from the signal 25 to approximately perform time differentiation to obtain a signal 27. As a result, the signal 27 is a demodulated signal proportional to the differentiated amount of the phase of the inputted FM signal, i.e. to the instantaneously frequency of the inputted FM signal. That is, the phase detection circuit 2 and the differentiation circuit 3 constitute the conventional FM demodulator.

0024The demodulated signal 27 is inputted to an over-­modulation detection circuit 51 and an over-modulation correction circuit 52.

0025The construction of the over-modulation detection circuit 51 is shown in Fig. 3. A comparator 31 compares the value of the signal 27 at each sampling point with a predetermined reference value and produce a signal 55 which becomes a high level when the value of the signal 27 is smaller than the reference value (i.e., an over-­modulation is occurring) and otherwise a low level. Thus, the signal 55 becomes a pulse signal whose duration corresponds to a duration of one over-modulation phenomenon. Usually, one over-modulation phenomenon lasts for one to three sampling periods. As described before, an error of the demodulated signal caused by one over-­modulation phenomenon can be corrected by adding a value corresponding to 2π to the demodulated signal. So, an over-modulation detection signal may be produced once at any one sampling point within the duration of or near the pulse signal 55. A detection signal generating circuit 32 generates such a detection signal 53. That is, the detection signal 53 is a single pulse occurring at one sampling point within the duration of or near the pulse signal 55. The detection signal 53 may be easily generated by using the sampling clock and the leading or trailing edge of the pulse signal 53.

0026Referring again to Fig. 2, the detection signal 53 described above is led to the correction circuit 52.

0027A construction of the correction circuit 52 is shown in Fig. 4. A selecting circuit 71 selects one of a fixed signal 73 representing value 0 and a fixed signal 74 representing 2π according to the detection signal 53. If detection signal 53 indicates an occurrence of the over-modulation phenomenon, the selecting circuit 71 selects the signal 74. If the detection signal 53 does not indicate an occurrence of the over-modulation phenomenon, the selecting circuit 71 selects the signal 73. So, an addition circuit 72 adds the signal representing 2π to the signal 27 only when an over-modulation phenomenon occurs.

0028As a result, an output signal 54 of the addition circuit 72 becomes a demodulated signal free from the over-modulation phenomenon and is outputted from an output terminal 4 in Fig. 2.

0029The comparator 71 in Fig. 3 may be arranged to detect the over-modulation phenomenon by detecting that the signal 27 is negative. By this, the implementation can be made by a simpler circuit.

0030By the way, as in the case of ℓ = 0.9 in Fig. 1, when, even if no over-modulation phenomenon has been occured, the lower sideband signal of the reproduced FM signal is relatively large to the upper sideband signal, the dynamic range of the demodulated signal becomes extremely large. Therefore the number of bits is needed to be large for expressing the signal, involving a problem of enlargement in the succeeding circuit scale. On the other hand, if amplitude limitation is made by an ordinary amplitude limiter to reduce the dynamic range, the level of the direct current component of the signal is changed. Now, with reference to the drawings, description is made on a second example of the correction circuit 52 in Fig. 2 which is designed to solve this problem.

0031Fig. 5 is a block diagram showing the second construc­tion of the correction circuit 52 in Fig. 2. In Fig. 5, the same numbers are assigned to the same blocks and signals as those of the first construction shown in Fig. 4. Description will be made below on the points different from the correction circuit in Fig. 4.

0032In Fig. 5, a signal 56 from the addition circuit 72 is fed to one input of an addition circuit 61. An output signal of the addition circuit 61 is led to an amplitude limiter 62. The amplitude limiter 62 limits the amplitude of its input signal to a predetermined value to make an output signal 54. On the other hand, in a subtraction circuit 63, an amplitude removed with the amplitude limiter 62 is obtained by taking a difference between the input signal and the output signal of the amplitude limiter 62. The output signal of the subtraction circuit 63 is delayed by one sampling period by a delay circuit 64, and then fed back to the other input terminal of the addition circuit 61 to be added to the signal 56. By this process, because the component removed by the amplitude limiter 62 is returned again and added to the signal 56, the direct current component of the signal 54 is equal to that of the signal 56. Further, the signal 56 has a small dynamic range limited with the amplitude limiter 62, so that the number of bits necessary for expressing the demodulated signal may be small, thus making it possible to minimize the scale of the succeeding circuit.