JP2004055017A - Adaptive equalizer and reproducing device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To always make excellent equalization by avoiding the process which uses defective coefficients. <P>SOLUTION: The signal taken out from a magnetic tape 1 with a reproducing head 2 and a reproducing amplifier 3 is supplied to an analog equalizer 5 through an automatic gain control circuit 4. The signal processed for analog equalization is supplied to an A/D converter 6, and the dispersed signal is supplied to a phase-locked loop 7 and also to a first and second adaptive equalizer circuits 8 and 9. And the output signal of the second adaptive equalizer circuit 9 is supplied to a decoder 10 of, for example, a Viterbi decoding algorithm, and the signal from this decoder 10 is supplied to, for example, an error correction code decoding circuit 11, and reconstructed into digital data simultaneously with error correction. This reconstructed digital data are converted into analog signals in a D/A converter 12 through an expander circuit (illustration is omitted) etc., and the analog signals, for example, such as video and audio signals are taken out from the output-terminals 13. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an adaptive equalizer and a reproducing apparatus suitable for use in, for example, reproducing a signal digitally recorded on a recording medium. More specifically, it is intended that the equalization of the reproduced signal be performed satisfactorily and that the error rate of the reproduced signal be prevented from being deteriorated.
[0002]
[Prior art]
For example, in reproducing a signal digitally recorded on a recording medium, a device as shown in FIG. 29 is conventionally used. That is, in FIG. 29, a signal extracted from a magnetic tape 61 through a reproducing head 62 and a reproducing amplifier 63 is supplied to an analog equalizer 65 through an automatic gain control circuit 64, and the analog equalized signal is converted into an A / D converter. 66. Then, the digital signal converted by the A / D converter 66 is supplied to a phase lock loop (hereinafter abbreviated as PLL) 67 to extract a data clock of the reproduction signal, and the extracted data clock is subjected to A / D conversion. The digital signal is supplied to the reproducing unit 66 and the reproduced signal is converted into a digital signal.
[0003]
Further, the reproduced signal digitally converted by the A / D converter 66 is supplied to an adaptive equalizing circuit 68, and the output signal of the adaptive equalizing circuit 68 is supplied to a decoder 69, for example, to decode by applying a Viterbi decoding algorithm. Is performed. In the Viterbi decoding algorithm, for example, the symbol sequence with the maximum path metric is estimated using the impulse response of each transmission path estimated with different weighting coefficients, and the symbol sequence with the maximum calculated path metric is estimated. Is output. A target value corresponding to the output symbol sequence is fed back to the adaptive equalization circuit 68.
[0004]
Here, the adaptive equalization circuit 68 has a basic structure as shown in FIG. 30, for example. That is, in FIG. 30, the reproduction signal supplied to the input terminal 80 is supplied to a plurality of unit delay means 81 to 84 connected in series. In FIG. 30, the unit delay means has four stages for the sake of simplicity. However, in an actual apparatus, ten or more stages are used. The signals at the input and output terminals of these unit delay means 81 to 84 are added by the adder 90 through the weighting means 85 to 89 for the weighting coefficients C1 to C5, respectively. As a result, a so-called FIR filter is formed, and the reproduced signal is equalized.
[0005]
Further, the ternary (+1, 0, -1) target value extracted from the above-described decoder 69 is supplied to the arithmetic circuit 91, and is subtracted from the output signal of the adder 90 to extract an error from the target value. It is. Then, the error and the reproduced signal supplied to the input terminal 80 are supplied to the arithmetic circuit 92, and the impulse response of the transmission path is estimated using, for example, a least-square (Least Mean Square) algorithm. At the same time, the weighting coefficients C1 # to C5 # set in the weighting means 85 to 89 described above so as to minimize the error are obtained.
[0006]
As a result, for example, a reproduced signal obtained by estimating an impulse response of a transmission path in magnetic recording and reproduction is equalized. In this case, the arithmetic circuit 92 takes into account the estimation of the impulse response of the transmission path by, for example, the LMS algorithm and the error from the target value obtained by, for example, the Viterbi decoding algorithm in the decoder 69, for equalization. By adaptively determining the weighting coefficient of, for example, in magnetic recording and reproduction of a digital signal, good reproduction signal equalization can be performed.
[0007]
In FIG. 29, a signal from the above-described decoder 69 is supplied to, for example, an error correction code (hereinafter, abbreviated as ECC) demodulation circuit 70, and digital data is reconstructed simultaneously with error correction. Further, the reconstructed digital data is supplied to a decompression circuit (not shown) and the like, and processing such as decompression and restoration of the compression performed at the time of recording is performed. Then, this signal is converted into an analog signal by the D / A converter 71 and is taken out to the output terminal 72. In this way, for example, digitally recorded video and audio signals are reproduced.
[0008]
[Problems to be solved by the invention]
However, in such an apparatus, for example, it is necessary to always determine the validity of the coefficient of the adaptive equalization circuit 68. That is, since the above-described device is provided with a feedback system, there is a high possibility that oscillation, divergence, and the like will occur, especially when trying to increase the loop gain to speed up the convergence of coefficients and the like. Therefore, in order to eliminate such a fear, it is necessary to always determine the validity of the coefficient of the adaptive equalization circuit 68 and to control the coefficient so as not to be defective.
[0009]
However, in the above-described apparatus, for example, when trying to evaluate the coefficient of the adaptive equalization circuit 68 based on the output of the adder 90, there is a possibility that the error rate of the reproduced signal is essentially deteriorated. This is because the output of the adder 90 of the conventional adaptive equalization circuit 68 is sequentially supplied to the subsequent decoder 69 and the ECC demodulation circuit 70. If an abnormality occurs in this output, the abnormality is directly The error rate is supplied to the ECC demodulation circuit 70 and adversely affects the error rate.
[0010]
That is, for example, in a configuration in which the coefficient of the adaptive equalization circuit 68 is evaluated based on the output of the adder 90, when it is determined that the coefficient becomes defective, the determination may be delayed, and There is a possibility that the error rate of the signal is deteriorated. To solve such a problem, methods such as lowering the loop gain or increasing the accuracy of the determination are conceivable. However, any of these methods delays the convergence of the above-described coefficients and the like.
[0011]
The present application has been made in view of such a point, and the problem to be solved is that the conventional apparatus evaluates and determines the validity of the coefficient of the adaptive equalization circuit based on the output. Therefore, the error rate of the reproduced signal may be deteriorated. On the other hand, methods such as increasing the accuracy of evaluation judgment or lowering the loop gain are also conceivable, but in any case, the convergence of the coefficients and the like is slow, and good adaptive processing cannot be performed. It is.
[0012]
[Means for Solving the Problems]
Therefore, in the present invention, a first adaptive equalizing circuit and a second adaptive equalizing circuit for performing equalization of a reproduced signal are provided, and a weighting coefficient calculated by the first adaptive equalizing circuit is provided. A pass / fail decision is made, and the copying of the weighting factor to the second adaptive equalization circuit is controlled in accordance with the result of the decision. According to this, the second adaptive equalization circuit is controlled. By performing the processing of the reproduced signal by using, the processing by the bad coefficient can be prevented, the good equalization can be always performed, and the convergence of the coefficient and the like can be quickened.
[0013]
BEST MODE FOR CARRYING OUT THE INVENTION
That is, the adaptive equalizer of the present invention has a first adaptive equalizer and a second adaptive equalizer that equalize a reproduced signal. It is provided with a judging means for judging the quality of the calculated weighting coefficient, and a control means for controlling copying of the weighting coefficient to the second adaptive equalization circuit according to the result of the judgment.
[0014]
Further, the playback apparatus of the present invention is a playback apparatus for extracting a playback signal from a recording medium, and has a first adaptive equalization circuit and a second adaptive equalization circuit for equalizing the playback signal. Then, the quality of the weighting coefficient calculated by the first adaptive equalization circuit is determined, and copying of the weighting coefficient to the second adaptive equalization circuit is controlled according to the result of the determination. .
[0015]
Hereinafter, the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing a configuration of an embodiment of a reproducing apparatus to which the present invention is applied.
[0016]
In FIG. 1, a signal extracted from a magnetic tape 1 through a reproducing head 2 and a reproducing amplifier 3 is supplied to an analog equalizer 5 through an automatic gain control circuit 4. The analog-equalized signal is supplied to an analog-to-digital converter (hereinafter, abbreviated as A / D conversion) device 6. A signal discretized by the A / D converter 6 is supplied to a phase-locked loop (hereinafter, abbreviated as PLL) 7 so as to perform a reproduction signal when the A / D converter 6 performs A / D conversion. Is extracted.
[0017]
The digital signal converted by the A / D converter 6 is supplied to a first adaptive equalizer 8 and a second adaptive equalizer 9. The output signal of the second adaptive equalization circuit 9 is supplied to a decoder 10 using, for example, a Viterbi decoding algorithm. In the Viterbi decoding algorithm, for example, the symbol sequence with the maximum path metric is estimated using the impulse response of each transmission path estimated with different weighting coefficients, and the symbol sequence with the maximum calculated path metric is estimated. Is output.
[0018]
Further, a signal from the decoder 10 is supplied to, for example, an error correction code (hereinafter, abbreviated as ECC) demodulation circuit 11, and digital data is reconstructed simultaneously with error correction. The reconstructed digital data is supplied to a decompression circuit (not shown) and the like, and decompression is performed to restore the compression performed at the time of recording. Further, this signal is converted into an analog signal by the D / A converter 12, and analog signals of, for example, video and audio signals are taken out to the output terminal 13. In this way, for example, digitally recorded video and audio signals are reproduced.
[0019]
In this apparatus, the adaptive equalizers 8 and 9 have a configuration as shown in FIG. 2, for example. In FIG. 2, a reproduction signal supplied to an input terminal 20 includes a plurality of serially connected unit delay units 21 to 24 constituting a first adaptive equalizing circuit 8 and a second adaptive equalizing circuit 9. Are provided to a plurality of unit delay means 31 to 34 connected in series. In FIG. 2, the unit delay means has four stages for the sake of simplicity. However, in an actual apparatus, ten or more stages are used.
[0020]
The signals at the input and output terminals of these unit delay means 21 to 24 are added by the adder 30 through the weighting means 25 to 29 of the weighting coefficients C1Ｃ to C5, respectively. The signals at the input and output terminals of the unit delay means 31 to 34 are added by the adder 40 through the weighting means 35 to 39 for the weighting coefficients C1Ｃ to C5Ｃ, respectively. As a result, an FIR filter is formed by the unit delay means, the weighting means and the adder, and the reproduced signal is equalized.
[0021]
Further, the output signal of the adder 30 is supplied to the temporary determination circuit 41. The ternary (+1, 0, -1) target value extracted from the tentative determination circuit 41 is supplied to the arithmetic circuit 42, and is subtracted from the output signal of the adder 30, whereby an error from the target value is obtained. Taken out. Then, the error and the reproduction signal supplied to the input terminal 20 are supplied to the coefficient calculation circuit 43. The coefficient operation circuit 43 estimates the impulse response of the transmission path using, for example, a least-squares method (Least {MeanSquare}: LMS) algorithm. The set weighting coefficients C1 # to C5 # are obtained.
[0022]
Further, the validity of the obtained weighting coefficients C1 # to C5 # is determined by the control circuit 44. Then, when the validity of the weighting coefficients C1Ｃ to C5 認 め is recognized by the control circuit 44, the weighting coefficients C1 to C5 are transmitted through the transmission circuit 45 and set in the weighting means 35 to 39 of the second adaptive equalization circuit 9. Is done. In this manner, the weighting coefficients C1 # to C5 # are set in the weighting means 35 to 39 of the second adaptive equalization circuit 9. That is, the output signal of the second adaptive equalization circuit 9 in which the weighting coefficients C1 # to C5 # are set in the weighting means 35 to 39 in this way is supplied to the decoder 10 using, for example, a Viterbi decoding algorithm.
[0023]
In this device, the validity of the weighting coefficients C1 # to C5 # in the control circuit 44 is determined, for example, as described below. In the following description, the circuit shown in FIG. 2 will be described with reference to FIG. However, in the configuration of FIG. 3, the circuits of the unit delay means 21 to 24, the weighting means 25 to 29, and the adder 30 in the circuit of FIG.
[0024]
The output signal of the equalization circuit 200 is supplied to the provisional decision circuit 41 and also supplied to the arithmetic circuit 42, and the target value from the provisional decision circuit 41 is subtracted from the output signal of the equalization circuit 200. The error between the extracted output signal and the target value is supplied to the coefficient arithmetic circuit 43 together with the reproduced signal supplied to the input terminal 20, and the weighting coefficients C1 # to C5 # are calculated. Further, the calculated weighting coefficients C1 # to C5 # are supplied to the second adaptive equalization circuit 9 through the transmission circuit 45, and the transmission in the transmission circuit 45 is controlled by the control circuit 44.
[0025]
The input signal supplied to the equalization circuit 200 is stored in the memory 46. For example, when the calculated weighting coefficients C1 # to C5 # are determined to be defective by the control circuit 44, and the same data is supplied to the coefficient calculation circuit 43 again when the weight calculation coefficients C1 # to C5 # are calculated again by the coefficient calculation circuit 43. For example, the input signal stored in the memory 46 can be used to calculate the repetition coefficients C1 # to C5 #.
[0026]
In the configuration of FIG. 3, various signals are further supplied to the control circuit 44, for example, as shown in FIGS. 4 to 7, and the validity of the weighting coefficients C1 # to C5 # is determined.
[0027]
Therefore, in the configuration of FIG. 4, the determination is made using the distance of the coefficient. FIG. 8 shows a conceptual diagram. That is, in FIG. 4, the weighting coefficients C1 # to C5 # calculated by the coefficient calculation circuit 43 are supplied to the control circuit 44. Here, the center coefficient of the first operation of the adaptive equalization circuit is 1, and the other coefficients are all 0. Then, the optimum coefficient of the FIR filter reflecting the initial characteristics of the transmission path of the reproducing apparatus and the characteristics of the standard recording medium is obtained, and is stored in the nonvolatile memory in the control circuit 44 as the initial coefficient.
[0028]
Further, in the subsequent operation, the distance between the stored initial coefficient and the newly obtained coefficient is obtained according to the following equation.
D = {α (k)} (C (k)} − Ci (k))²
Where D is the distance of the coefficient
C (k): current coefficient
Ci (k): initial coefficient
α (k): Constant group
When the number of taps is N, k = 0, 1,... N-1
[0029]
Here, the distance of the coefficient changes due to the aging of the reproducing apparatus, the variation in the characteristics of the recording medium, and the like. When the aging and the variation in the variation of the recording medium are defined, the variation in the coefficient distance falls within a predetermined range. It fits. Therefore, for example, when the coefficient distance becomes larger than a predetermined range, it can be generally considered that the coefficient has changed abnormally, that is, for example, that the coefficient has diverged.
[0030]
Therefore, for example, the waveform distortion (equalization error) to be absorbed by the FIR filter is mainly composed of the frequency characteristic of the reproduction signal sensitivity and the phase rotation depending on the growth angle of the tape magnetic layer as the recording medium. Is defined as follows, for example.
Frequency characteristics: -6 ≦ 1T / 2T (dB) ≦ + 4
Phase rotation: -15 ≦ Phase (°) ≦ + 15
[0031]
That is, when the sampling frequency is Fs, the gain at the frequency of Fs × １ ／ is represented by 1T, and the gain at the frequency of Fs × １ ／ is represented by 2T. If, for example, foreign matter adheres to the surface of the reproducing head and the spacing with the tape becomes large, the gain in the high band becomes lower than that in the middle and low bands, so that 1T / 2T (dB) becomes relatively small. . The lowest value of -6 (dB) indicates that the gain in the high frequency band is approximately halved.
[0032]
On the other hand, the phase rotation is a waveform distortion depending on the characteristics of the tape, and this value reflects the deposition angle of the magnetic layer of the tape. That is, in the above definition, for example, a state of −15 (°) is a case where the deposition angle is relatively inclined in a direction parallel to the tape surface by 15 (°) relative to the reference tape.
[0033]
Therefore, a method of obtaining the above-described threshold value representing the predetermined range will be described by taking an experiment using a certain tape reproducing device as an example. That is, this device is capable of outputting a reproduction signal (digital data) to which arbitrary characteristics are added. Then, the reproduced data was taken into an external personal computer (hereinafter abbreviated as PC), a simulation equivalent to an adaptive equivalent circuit was performed, and an optimum coefficient and a coefficient distance were obtained. However, in the above equation for calculating the coefficient distance D, the constant set α (k) is calculated as 5 only for the center coefficient and as 1 for the rest.
[0034]
Here, FIG. 9 shows a coefficient distance for obtaining an initial coefficient in a reproduction signal based on a reference parameter [1T / 2T (dB) = 0, Phase (°) = 0]. In FIG. 9, the coefficients converge at 50,000 or more loops, and the coefficient distance shows a constant value of about 0.25. The coefficient is stored in the memory, and from the next trial, the operation of adaptive equalization is performed using this coefficient as an initial value. In other words, FIG. 10 shows the result of performing the calculation again using the reference reproduction signal, and it can be seen that the coefficient is largely small at 0.1 or less and the coefficient does not change much.
[0035]
Next, it will be verified how the coefficient distance changes within the range of the above parameters. That is, FIG. 11 shows the case of the parameter [1T / 2T (dB) =-6, Phase (°) = 0], and FIG. 12 shows the case of the parameter [1T / 2T (dB) =-6, Phase (°) = -15. FIG. 13 shows the case of the parameter [1T / 2T (dB) = + 4, Phase (Ｐ °) = 0], and FIG. 14 shows the case of the parameter [1T / 2T (dB) = + 4, Phase (°) = −15. ]. In these cases, the coefficient distance is maximum in the case of FIG. 11 and is about 0.8.
[0036]
Therefore, from these experiments, it is understood that the coefficient distance does not exceed 0.8 within the range of the specified parameter, and this indicates that the above-described threshold value may be set to 0.8. That is, the control circuit 44 calculates the coefficient distance between the weighting coefficients C1 to C5 supplied from the coefficient calculation circuit 43 and the initial coefficient, and when the value exceeds 0.8, it is determined that the coefficient has abnormally changed. It can be judged.
[0037]
In order to examine the effect of the above-described coefficient quality determination, a simulation of adaptive equalization was performed using input data including a reproduction signal of a non-recording tape. Normally, when the coefficient is changed based on a signal having a characteristic deviating from a normal reproduction signal, such as a signal of a non-recording tape, an optimum coefficient may not be obtained in some cases.
[0038]
That is, for example, in the simulation shown in FIG. 15, it can be seen that the output signal immediately after reproducing the non-recorded portion converges to 0, and the magnitude does not recover even if a normal signal is input thereafter. This is because the coefficient of the FIR filter has changed abnormally and cannot return to the optimum value. Further, when the coefficient distance in this simulation was examined, as shown in FIG. 16, the coefficient distance increased in the non-recorded portion and greatly exceeded 1.0.
[0039]
On the other hand, when the quality of the coefficient is determined using the above-described coefficient distance threshold value of 0.8, and when the coefficient is bad, a process of holding the coefficient without updating is added. As shown, it has been found that even if a non-recording signal is input, if a normal signal is input thereafter, the coefficient recovers to the optimum value. In this manner, by performing the quality judgment using the coefficient distance, it is possible to prevent a situation in which the coefficient of the FIR filter abnormally changes and cannot return to the optimum value, and it is possible to always prevent the situation. A good coefficient can be updated.
[0040]
Next, in the configuration of FIG. 5, the determination is made by comparing the magnitudes of the input and output signals. That is, in FIG. 5, the input signal and the output signal of the equalization circuit 200 are supplied to the control circuit 44. When the automatic gain control circuit 4 is provided before the adaptive equivalent circuits 8 and 9 as shown in FIG. 1, the magnitude of the input signal of the adaptive equivalent circuit is Since the decoder 10 has been optimized to some extent, the fact that the magnitude of the output signal of the adaptive equivalent circuit differs greatly from the input signal is unlikely in a normal operation range.
[0041]
Therefore, the control circuit 44 is provided with means for monitoring the ratio of the magnitude of the output signal to the magnitude of the input signal. If the ratio deviates greatly from 1 by this monitoring means, it can be determined that the coefficient of the FIR filter is abnormal.
[0042]
In the same manner as described above, a simulation of adaptive equalization was performed using input data including a reproduction signal of a non-recording tape, and it was examined how the magnitude ratio of the input / output signal changes due to a coefficient abnormality. FIG. 19 shows the result. However, the ratio R is obtained by the following equation.
R = ΣY²÷ ΣX²
Where R is the magnitude of the output signal relative to the input signal
X: Input signal
Y: output signal
Σ is a temporal integration operation, which is performed every 180 loops.
[0043]
Therefore, as is clear from FIG. 19, the ratio R largely deviates from 1 immediately after reproducing the non-recorded portion and converges to about 0.1. On the other hand, in this simulation, a change was made so that the coefficient was not updated when the ratio R became 0.5 or less, and as shown in FIGS. It was found that the coefficient was restored to the optimum value when input.
[0044]
In this way, by comparing the magnitude of the input / output signal and determining the quality of the coefficient, it is possible to prevent a situation in which the coefficient of the FIR filter changes abnormally and cannot return to the optimum value. , And a good coefficient update can always be performed. In the above description, the coefficient is not updated when the ratio R is 0.5 or less. However, the ratio R may be larger than 1 in the same manner. You also need to set a value threshold.
[0045]
In the configuration of FIG. 6, the determination is performed using the probability density distribution of the temporary determination value. That is, in FIG. 6, the judgment value of (+1, 0, −1) is supplied to the control circuit 44 from the temporary judgment circuit 41. Here, in the reproducing apparatus used in the above-described experiment, for example, equalization by a partial response class 4 (hereinafter abbreviated as PR4) is performed. In this case, a normal reproduced signal is input to the determination circuit. In this case, it has been found that the probability density of the judgment value of (+1, 0, -1) approaches (0.25, 0.50, 0.25).
[0046]
Therefore, the control circuit 44 obtains the probability density of the target value (+1, 0, -1) determined by the temporary determination circuit 41. When the probability density deviates from (0.25, 0.50, 0.25), it can be determined that the coefficient is abnormal or the input signal is abnormal.
[0047]
In the same manner as described above, a simulation of adaptive equalization was performed using input data including a reproduction signal of a non-recording tape, and the probability density distribution of the tentative determination value was examined. The result is shown in FIG. However, the probability density is obtained by the following equation.
P (+1) = d (+1) ÷ (d (+1) + d (0) + d (−1))
P (0) = d (0) ÷ (d (+1) + d (0) + d (−1))
P (-1) = d (-1) ÷ (d (+1) + d (0) + d (-1))
Here, P (n) is the probability density of the tentative judgment value n.
d (n): number of temporary judgment values n
[0048]
That is, as shown in FIG. 15, since the output signal of the adaptive equalization circuit after reproducing the non-recorded portion has converged to around 0, the tentative judgment values based on this output signal are all 0. turn into.
[0049]
On the other hand, if the coefficient is updated only when all of the following three conditions are satisfied, as shown in FIGS. 23 and 24, if the normal signal is input after the non-recording signal, It was found that the coefficient recovered to the optimum value.
0.25-Pth ≦ P (+1) ≦ 0.25 + Pth
0.25-2Pth ≦ P (0) ≦ 0.25 + 2Pth
0.25-Pth ≦ P (−1) ≦ 0.25 + Pth
However, Pth = 0.1
[0050]
In this manner, by performing the quality judgment of the coefficient using the probability density distribution of the tentative judgment value, it is possible to prevent a situation in which the coefficient of the FIR filter abnormally changes and cannot be returned to the optimum value. , And a good coefficient update can always be performed.
[0051]
Further, in the configuration of FIG. 7, the determination is performed using the average value of the square errors. Here, the difference between the output signal of the adaptive equalization circuit and the target value tentatively determined based on the output signal is referred to as a detection point error. It is possible to determine whether or not updating is possible. That is, in FIG. 7, the ternary (+1, 0, -1) target value extracted from the tentative determination circuit 41 and the output signal of the arithmetic circuit 42 are supplied to the control circuit 44.
[0052]
Therefore, in the experiment, for example, the average value of the square error is obtained by the following equation.
E (+1) = Σe (+1)²÷ d (+1)
E (0) = Σe (0)²÷ d (0)
E (-1) = Σe (-1)²÷ d (-1)
Here, E (n) is the average value of the square error when the provisional determination value is n.
e (n): error at the provisional judgment value n
d (n): number of temporary judgment values n
[0053]
In addition, Σ is a temporal integration operation, and each square error is calculated every time d (+1) + d (0) + d (−1) = 180 loops. However, if there is d (n) less than 180 loops, the corresponding E (n) is not calculated.
[0054]
Therefore, in the same manner as described above, a simulation of adaptive equalization was performed using input data including a reproduction signal of a non-recording tape, and values of E (n) were obtained. As a result, as shown in FIG. 25, the value of E (n) is about 0.01 while the optimum coefficient is obtained, but E (0) immediately after the non-recorded portion is input. Obtained a value exceeding 0.06, and subsequently decreased by updating the coefficient. On the other hand, as for E (-1) and E (+1), even if a non-recorded portion and a subsequent normal signal are input, the values greatly exceed 0.1.
[0055]
Here, when the square error in the specified range of the above-mentioned waveform distortion was examined, it was about 0.06 at most. Therefore, it is assumed that the coefficient is updated only when all of the following three equations are satisfied. As shown in FIGS. 26 and 27, when a normal signal is input after the non-recording signal, it has been found that the coefficient recovers to the optimum value.
E (+1) ≦ Eth
E (0) ≦ Eth
E (-1) ≦ Eth
However, Eth = 0.06
[0056]
In this way, by performing the quality judgment of the coefficient using the average value of the square error, it is possible to prevent a situation in which the coefficient of the FIR filter abnormally changes and cannot return to the optimum value. , A good coefficient can always be updated.
[0057]
In the circuits shown in FIGS. 3 and 4 to 7 described above, various signals are supplied to the control circuit 44, and the appropriateness of the weighting coefficients C1 to C5 calculated by the coefficient operation circuit 43 is determined. Done. When the calculated coefficient is defective, the transmission in the transmission circuit 45 is controlled, and a situation in which a defective coefficient is supplied to the second adaptive equalization circuit 9 can be prevented. FIG. 28 is a table showing the quality of the coefficient in each of the circuits shown in FIGS.
[0058]
Further, according to the above-described apparatus, the above-described coefficient evaluation can be performed without affecting the error rate of signal reproduction at all. That is, unless the coefficient is transmitted to the second adaptive equalization circuit 9 through the transmission circuit 45, the error rate does not have any influence. In addition, the coefficient can be freely calculated and evaluated.
[0059]
As a result, the coefficient update loop gain can be set larger than usual, and the quality of the coefficient can be determined quickly. Further, if the input signal is stored in the memory 46, it is possible to perform a more accurate calculation using the same data again after the pass / fail judgment. Note that the memory 46 is not required to be provided when the circuit configuration is implemented by hardware and the above-described operation is not performed again, but is required when the first adaptive equalization circuit 8 is implemented by software. Become.
[0060]
Further, the initial value of the coefficient set in the FIR filter can be set at any time, but the value needs to be appropriate. As the initial value, for example, it is conceivable that the weight of the center coefficient is set to 1 and the other values are set to 0, and the equalizing circuit is a simple passing circuit having no equalizing ability. Alternatively, a past optimum value may be stored and used as an initial value.
[0061]
Therefore, in the above-described embodiment, the first adaptive equalizer and the second adaptive equalizer for equalizing the reproduced signal are provided, and the weighting coefficient calculated by the first adaptive equalizer is calculated. By judging pass / fail and controlling the copying of the weighting coefficient to the second adaptive equalizing circuit in accordance with the result of this judgment, processing of the reproduced signal using the second adaptive equalizing circuit is performed. Is performed, processing by a bad coefficient is prevented, good equalization can always be performed, and convergence of a coefficient and the like can be quickened.
[0062]
As a result, in the conventional apparatus, the validity of the coefficient of the adaptive equalization circuit is evaluated and determined based on the output, and thus the error rate of the reproduced signal may be deteriorated. On the other hand, methods such as increasing the accuracy of evaluation judgment or reducing the loop gain are conceivable, but any of these methods slows down the convergence of the coefficients or the like and fails to perform a good adaptive process. According to the present invention, these problems can be easily solved.
[0063]
The present invention is not limited to the above-described embodiment, and various modifications can be made without departing from the spirit of the present invention.
[0064]
【The invention's effect】
Therefore, according to the first aspect of the present invention, there is provided a first adaptive equalizing circuit and a second adaptive equalizing circuit for performing equalization of a reproduced signal, and the first adaptive equalizing circuit calculates the first adaptive equalizing circuit. Determining means for determining whether the weighting coefficient is good or bad, and control means for controlling copying of the weighting coefficient to the second adaptive equalizing circuit in accordance with a result of the determination. By performing the processing of the reproduced signal using the equalizing circuit, processing by a bad coefficient is prevented, good equalization can be always performed, and convergence of the coefficient and the like can be quickened.
[0065]
According to the second aspect of the present invention, the first adaptive equalizing circuit and the second adaptive equalizing circuit each include a delay unit including a plurality of unit delays for sequentially delaying the reproduced signal, A plurality of weighting means for weighting each of the delayed signals, an adding means for adding the weighted signals, and a determining means for identifying the added signal to a predetermined target value; The type equalizing circuit includes an arithmetic circuit for calculating a weighting coefficient of the weighting means based on a difference between an output signal of the adding means and an output signal of the determining means, and a good / bad determining circuit for determining whether the calculated coefficient is good or bad. And a control circuit for controlling the transmission circuit by a pass / fail judgment circuit. The second adaptive equalization circuit includes a first adaptive type equalizer. Coefficients from the equalization circuit By having a receiving circuit signal to one in which it is possible to easily realize the adaptive equalizer.
[0066]
According to the third aspect of the present invention, when the first adaptive equalization circuit has n sets of coefficients, the coefficient initial value Ci (k) (where k = 0, 1,...) Storage means for storing (n-1), and using a constant set α (k) (where k = 0, 1,..., N-1), a coefficient C (k) (where k = 0 , 1,... N-1)
{Α (k)} (C (k)}-Ci (k))²(However, k = 0, 1, ... n-1)
Has a pass / fail judgment circuit that judges that the coefficient is bad when the coefficient distance exceeds a predetermined range, and controls the coefficient not to be transmitted when the result of the pass / fail judgment circuit is bad. The determination using the distance is performed, and a good weighting coefficient can be calculated.
[0067]
According to the fourth aspect of the present invention, the first adaptive equalizing circuit and the second adaptive equalizing circuit each include a delay unit including a plurality of unit delays for sequentially delaying a reproduction signal, A plurality of weighting means for weighting each of the delayed signals, an adding means for adding the weighted signals, and a determining means for identifying the added signal to a predetermined target value; And comparing means for comparing the magnitude of the signal input to the type equalizing circuit with the magnitude of the signal added by the adding means, and transmitting the coefficient when the result of the comparing means is out of a predetermined range. By performing control so as not to make a determination, the magnitudes of the input and output signals are compared, and a good weighting coefficient calculation can be performed.
[0068]
According to the fifth aspect of the present invention, the first adaptive equalizing circuit and the second adaptive equalizing circuit each include a delay unit including a plurality of unit delays for sequentially delaying the reproduced signal, A plurality of weighting means for weighting each of the delayed signals, an adding means for adding the weighted signals, and a determining means for identifying the added signal to a predetermined target value; Statistical processing means for statistically processing the target value obtained by the determining means of the type equalizing circuit, and controlling so as not to transmit the coefficient when the result of the statistical processing means is out of a predetermined range. As a result, the determination using the probability density distribution of the temporary determination value is performed, and a good calculation of the weighting coefficient can be performed.
[0069]
According to the invention of claim 6, the first adaptive equalizing circuit and the second adaptive equalizing circuit each include a delay unit including a plurality of unit delays for sequentially delaying a reproduction signal, A plurality of weighting means for weighting each of the delayed signals, an adding means for adding the weighted signals, and a determining means for identifying the added signal to a predetermined target value; A statistical processing means for obtaining a difference between the signal added by the adding means of the type equalizing circuit and the target value obtained by the determining means of the first adaptive equalizing circuit, and statistically processing the difference; By controlling not to transmit the coefficient when the processing result is out of the predetermined range, the determination using the average value of the square error is performed, and the calculation of the good weighting coefficient can be performed. You can do it.
[0070]
According to a seventh aspect of the present invention, there is provided a reproducing apparatus for extracting a reproduction signal from a recording medium, comprising a first adaptive equalization circuit and a second adaptive equalization circuit for equalizing the reproduction signal. Then, the quality of the weighting coefficient calculated by the first adaptive equalization circuit is determined, and the copying of the weighting coefficient to the second adaptive equalization circuit is controlled in accordance with the result of the determination. The processing of the reproduced signal using the adaptive equalization circuit of the above prevents processing by a bad coefficient, can always perform good equalization, and can speed up the convergence of the coefficient and the like. It is.
[0071]
According to the eighth aspect of the present invention, the first adaptive equalizing circuit and the second adaptive equalizing circuit each include a delay unit including a plurality of unit delays for sequentially delaying the reproduced signal, A plurality of weighting means for weighting each of the delayed signals, an adding means for adding the weighted signals, and a determining means for identifying the added signal to a predetermined target value; The type equalizing circuit includes an arithmetic circuit for calculating a weighting coefficient of the weighting means based on a difference between an output signal of the adding means and an output signal of the determining means, and a good / bad determining circuit for determining whether the calculated coefficient is good or bad. And a control circuit for controlling the transmission circuit by a pass / fail judgment circuit. The second adaptive equalization circuit includes a first adaptive type equalizer. Coefficients from the equalization circuit By having a receiving circuit signal to one in which it is possible to easily realize a reproducing apparatus using an adaptive equalizer.
[0072]
According to the ninth aspect of the present invention, when the first adaptive equalizer has n sets of coefficients, the coefficient initial value Ci (k) (where k = 0, 1,...) Storage means for storing (n-1), and using a constant set α (k) (where k = 0, 1,..., N-1), a coefficient C (k) (where k = 0 , 1,... N-1)
{Α (k)} (C (k)}-Ci (k))²(However, k = 0, 1, ... n-1)
Has a pass / fail judgment circuit that judges that the coefficient is bad when the coefficient distance exceeds a predetermined range, and controls the coefficient not to be transmitted when the result of the pass / fail judgment circuit is bad. The determination using the distance is performed, and a good weighting coefficient can be calculated.
[0073]
According to the tenth aspect of the present invention, the first adaptive equalizing circuit and the second adaptive equalizing circuit each include a delay unit including a plurality of unit delays for sequentially delaying a reproduction signal, A plurality of weighting means for weighting each of the delayed signals, an adding means for adding the weighted signals, and a determining means for identifying the added signal to a predetermined target value; And comparing means for comparing the magnitude of the signal input to the type equalizing circuit with the magnitude of the signal added by the adding means, and transmitting the coefficient when the result of the comparing means is out of a predetermined range. By performing control so as not to make a determination, the magnitudes of the input and output signals are compared, and a good weighting coefficient calculation can be performed.
[0074]
According to the eleventh aspect of the present invention, the first adaptive equalizing circuit and the second adaptive equalizing circuit each include a delay unit including a plurality of unit delays for sequentially delaying a reproduction signal, A plurality of weighting means for weighting each of the delayed signals, an adding means for adding the weighted signals, and a determining means for identifying the added signal to a predetermined target value; Statistical processing means for statistically processing the target value obtained by the determining means of the type equalizing circuit, and controlling so as not to transmit the coefficient when the result of the statistical processing means is out of a predetermined range. As a result, the determination using the probability density distribution of the temporary determination value is performed, and a good calculation of the weighting coefficient can be performed.
[0075]
According to the twelfth aspect of the present invention, the first adaptive equalizing circuit and the second adaptive equalizing circuit each include a delay unit including a plurality of unit delays for sequentially delaying the reproduction signal, A plurality of weighting means for weighting each of the delayed signals, an adding means for adding the weighted signals, and a determining means for identifying the added signal to a predetermined target value; A statistical processing means for obtaining a difference between the signal added by the adding means of the type equalizing circuit and the target value obtained by the determining means of the first adaptive equalizing circuit, and statistically processing the difference; By controlling not to transmit the coefficient when the processing result is out of the predetermined range, the determination using the average value of the square error is performed, and the calculation of the good weighting coefficient can be performed. You can do it.
[0076]
As a result, in the conventional apparatus, the validity of the coefficient of the adaptive equalization circuit is evaluated and determined based on the output, and thus the error rate of the reproduced signal may be deteriorated. On the other hand, methods such as increasing the accuracy of evaluation judgment or reducing the loop gain are conceivable, but any of these methods slows down the convergence of the coefficients or the like and fails to perform a good adaptive process. According to the present invention, these problems can be easily solved.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating a configuration of an embodiment of a playback device to which the present invention has been applied.
FIG. 2 is a block diagram showing a configuration of an embodiment of an adaptive equalizer to which the present invention is applied.
FIG. 3 is a block diagram schematically showing the circuit of FIG. 2;
FIG. 4 is a block diagram of a circuit that makes a determination using a distance of a coefficient.
FIG. 5 is a block diagram of a circuit for comparing and judging the magnitude of an input / output signal.
FIG. 6 is a block diagram of a circuit that makes a determination using a probability density distribution of a provisional determination value.
FIG. 7 is a block diagram of a circuit that makes a determination using an average value of a square error.
FIG. 8 is a diagram for explaining the circuit of FIG. 4;
FIG. 9 is a diagram for explaining the circuit of FIG. 4;
FIG. 10 is a diagram for explaining the circuit of FIG. 4;
FIG. 11 is a diagram for explaining the circuit of FIG. 4;
FIG. 12 is a diagram for explaining the circuit of FIG. 4;
FIG. 13 is a diagram for explaining the circuit of FIG. 4;
FIG. 14 is a diagram for explaining the circuit of FIG. 4;
FIG. 15 is a diagram for explaining the circuit in FIG. 4;
FIG. 16 is a diagram for explaining the circuit of FIG. 4;
FIG. 17 is a diagram for explaining the circuit in FIG. 4;
FIG. 18 is a diagram for explaining the circuit in FIG. 4;
FIG. 19 is a diagram for explaining the circuit in FIG. 5;
FIG. 20 is a diagram for explaining the circuit of FIG. 5;
FIG. 21 is a diagram for explaining the circuit of FIG. 5;
FIG. 22 is a diagram for explaining the circuit in FIG. 6;
FIG. 23 is a diagram for explaining the circuit in FIG. 6;
FIG. 24 is a diagram for explaining the circuit in FIG. 6;
FIG. 25 is a diagram for explaining the circuit in FIG. 7;
FIG. 26 is a diagram for explaining the circuit in FIG. 7;
FIG. 27 is a diagram for explaining the circuit in FIG. 7;
FIG. 28 is a table summarizing descriptions of the circuits of FIGS. 4 to 7;
FIG. 29 is a block diagram illustrating a configuration of a conventional playback device.
FIG. 30 is a block diagram showing a configuration of a conventional adaptive equalization circuit.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Magnetic tape, 2 ... Reproduction head, 3 ... Reproduction amplifier, 4 ... Automatic gain control circuit, 5 ... Analog equalizer, 6 ... A / D converter, 7 ... Phase lock loop, 8 ... 1st adaptive type Equalization circuit, 9: second adaptive equalization circuit, 10: decoder using Viterbi decoding algorithm, 11: error correction code demodulation circuit, 12: D / A converter, 13: output terminal, 20: input terminal, 21 to 24, 31 to 34 ... a plurality of unit delay means connected in series, 25 to 29, 35 to 39 ... weighting means for weighting coefficients C1 # to C5 #, 30, 40 ... adder, 41 ... temporary judgment circuit, 42 ... Calculation circuit, 43: coefficient calculation circuit, 44: control circuit, 45: transmission circuit, 46: memory

Claims (12)

A first adaptive equalizer and a second adaptive equalizer that equalize the reproduced signal;
Determining means for determining whether the weighting coefficient calculated by the first adaptive equalization circuit is good or bad;
Control means for controlling copying of the weighting coefficients to the second adaptive equalization circuit in accordance with the result of the determination. The adaptive equalizer according to claim 1,
The first adaptive equalizer and the second adaptive equalizer include:
Delay means comprising a plurality of unit delays for sequentially delaying the reproduction signal,
A plurality of weighting means for weighting each of these delayed signals,
Adding means for adding these weighted signals;
Determining means for identifying the added signal to a predetermined target value,
The first adaptive equalization circuit includes:
An arithmetic circuit that calculates a weighting coefficient of the weighting means based on a difference between an output signal of the adding means and an output signal of the determining means;
A pass / fail judgment circuit for judging pass / fail of the calculated coefficient;
A transmission circuit for transmitting the calculated coefficient to the second adaptive equalization circuit;
A control circuit for controlling the transmission circuit by the pass / fail determination circuit,
The second adaptive equalization circuit includes:
An adaptive equalizer, comprising: a receiving circuit that receives the coefficient from the first adaptive equalizer. The adaptive equalizer according to claim 1,
If the first adaptive equalization circuit has n sets of the coefficients,
Storage means for storing an initial value Ci (k) (where k = 0, 1,..., N-1) of the coefficient;
Using a constant set α (k) (where k = 0, 1,..., N−1),
The coefficient distance of the coefficient C (k) (where k = 0, 1,.
√Σα (k) (C (k) −Ci (k)) ² (where k = 0, 1,..., N−1)
Defined by
Having a pass / fail determination circuit that determines that the coefficient distance is defective when the coefficient distance exceeds a predetermined range,
An adaptive equalizer, which performs control so as not to transmit the coefficient when the result of the pass / fail judgment circuit is defective. The adaptive equalizer according to claim 1,
The first adaptive equalizer and the second adaptive equalizer include:
Delay means comprising a plurality of unit delays for sequentially delaying the reproduction signal,
A plurality of weighting means for weighting each of these delayed signals,
Adding means for adding these weighted signals;
Determining means for identifying the added signal to a predetermined target value,
Comparing means for comparing the magnitude of the input signal to the first adaptive equalization circuit with the magnitude of the signal added by the adding means,
An adaptive equalizer, wherein the control is performed so as not to transmit the coefficient when the result of the comparing means deviates from a predetermined range. The adaptive equalizer according to claim 1,
The first adaptive equalizer and the second adaptive equalizer include:
Delay means comprising a plurality of unit delays for sequentially delaying the reproduction signal,
A plurality of weighting means for weighting each of these delayed signals,
Adding means for adding these weighted signals;
Determining means for identifying the added signal to a predetermined target value,
Statistical processing means for statistically processing the target value obtained by the determination means of the first adaptive equalization circuit;
An adaptive equalizer, wherein the coefficient is not transmitted when the result of the statistical processing means deviates from a predetermined range. The adaptive equalizer according to claim 1,
The first adaptive equalizer and the second adaptive equalizer include:
Delay means comprising a plurality of unit delays for sequentially delaying the reproduction signal,
A plurality of weighting means for weighting each of these delayed signals,
Adding means for adding these weighted signals;
Determining means for identifying the added signal to a predetermined target value,
A signal added by the adding means of the first adaptive equalization circuit;
Calculating the difference between the target values obtained by the determination means of the first adaptive equalization circuit;
It has statistical processing means for statistically processing the difference,
An adaptive equalizer that controls not to transmit the coefficient when the result of the statistical processing deviates from a predetermined range. A playback device for extracting a playback signal from a recording medium,
A first adaptive equalizer and a second adaptive equalizer that equalize the reproduced signal;
Determining whether the weighting coefficient calculated by the first adaptive equalizing circuit is good or bad;
A reproducing apparatus for controlling copying of the weighting coefficient to the second adaptive equalization circuit in accordance with a result of the determination; The playback device according to claim 7,
The first adaptive equalizer and the second adaptive equalizer include:
Delay means comprising a plurality of unit delays for sequentially delaying the reproduction signal,
A plurality of weighting means for weighting each of these delayed signals,
Adding means for adding these weighted signals;
Determining means for identifying the added signal to a predetermined target value,
The first adaptive equalization circuit includes:
An arithmetic circuit that calculates a weighting coefficient of the weighting means based on a difference between an output signal of the adding means and an output signal of the determining means;
A pass / fail judgment circuit for judging pass / fail of the calculated coefficient;
A transmission circuit for transmitting the calculated coefficient to the second adaptive equalization circuit;
A control circuit for controlling the transmission circuit by the pass / fail determination circuit,
The second adaptive equalization circuit includes:
A reproducing apparatus comprising: a receiving circuit that receives the coefficient from the first adaptive equalization circuit. The playback device according to claim 7,
If the first adaptive equalization circuit has n sets of the coefficients,
Storage means for storing an initial value Ci (k) (where k = 0, 1,..., N-1) of the coefficient;
Using a constant set α (k) (where k = 0, 1,..., N−1),
The coefficient distance of the coefficient C (k) (where k = 0, 1,.
√Σα (k) (C (k) −Ci (k)) ² (where k = 0, 1,..., N−1)
Defined by
Having a pass / fail determination circuit that determines that the coefficient distance is defective when the coefficient distance exceeds a predetermined range,
A reproducing apparatus, wherein the coefficient is not transmitted when the result of the pass / fail judgment circuit is defective. The playback device according to claim 7,
The first adaptive equalizer and the second adaptive equalizer include:
Delay means comprising a plurality of unit delays for sequentially delaying the reproduction signal,
A plurality of weighting means for weighting each of these delayed signals,
Adding means for adding these weighted signals;
Determining means for identifying the added signal to a predetermined target value,
Comparing means for comparing the magnitude of the input signal to the first adaptive equalization circuit with the magnitude of the signal added by the adding means,
A reproduction apparatus, wherein the coefficient is not transmitted when the result of the comparing means deviates from a predetermined range. The playback device according to claim 7,
The first adaptive equalizer and the second adaptive equalizer include:
Delay means comprising a plurality of unit delays for sequentially delaying the reproduction signal,
A plurality of weighting means for weighting each of these delayed signals,
Adding means for adding these weighted signals;
Determining means for identifying the added signal to a predetermined target value,
Statistical processing means for statistically processing the target value obtained by the determination means of the first adaptive equalization circuit;
A reproduction apparatus, wherein the coefficient is not transmitted when the result of the statistical processing means is out of a predetermined range. The playback device according to claim 7,
The first adaptive equalizer and the second adaptive equalizer include:
Delay means comprising a plurality of unit delays for sequentially delaying the reproduction signal,
A plurality of weighting means for weighting each of these delayed signals,
Adding means for adding these weighted signals;
Determining means for identifying the added signal to a predetermined target value,
A signal added by the adding means of the first adaptive equalization circuit;
Calculating the difference between the target values obtained by the determination means of the first adaptive equalization circuit;
It has statistical processing means for statistically processing the difference,
A reproduction apparatus, wherein the coefficient is not transmitted when the result of the statistical processing is out of a predetermined range.

Images