JPH0610844B2 - Magnetic recording method - Google Patents

Description

The present invention relates to a magnetic recording system, and more particularly to a magnetic recording system for signals including high frequencies such as digital signals. The magnetic medium according to the present invention is a material for performing bias recording, and any of its applied types such as tapes, desks and cards may be used.

Generally, the magnetic recording method includes non-bias recording and bias recording, and the latter is often used because the latter can suppress the recording signal level lower than the former.

In general, analog signal recording has been performed in recording music signals.

FIG. 1 (a) is a diagram showing a saturation characteristic curve a of the frequency f of the medium (magnetic tape) and the recording level h, and (b) is the characteristic of each frequency f ′, f ″ ... Of the input x and the output y. Even if a certain amount of level or more is applied to the medium and head defined in the figure, the recording / reproducing level becomes higher than this value, indicating that recording / reproducing cannot be performed.

An excessive signal that enters the saturation region h ₁₁ not only does not allow signal recording but also operates in a direction to increase the demagnetization effect and further reduce the output and generate harmonic distortion due to the non-linear region recording.
This harmonic is known to generate so-called beat noise with the bias signal.

However, in the analog recording of a music signal, the level of the high frequency included in the music signal is relatively small as compared with other low and middle frequencies, so that even if the bead noise is mixed, it has not been a serious problem. Moreover, even if it occurs, it does not become an essential obstacle that the sound quality is deteriorated and the recording and reproduction are impossible.

In addition, in order to reduce the occurrence of this beat noise, Fig. 2 (a)
As shown in the figure, the frequency of the bias signal b is selected to be 5 times or more as high as the maximum frequency of the music signal (usually 20 kHz), and as shown in FIGS. The filters F ₁ and F ₂ that cut off high frequencies at high frequencies are provided to prevent unwanted input signals from being applied to the head.

In addition, the invention of the applicant, Japanese Patent Publication No. 16684/53, was proposed.

Due to such consideration, the beat noise included in the music signal has been designed to be suppressed to a lower limit of -60 to -70db, which is the lower limit value of the dynamic range possessed by the music signal.

However, in digital recording, the waveform to be recorded is basically a waveform containing a large amount of harmonics (rectangular wave). Therefore, even if the signal level does not reach the saturation region, this tendency of beat occurrence is large. . If higher density recording is used,
Recording is required near the maximum frequency of the saturation characteristic (f characteristic) defined by the head and magnetic tape, and recording is performed at a high level as close to the saturation region as possible in order to obtain a good S / N ratio. Harmonic distortion is likely to occur, thus further promoting beat generation. Figure 3 shows this state. S is signal, N is beat noise, tape speed is 4.75.
cm / sec.

In the third example shown in the figure, the signal S is reproduced at an S / N ratio of 30 dB or more compared to the background noise such as hiss of the tape. However, the harmonics contained in the signal S and the bias signal frequency are close to about 20 kHz. Since strong beat noise N is generated, the S / N ratio is only 10db. In reproducing the digital signal NRZ (non-return, zero signal), it is modulated by the modulating means M ₁ as shown in FIG. 4, this is recorded / reproduced by the head tape, and the signal component is extracted by the signal detecting circuit of the demodulating means M. , The original digital signal according to the extracted signal
The input signal such as NRZ must be demodulated. In the figure, H ₁ is a recording head, H is a reproducing head, N ₁ is a recording signal amplifying means, and N is a reproducing signal amplifying means. The S / N ratio required for this extraction is generally 30 to 20db. However, when the beat signal as shown in FIG. 3 is present in the reproduced signal, the S / N ratio becomes about 10 db, which makes extraction impossible. This does not mean a sound quality failure in the conventional analog recording, but means an essential failure in recording / reproduction that signal recording / reproduction is impossible.

Also in the case of a music signal or the like, when the modulation circuit on the recording side shown in FIG. 4 modulates the digital signal NRZ, the group code modulation (for example, 4-7 modulation) wave of the recording current waveform is shown in FIG. 4-in such a random signal
Since a waveform having a 7-modulation signal spectrum and containing a high range of harmonics is recorded, the factor of beat generation tends to increase as compared with analog recording. In addition, 4-shown as an example
The minimum period Tmin and the maximum period Tmax of 7 modulation are as follows.

As described above, in recording / reproducing a large number of modulated waveforms containing harmonics, beat noise reduction has been a major issue in digital signal bias recording.

It is an object of the present invention to reduce so-called beat noise that is generated in this bias recording due to non-linearity of a recording signal, a bias signal, a transmission path including a head and a magnetic tape.

As a result of research, the present inventor succeeded in remarkably reducing the beat disturbance by providing a modulating filter and a head drive means system, that is, a signal recording system, with a wedding filter having specific characteristics in the recording and reproduction having the above-mentioned harmonic. Then, the present invention has been completed.

Examples of the present invention will be described in detail below.

As a general method, increasing the frequency of the bias signal can reduce the beat noise, but the upper limit of the frequency is limited by the heads and circuits used, and in the embodiment, it is set to 1 MHz. There is a limit to the reduction.

Next, when demodulating through a modulated waveform of a signal containing harmonics as shown in FIG. 5 and a recording / reproducing transmission transmission path, it is ideal that the transmission path has a wide frequency band, but the Nyquist condition 10
When the concept of 0% roll-off is applied, it can be seen that the original waveform can be demodulated even in the highest frequency f _max band that is the reciprocal of the minimum period Tmin of the modulated waveform. That is, the minimum period Tm of the modulation waveform
It shows that a band of 120 kHz is required when the maximum frequency f _max corresponding to in is 60 kHz. (This is 70kHz
In other words, it means that it is difficult to restore in a transmission line having a band less than the frequency, and in the case of recording / reproducing, it is impossible to reproduce. ) In consideration of this, in the method of performing beat reduction by using the high-frequency cutoff filters F ₁ and F ₂ as shown in FIGS. 2B and 2C, the cutoff frequency is at least the minimum period Tmin. It suffices to set the filter so that a frequency twice as much as the corresponding frequency is passed.

Further, in the thought of the Nyquist condition, it is assumed that the group delay characteristic in the transmission line is constant. Therefore, it is desirable that the group delay characteristics of the filter do not change significantly within the band set under the Nyquist condition. However, when a high-order filter is used in consideration of the beat reduction effect, the group delay characteristic generally tends to deteriorate, and therefore, in order to carry out the reduction method with a general filter, it is more than double and as high as possible. It is necessary to set the cutoff characteristic of the filter to the frequency and suppress the group delay change in the band up to twice.

In conventional analog recording for music signals, even when there is beat noise at a low value equivalent to the background noise (his noise etc.) level, the beat noise in the noise is heard, so beat noise reduction is necessary to prevent this obstacle. Even though the noise level is set to 50db, it has been considered to be close to 70db which is inaudible.

On the other hand, in digital recording, the level at which the signal can be extracted is 20 to 30 dB, and in consideration of high density, background noise (his noise etc.) is allowed up to around this value for recording and reproduction. In the high-density recording, the high frequency part is particularly important among the transmission frequency characteristics, but the recording tends to be performed in the low S / N ratio region. Therefore, this background noise and an input signal above the possible level have an important meaning in digital recording, but a signal below this level has no essential influence on recording and reproduction. Therefore, even if the beat noise is equal to or lower than the background noise level (example 30db) or lower than the extractable level (example 28db), it is not demodulated and does not cause a sound quality failure like analog. Means

Therefore, the method of implementing the beat reducing means shown in FIG. 2 in the digital recording has the effect that the high frequency cutoff of the filter is performed so that the beat noise level becomes equal to or lower than the signal detectable level (eg -28db) of the demodulator. The characteristics should be set as close to the high frequency side as possible.

FIG. 6 shows a saturation curve c of a transmission frequency f characteristic of a transmission line composed of a magnetic head and a tape. In this case, the frequency of the bias signal b is 1 MHz, and the modulation method is 4-7 modulation (Fig. 5). An example of high-density recording is shown. The background noise level h ₁ indicates the sliding noise when the tape is run and the noise generated by the reproducing circuit and the head. In this case, the background noise level is -30db.
Is.

FIG. 7 shows the characteristic curve d of the input x and the output y of the apparatus when a 55 kHz signal is recorded for high density recording. In this case, the recording is performed at the high frequency end of f characteristic shown in FIG. The recording near the saturation point c ₁ in FIG. 6 corresponds to d ₁ in FIG. 7 in order to increase the S / N ratio as much as possible. In the present invention, the input / output levels at these points c ₁ and d ₁ are set to the reference level 0db.
I am trying. The one shown in FIG. 7 is a characteristic curve of a signal of 55 kHz, but the saturation characteristic is determined by the magnetic medium, head, bias amount, etc., and therefore shows different values for each frequency.

At the frequency of 1 / n (n = integer) of the bias frequency (1 MHz), beat noise is generated in the vicinity frequency such as 1/2 500 kHz or 1/4 250 kHz, but such frequency is high. Therefore, as shown in FIG. 6, it does not appear as a recording / reproducing output on this transmission line and is not recorded. That is, the above-mentioned frequency is in a band where the saturation characteristic curve c becomes lower than the noise level h ₁ and cannot be detected by the reproducing head. However, if these frequencies are input and recorded, distortion will occur in the head and medium, and harmonics will be generated. The harmonics and the bias generate a beat and appear as beat noise in the reproduction band. For example, when the frequency of the input signal is 510kHz and the bias frequency is 1MHz, the frequency of the double harmonic of the input signal is 510 × 2 = 1020kHz, and the beat noise of this and the bias frequency, that is, 20kH of the difference frequency is reproduced. Come on. FIG. 8 shows this state.

FIG. 8 shows the input x ₁ to the beat output y ₁ of the transmission line. Frequencies with a relatively large number of beat occurrences, that is, 1/2 and 1/4 of the bias frequency, 500 kHz and 250 kHz, are selected and their respective neighborhoods are selected. An input x ₁ of a sine wave having a frequency in the range of ± 10% is applied to a point of the circuit shown in FIG. 4 and an output y _{1 at} that point is measured. Graph e ₀ indicates 500 kHz and graph f ₀ indicates 250 kHz. . Since the recording only generates a distorted wave, its input / output characteristics are represented by the form of beat noise as an output.

When the detectable level of the demodulation means M is set to -28db, it is necessary to suppress the beat noise to at least -30db. Therefore, as shown in Fig. 8, the input noise in the vicinity of 500kHz is -25db or less, as shown by point e _1. It is understood that one condition is that the input noise near 250 kHz needs to be suppressed to -18 db or less as shown by the point f ₁ .

The waveform to be recorded is the modulated wave g.
Considering the spectrum of the modulated wave g shown in the figure, the above conditions become points e ₂ and f ₂ . That is, points e ₂ and f ₂
Shows the input signal level for suppressing the beat noise below 30db.

As shown in the ninth figure, 6 dB, 500 near the point f ₂ of 250 kHz
The 10db modulated wave is high near the point e ₂ at kHz, and therefore,
Beat noise appears within the range of possible levels, indicating that demodulation is impossible.

From this, the filter means is 6db, 50 at around 250kHz.
It can be seen that the level of the modulated wave should be lowered by 10db near 0kHz.

FIG. 10 shows the frequency characteristic of the weighting filter in one embodiment of the present invention. The characteristic curve k has a cutoff at 200 kHz in the fourth order Putters and is 850 kHz.
Attenuates linearly toward.

When the modulated wave g shown in FIG. 9 is input to the weighting filter, it is suppressed as shown by the curve 1 in FIG. On the other hand, points e ₁ and f ₁ in FIG. 8 are shown in FIG. 10 as e ₃ and f ₃ , and the levels of these points e ₃ and f ₃ are lower than that of the curve l. Therefore, if this weighting filter is used, the beat noise can be suppressed to 30db or less, which shows that the modulated wave can demodulate the signal without the influence of the beat noise. In addition,
With this setting, the component spectrum of the modulation waveform is automatically determined when the modulation method is determined, so that the filter cutoff characteristics can be easily derived in this way.

FIG. 10 shows the weighting filter according to the present invention with a simple structure of Putters, but various filters satisfying the conditions of the present invention can be applied.

FIGS. 11, 12, and 13 show another embodiment of the present invention.
Characteristic curve k ₁ of the weighting filter shown in FIG.
Has a steep attenuation characteristic up to around 300kHz
It is oct 6db.

Characteristic curve k ₂ of the weighting filter shown in FIG.
Shows a characteristic curve k ₁ similar to that shown in FIG. 11, which is further provided with a steep descending characteristic near 600 kHz, and is a combination of a plurality of high-frequency cutoff characteristics.

Characteristic curve k ₃ of the weighting filter shown in FIG.
Is a filter characteristic of cot 6db with a band-stop type filter characteristic having resonance at a frequency where a large amount of beat is generated.

Both of them are the modulated wave g shown in FIG. 9 and the points e ₁ and f ₁ shown in FIG.
When the noise of is input, the curves l ₁ , l ₂ , l ₃ and the point f ₄ ,
f ₅ , f ₆ , e ₄ , e ₅ , e ₆ and the curves l ₁ ,
Since l ₂ and l ₃ are lower than the points e ₄ , e ₅ , e ₆ , f ₄ , f ₅ , and f ₆ , the occurrence of beat noise in the high range of the modulated signal (multi-frequency of beat generation amount) is It is less than 30db, which shows that the signal can be demodulated without the influence of beat disturbance.

Although the embodiment has been described with 4-7 modulation, the present invention uses this method (4
-7 modulation), any modulation method can be used, and the filter may be provided in consideration of the frequency component of each modulation method. The same applies to the case of directly recording the NRZ signal such as a rectangular wave.

The frequency characteristics of the weighting filter, which is the subject matter of the present invention, are as shown in the fourteenth diagram.

1. According to Nyquist's theorem, the cut-off frequency f _c of the characteristic curve k is when the minimum period of the modulation signal obtained by the modulating means is Tmin. Is more than twice the frequency.

2. The input / output level when recording and reproducing the signal of the minimum period Tmin of the modulation signal is used as a reference level, and as shown in FIG. 15, it corresponds to the integer (n ₁ , n ₂ ... ≠ 0) of the frequency f _b of the bias signal b. In the input x vs. beat noise output y characteristics of the device near the frequency of 1, The output levels Y ₁ , Y ₁ of the input signals of ... Represent the value (Y ₁ , Y ₂ = Y ₀ ) equal to the noise margin level Y ₀ of the demodulating means, and the input level is Ln (L ₁ , L ₁₎ . ₂ ..), the high frequency component In of the output 1 when the modulation signal g given to the recording head is input has a level difference .delta. Below the level Ln with respect to the signal of the minimum period Tmin. It has such damping characteristics.

FIG. 16 shows a specific circuit of the weighting filter having the cutoff characteristic according to the present invention. A digital signal such as NRZ (non-return, zero) is given to the input signal terminal 1. Between the modulation means 2 for converting the input signal into a modulation signal, the driving means 3 for driving the recording head 5 and the bias supply means 6 for supplying a bias signal to the recording head, and between the modulation means 2 and the driving means 3. Further, the weighting filter means 7 for beat noise reduction shown in the present invention is provided. With such a configuration, the modulated wave signal is supplied to the driving means 3 with a weight so as to suppress the beat noise to 30 db or less, so that the digital signal is reproduced without any trouble during demodulation.

FIG. 17 shows an example in which the filter means is provided between the driving means 3 and the head 5, and similar effects can be obtained.

FIG. 18 shows the drive means 3 having a filter characteristic by using, for example, a feed back loop.

The filter means 7 is composed of a passive filter or an active filter including a capacitor, a resistor, a coil and the like.

[Brief description of drawings]

FIG. 1 (a) is a graph showing a saturation characteristic curve of frequency and recording level of a general magnetic tape, (b) is a graph showing the relationship between the same input and output at each frequency, and FIG. 2 (a) is a conventional graph. A graph showing frequency characteristics of the filter, (b) and (c) are circuit diagrams showing two examples of specific circuits of the filter, FIG. 3 is a graph showing frequency characteristics of the digital signal, and FIG. A block diagram showing an example of a magnetic medium digital recording / reproducing apparatus, FIG. 5 is a graph showing frequency characteristics of an NRZ digital signal, and FIG. 6 is a frequency of a transmission line composed of a magnetic head and a tape similar to FIG. 1 (a). Graph showing characteristic song, Fig. 7
Fig. 8 is a graph showing the input and output characteristics of a 55 kHz signal, Fig. 8 is a graph showing the relationship between the input and the beat output in the input signal in the vicinity of a frequency that is an integer fraction of the frequency of the bias signal, and Fig. 9 is a digital signal. Of the modulated wave spectrum of the
FIG. 10 is a graph showing frequency characteristics of a weighting filter according to the present invention, and FIGS. 11, 12, and 13 are graphs showing three examples of frequency characteristics of a weighting filter according to another embodiment of the present invention, and FIG. FIG. 15 is a graph for explaining the characteristics of the weighting filter according to the present invention. FIG. 15 is a graph showing the relationship between the input and the beat output in the input signal in the vicinity of the frequency which is an integer fraction of the frequency of the bias signal.
Figures 6, 17 and 18 show 3 of the specific circuit of the filter according to the present invention.
It is a block diagram which shows an example, respectively. 2, M ₁ ... Modulation means, 3 ... Recording head drive means, 6
...... AC bias supplying means, 5, H ₁ ...... Recording head,
H ...... reproducing head, N ...... reproduced signal amplifying means, M ...... demodulation means, f c _...... cut-off frequency, f b _...... bias frequency, x ...... input level, y ...... beat noise power level, Y ₀ ...... noise margin level, Y _1, Y ₂ ...... noise margin level value, g ...... modulated signal, l n _...... high-frequency component, [delta] ......
Level difference, 7 ... Weighting filter.

Claims (1)

[Claims] 1. A magnetic medium digital recording / reproducing apparatus having a modulating means, a recording head driving means, an AC bias supplying means, a recording head, a reproducing head, a reproducing signal amplifying means and a demodulating means, the cut-off frequency being obtained by the modulating means. When the minimum period of the modulated signal is Tmin Of a frequency equal to or higher than twice the frequency of the modulated signal and the input / output level when recording / reproducing a signal having the minimum period Tmin of the modulation signal is used as a reference level, In the input level vs. beat noise output level characteristic, when the input level indicating the value at which the output level is equal to the noise margin level of the demodulating means is Ln, the high frequency component of the modulation signal given to the recording head has the minimum period Tmin. A weighting filter having the cutoff frequency and the attenuation characteristic is provided in the signal system path between the modulation means and the recording head so as to suppress the level of the signal below the input level Ln in the vicinity of the frequency. Magnetic recording method.