JPH01135209A - Noise correcting circuit - Google Patents

Abstract

PURPOSE:To execute the optimum noise correction regardless of the level and frequency of a signal by comparing the level of a coefficient signal outputted from a second holding means with a reference level and controlling a supply to the adding means of the coefficient signal in accordance with the comparing result. CONSTITUTION:A coefficient signal held to a capacitor C12 is supplied to an absolute value circuit 19 and the absolute value of the coefficient signal is obtained. The absolute value level and the reference voltage level of a constant voltage source 20 are compared with a level comparator 21 and based on the comparing result, a switching circuit 22 to lie a voltage-current converting circuit 17 and a switching circuit 18 is turn-on and turn-off controlled. At this time, when the absolute value level is lower than a reference voltage level, the circuit 22 is turned off and at the time of the opposite case, the circuit 22 is turned on. At the time of reproducing an ordinary voice signal when the absolute value level of the coefficient signal is higher than the reference voltage level, the circuit 22 is turned on, the output current from the circuit 17 is supplied to the charging and discharging of a capacitor C11 and a linear approximate system noise correction is executed.

Description

[Detailed Description of the Invention] [Objective of the Invention] (Industrial Application Field) This invention is an object of the present invention to prevent noise from being mixed in at regular intervals, such as switching noise when switching heads in a rotating head type video tape recorder. The present invention relates to a noise correction circuit used in a signal transmission system.

(Prior Art) As is well known, in home video tape recorders, in order to achieve high sound quality (high-fidelity), the audio signal is FM modulated and, like the video signal, it is transmitted through multiple rotations. Recording on a tape using a helical scan method using a head and reproducing the tape is carried out when fishing on a boat.

This type of audio recording and reproducing system has the advantage that it can achieve a wider band than a system with a fixed audio head, and is extremely advantageous in improving sound quality. On the other hand, during playback, multiple rotating heads are selectively brought into contact with and separated from the tape, and a switch is used to select the FM-modulated audio signal, so switching noise caused by switching heads at regular intervals is added to the audio signal. The problem of contamination arises.

Two means are currently being considered to reduce this type of switching noise. First, the first means is a sample hold method in which the audio signal immediately before the noise occurs is held, and this held audio signal is output during the noise generation period.

Further, the second means holds the audio signal immediately before the noise occurs, differentiates the audio signal immediately before the noise occurs to generate a coefficient signal for linear approximation, and combines the held audio signal and the coefficient signal. This is a linear approximation method that linearly approximates the audio signal waveform during the noise generation period based on the following.

In this case, since the switching noise mixed into the audio signal is generated when the heads are switched, the timing of its occurrence can be known in advance. For this reason, a correction pulse is generated that becomes an H level when the head is switched and becomes an L level in other situations, and based on this correction pulse, the hold timing of the audio signal, the extraction timing of the coefficient signal, etc. are controlled. .

The sample-hold method has the advantage of a simple circuit configuration, and the linear approximation method has the advantage of being able to perform highly accurate noise correction, so both methods are often used. There is.

FIG. 6 shows a conventional noise correction circuit using a linear approximation method. Reference numeral 11 denotes an input terminal to which an FM-modulated audio signal obtained by switching a plurality of rotary heads (not shown) is supplied. As described above, this FM-modulated audio signal is mixed with switching noise at regular intervals due to switch switching.

Here, the audio signal supplied to the input terminal 11 is supplied to the buffer circuit 13 via the switch circuit 12 and the capacitor C1l. The output of this buffer circuit 13 is supplied to an audio demodulation/reproduction system (not shown) via an output terminal 14.

Further, the audio signal supplied to the input terminal 11 is supplied to a differentiating circuit 15. The differentiating circuit 15 generates a coefficient signal for linear approximation by differentiating the input audio signal. The coefficient signal output from the differentiating circuit 15 is then transmitted to the switch circuit 16. Capacitor C12, voltage-current conversion circuit 17 and switch circuit 1
8 to the buffer circuit 13.

Here, the switch circuit 12. lft is controlled so that it is in an on state during an L level period of the correction pulse, which is outside the switching noise generation period, and is in an off state, during an H level period of the correction pulse, which is during a switching noise generation period. The switch circuit 18 is the switch circuit 1
Contrary to 2.18, it is controlled to be in the OFF state during the L level period of the correction pulse and to be in the ON state during the H level period of the correction pulse.

According to such a configuration, first, outside the switching noise generation period, the audio signal supplied to the input terminal 11 is transmitted to the switch circuit 12 . The signal is directly supplied to the audio demodulation/reproduction system via the capacitor C11° buffer circuit 13 and the output terminal 14, and is subjected to FM demodulation processing.

Furthermore, when the switching noise generation period begins and the correction pulse is inverted from the L level to the H level, the audio signal immediately before the noise generation is held in the capacitor C1l.

Furthermore, the coefficient signal output from the differentiating circuit 15 immediately before the noise occurs is held in the capacitor C12.

The coefficient signal held in the capacitor C12 is converted into a current corresponding to the level by the voltage-current conversion circuit 17, and then passed through the switch circuit 18 to the capacitor C1.
Charge and discharge 1. Then, the voltage across the capacitor C1l is firstly integrated with the coefficient signal immediately before the noise occurs, with the level of the previously held audio signal as the initial value, and a linearly approximated signal is generated here. Then, the signal is supplied to the buffer circuit 13.

Note that the sample-and-hold type noise correction circuit is
From the figure, differentiation circuit 15. Switch circuit 16. Condenser 6 = constructed by omitting the capacitor C12, the voltage-current conversion circuit 17, and the switch circuit 18.

Here, the voltage-current conversion circuit 17 is required to have a characteristic such that when the input voltage is "0", the output current is also "0", as shown by the solid line in FIG. In other words, when we consider the case of playing back an unrecorded tape, the playback signal will be mixed with switching noise at regular intervals as shown in Figure 8(a), and the playback signal will be mixed with switching noise as shown in Figure 8(a). When the linear approximation correction described above is performed during the H level period of the correction pulse shown in b), when the voltage-current conversion circuit 17 with ideal characteristics is used, the switching noise is completely eliminated as shown in Fig. A removed reproduced signal can be obtained.

Note that even with a sample-and-hold type noise correction circuit, when playing back an unrecorded tape, as shown in Figure 8 (C),
A reproduced signal with switching noise removed can be obtained.

However, in reality, the voltage-current conversion circuit 17 has an offset, so its characteristics vary as shown by dotted lines A and B in FIG. 7, and this offset causes linear approximation. This is a factor that causes correction errors. For example, if the characteristics of the voltage-current conversion circuit 17 vary as shown by the dotted line A in FIG.
As shown in Figure (d), triangular noise is included.

In addition, the operating points for small and large signals are shown in Figure 7, respectively.
, ■, and the output after correction is as shown in Figure 8 (e
) and (f), triangular noise is included. The smaller the amplitude of the signal is, the greater the ratio of the noise level to the signal becomes, so that such a correction error becomes more audibly perceived as unpleasant sound.

That is, the larger the amplitude of the signal or the higher the frequency of the signal, the more effective the linear approximation method becomes than the sample-and-hold method.

The smaller the amplitude of the signal or the lower the frequency of the signal, the less there is audible difference between the linear approximation method and the sample hold method, but rather due to the offset of the voltage-current conversion circuit 17 of the linear approximation method. Considering the fact that correction errors occur in , it can be said that the sample-and-hold method can reproduce sound with less noise.

(Problems to be Solved by the Invention) As described above, in conventional noise correction circuits, the linear approximation method is superior in reproducing normal audio signals;
When reproducing extremely weak audio (no signal) or extremely low frequency sounds, the sample-hold method can be said to be superior because the linear approximation method voltage-to-current conversion circuit has an offset.

Therefore, the present invention was made in consideration of the above circumstances, and an object thereof is to provide an extremely good noise correction circuit that can always perform noise correction using an optimal method regardless of the signal level or frequency. do.

[Structure of the Invention] (Means for Solving the Problems) In other words, the noise correction circuit according to the present invention has -9= first holding means that holds a signal immediately before the noise occurs during the period in which the noise occurs. , a differentiating means for differentiating a signal immediately before the noise occurs to generate a coefficient signal for linear approximation, a second holding means for holding the coefficient signal output from the differentiating means during the noise generation period; addition means for performing noise correction by adding the coefficient signal output from the second holding means to the output signal of the holding means; and comparison means for comparing the level of the coefficient signal output from the second holding means with a reference level. and a control means for supplying and cutting off the supply of the coefficient signal output from the second holding means to the adding means in accordance with the comparison result of the comparison means.

(Function) According to the above configuration, the level, frequency, etc. of the signal are determined by comparing the level of the coefficient signal output from the second holding means with the reference level, and the Since the coefficient signal output from the second holding means is controlled to be supplied to and cut off from the adding means, noise correction is performed using a linear approximation method during normal signal reproduction, and when reproducing weak signals or low frequency signals. It is possible to automatically switch to perform noise correction using the sample-and-hold method, and noise correction can always be performed using the optimal method regardless of the signal level or frequency.

(Example) Hereinafter, an example of the present invention will be described in detail with reference to the drawings. In FIG. 1, the same parts as in FIG. 6 are indicated by the same symbols, and only the different parts will be described here. That is, the coefficient signal held in the capacitor C12 is supplied to the absolute value circuit 19, and the absolute value of the level of the coefficient signal is determined.

The level comparison circuit 21 compares this absolute value level with the reference voltage level of the constant voltage source 20, and based on the comparison result, a switch circuit 22 interposed between the voltage-current conversion circuit 17 and the switch circuit 18 , turn on.

I'm trying to control it off.

In this case, the switch circuit 22 is set to be in an off state when the absolute value level is lower than the reference voltage level, and to be in an on state when the absolute value level is higher than the reference voltage level.

According to such a configuration, during normal audio signal reproduction when the absolute value level of the coefficient signal is higher than the reference voltage level, the switch circuit 22 is turned on, and the current output from the voltage-current conversion circuit 17 is is used for charging and discharging the capacitor C1l, and noise correction using a linear approximation method is performed.

When reproducing a weak audio signal or an extremely low-pitched audio signal, the switch circuit 22 is turned off, and the voltage-current conversion circuit 1 is turned off.
The current output from the capacitor C1l is not used to charge or discharge the capacitor C1l, and the signal held in the capacitor C1l immediately before the noise generation is supplied as is to the buffer circuit 13.
Sample-and-hold noise correction is performed.

In other words, the characteristics of the charging/discharging current to the capacitor C1l with respect to the terminal voltage (coefficient signal level) of the capacitor C12 are as shown in the ideal characteristic shown by the solid line in FIG. It is separated from the capacitor C1l and becomes a sample hold method, and a linear approximation method is used except for the coefficient signal level near rOJ.

Now, if we consider the case of playing back an unrecorded tape, the playback signal will be mixed with switching noise at regular intervals, as shown in Figure 3(a). In (b), during the H level period of the signal correction pulse, the above-described sample-hold method or linear approximation method correction is performed.

Here, the characteristics of the charging/discharging current with respect to the coefficient signal level have a maximum positive and negative offset, as shown by dotted lines C and D in FIG. 2, due to the offset of the voltage-current conversion circuit 17. For example, if the characteristics of the charge/discharge current with respect to the coefficient signal level vary as shown by the dotted line C in Figure 2, the operating point during no signal will be at the position indicated by ■ in the figure, and the sample-and-hold method correction will occur. The output after the correction is one in which switching noise has been completely removed, as shown in FIG. 3(C).

In addition, the operating point at the time of a small signal is the position shown by ■ in Figure 2, and sample-and-hold noise correction is performed, and the output after correction has a triangular shape as shown in Figure 3 (d). Noise will be included. Furthermore, the operating point at the time of a large signal is the position shown by ■ in Figure 2, and noise correction is performed using a linear approximation method, and the output after correction has a triangular shape as shown in Figure 3 (e). Noise will be included.

Therefore, according to the configuration of the above embodiment, the audio signal level and frequency height are determined from the magnitude of the coefficient signal level, noise correction is performed using a linear approximation method in the case of a normal audio signal, and weak In the case of audio signals or ultra-low frequency audio signals, noise correction is performed using a sample-hold method. Even if the characteristic of the charge/discharge current with respect to the coefficient signal level has a maximum offset of positive or negative, the signal will be Noise correction can always be performed using the optimal method regardless of the level or frequency.

In the above embodiment, since sample-and-hold noise correction is performed when the signal is small, the noise contained in the small signal waveform shown in FIG. The noise generated is louder. However, even if the noise included in the small signal waveform becomes larger, the effect of completely removing the switching noise when there is no signal is a much greater audible effect.

Note that when a large signal is present, noise correction is performed using the linear approximation method as in the conventional case. - Next, FIG. 4 shows another embodiment of the present invention, in which the voltage-current conversion circuit 17 is changed. Above capacitor C
The coefficient signal held at 12 is supplied to the input terminal 23 via an inverting buffer circuit (not shown). and,
The voltage level applied to this input terminal 23 is determined by the resistance R1
1 is converted into current.

In this case, the potential at point E in FIG. 4 is R12 (IO2 - IO2) + R13 (IO4 - IBI
) changes. This voltage change becomes a dead zone for voltage-current conversion.

The positive component of the current converted in this way is led to the output terminal 25 via a current mirror circuit made up of transistors Qll and Q12, and the negative component is led to the output terminal 25 through a current mirror circuit made up of transistors Qll and Q12. It is guided to the output terminal 25 via a current mirror circuit made up of Q14.

The conversion characteristics of the voltage-current conversion circuit shown in FIG. 4 are as shown in FIG. 5. The solid line in FIG. 5 shows the characteristics when there is no offset, and the dotted lines F and G in the same figure show the characteristics when there is a maximum positive and negative offset, respectively.

Here, the width of the dead zone (sample and hold period) when the input voltage is around "0" is approximately given by the above equation, but in terms of the circuit, transistor Q15. Q18 is switched on and off softly according to the input voltage (coefficient signal level), so the transition from the linear approximation method to the sample hold method is
The conversion is performed more softly than that shown in FIG. 2, and the conversion characteristics are gentler.

For example, the characteristics of the output current with respect to the coefficient signal level are
If there is a variation as shown by the dotted line F in the figure, the operating point when there is no signal will be at the position indicated by ■ in the figure, and sample-hold correction will be performed, and the output after correction will be as shown in Figure 3 (C). As a result, switching noise is completely eliminated.

In addition, the operating point at the time of a small signal is the position shown by ■ in Figure 5, and sample-and-hold noise correction is performed, and the output after correction has a triangular shape as shown in Figure 3 (d). Noise will be included. The operating point at the time of a large signal is the position shown by ■ in Figure 5, and noise correction using a linear approximation method is performed, and the output after correction has triangular noise as shown in Figure 3 (e). will be included.

Therefore, even if a voltage-current conversion circuit as shown in FIG. 4 is used, substantially the same effect as in the above embodiment can be obtained. In particular, in the case of small signals, since the voltage-current conversion characteristics are soft, the waveform becomes slightly more linearly approximated than shown in Figure 3(d), and is slightly more effective than the noise correction circuit shown in Figure 1. become a target.

It should be noted that the present invention is not limited to the above-described embodiments, and can be implemented with various modifications without departing from the gist thereof.

[Effects of the Invention] Therefore, as detailed above, according to the present invention, it is possible to provide an extremely good noise correction circuit that can always perform noise correction using an optimal method regardless of the signal level or frequency. Can be done.

[Brief explanation of the drawing]

FIG. 1 is a block circuit configuration diagram showing one embodiment of the noise correction circuit according to the present invention, FIGS. 2 and 3 are characteristic diagrams and waveform diagrams for explaining the operation of the same embodiment, respectively, and FIG.
5 and 5 are respectively a circuit configuration diagram showing another embodiment of the present invention and a waveform diagram for explaining its operation, FIG. 6 is a block circuit diagram showing a conventional noise correction circuit, and FIGS. FIG. 8 is a characteristic diagram and a waveform diagram, respectively, for explaining the operation of the conventional noise correction circuit. = 18-ll...input terminal, 12...switch circuit, 13.
...Buffer circuit, 14...Output terminal, 15...Differential circuit, 16...Switch circuit, 17...Voltage-current conversion circuit, 18...Switch circuit, 19...Absolute value circuit , 20... constant voltage source, 21... level comparison circuit, 22... switch circuit, 23... input terminal, 2
4... Constant voltage source, 25... Output terminal. Applicant's agent Patent attorney Takehiko Suzue Figure 4 Figure 5

Claims (1)

[Claims] In a noise correction circuit that is provided in a signal transmission system in which noise is mixed in at regular intervals and performs a correction process for the noise during the period in which the noise occurs, a signal immediately before the noise occurs is held during the period in which the noise occurs. a first holding means;
a differentiating means for differentiating the signal immediately before the noise occurs to generate a coefficient signal for linear approximation; a second holding means for holding the coefficient signal output from the differentiating means during the period in which the noise occurs; The output signal of the first holding means is applied to the second holding means.
an addition means for performing noise correction by adding the coefficient signal output from the second holding means; a comparison means for comparing the level of the coefficient signal output from the second holding means with a reference level; and a comparison result of the comparison means. A noise correction circuit comprising: control means for supplying and cutting off the supply of the coefficient signal output from the second holding means to the addition means in accordance with the above.