JP2000276848A - Waveform equalizer and playback device - Google Patents

Abstract

(57) [Problem] To provide a waveform equalizer capable of supplying an equalized waveform having a small error rate during decoding even to a reproduced signal having asymmetry. SOLUTION: In an equalizer 11 using a transversal filter, a detector 31 for detecting asymmetry is provided, and an equalization coefficient of a multiplier 17 is changed by a coefficient selection circuit 32 depending on the presence or absence of asymmetry. As a result, even when a reproduced signal having asymmetry is equalized, it is possible to prevent the dispersion of values at the detection points from increasing. Therefore, by performing maximum likelihood decoding (Viterbi decoding) on the output signal whose waveform has been equalized using the equalizer of the present invention, it is possible to provide a reproducing apparatus having a high decoding capability.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a waveform equalizer used for reproducing digital data recorded on a recording medium such as an optical disk, and a reproducing apparatus using the same.

[0002]

2. Description of the Related Art As an optical recording medium for recording digital data, a compact disk (CD) and a mini disk (M
D), and 130 mm and 90 mm optical disk cartridges and the like according to the standards of the International Organization for Standardization (ISO) have been commercialized. In recent years, DVD-standard optical discs having a recording capacity about seven times as large as CDs have appeared, and are attracting attention as recording media for moving image data whose data amount will increase in the future.

Conventionally, a CD or the like employs a bit-by-bit decoding method in which a read analog signal is determined to be "1" if the analog signal is higher than a predetermined level and "0" if the analog signal is lower than a predetermined level. With a significantly improved recording medium, it is difficult to reproduce digital data with high reliability.

For this reason, in recent years, as a method of reproducing digital data, a PRML method combining a partial response method (PR method or PR) and a maximum likelihood decoding method (ML method or ML) using a Viterbi decoding method has attracted attention. Have been.

FIG. 14 is a block diagram showing a general configuration of a disk drive 1 using the PRML method.
The processing in the optical disk device 1 is performed along the arrows in the block diagram, and the disk data recorded on the optical disk medium 2 is read and reproduced. First, the optical pickup 3 irradiates the optical disc medium 2 with laser light to read digital data recorded on the optical disc medium 2 as an analog signal, and the reproduced signal φ1 is amplified by the preamplifier circuit 4. The amplified analog signal is shaped by a waveform shaping circuit 5 having a function of a low frequency pass filter (LPF) and an equalizer (EQ) for shaping the waveform, and then converted to an analog / digital (A / D) converter 6 Is converted into a digital signal. The digitized reproduction signal is further equalized by an equalization circuit 11 into a waveform of a preset partial response characteristic (PR characteristic). In this way, the output signal φs
Is obtained and Viterbi-decoded by the maximum likelihood decoding circuit 20, and the digital data recorded on the optical disk medium 2 is reproduced.

In the PRML optical disk apparatus 1, the transfer characteristic of the signal reproducing system 10 is a predetermined PR characteristic, for example, PR (1221), until the reproduced signal is read from the optical disk medium 2 and supplied to the maximum likelihood decoding circuit 20. Viterbi decoding is performed in the maximum likelihood decoding circuit 20 on the assumption that it has characteristics. Therefore, the equalizer 11 has a predetermined P
It is adjusted so as to output an output signal φs whose waveform is shaped into an R characteristic, for example, a PR (1221) characteristic.

FIG. 15 shows an example of the waveform equalizer 11. The waveform equalizer 11 includes an equalizing filter, a transversal filter, or a FIR (Finite Impulse Response).
e) Also called a filter, a plurality of delay elements 15 connected in series, sampling taps 16 for outputting values before and after these delay elements 15, and an equalization method for the values obtained by the sampling taps 16. And an output unit 19 that outputs an output signal φs by multiplying and adding an equalization coefficient (tap coefficient) C according to. The output unit 19 generally includes a multiplier 17 corresponding to each sampling tap 16,
An adder 18 is provided for adding the multiplication results of the multipliers 17 and outputting a sample value. The delay time of the delay element 15 is set to the delay time of the bit period of the reproduced signal φ1, that is, the delay time D for one bit. In order to obtain n samplings, n-1 delay elements 15 are used. Connected in series.

[0008]

As shown in FIG. 16, an 8/1 optical disk employed in a DVD standard optical disk is used.
According to the evaluation of the bit error rate (BER) using 6 codes and various PRML methods, the PR (1221) method has the smallest error rate. Also, it can be seen that the PR (1221) method becomes more advantageous as the S / N ratio increases. This corresponds to PTF for MTF (Modulation Transfer Function) which is the frequency characteristic of the optical system of DVD standard.
This is probably because the frequency characteristics of the R (1221) system are similar. Therefore, the playback device of the optical disk device that handles the DVD standard optical disk has a PR (1221)
In the waveform equalizer 11, the output signal φs according to the PR (1221) characteristic is preferably used.
Is determined so that the following equation is obtained.
(1221) It is important to minimize the equalization error from the characteristics.

However, the reproduced signal φ1 obtained from the optical disk or the like is not always reproduced in an ideal state due to various conditions such as the data generation state on the optical disk, the state of reproducing the data, and the characteristics of the optical pickup. Phenomena such as inclusion of noise and distortion of the signal appear. One of them is a phenomenon called asymmetry.
The level (eg, voltage intensity) of the reproduced signal becomes asymmetric (non-linear) with respect to the bit length. For example, for DVD,
This refers to a case where the center level of the 3T amplitude is shifted from the center level of the amplitude of 14T (T is the channel bit length), and the shift amount is from -5% to + 1% of the amplitude of 14T.
Exceeding the range of 5% is a major factor in increasing the error rate.

That is, if there is positive asymmetry, 1
In the 3T waveform, the amplitude of the upwardly convex waveform becomes smaller and the amplitude of the downwardly convex waveform becomes larger in the 3T waveform. Conversely, if there is negative asymmetry, the amplitude of the upward convex waveform of the 3T waveform increases and the amplitude of the downward convex waveform decreases. Therefore, an output obtained by equalizing the waveform of a reproduced signal including such asymmetry is a predetermined PR (1221)
There is a case where it does not match the predicted value when performing Viterbi decoding due to deviation from the characteristic. For this reason, for example, Japanese Patent Application Laid-Open
No. 72366 discloses that when a Viterbi decoder obtains a decoded data string using a square error between a received sample value and a predicted value as a branch metric, a linear function using a linear function approximates it.

In Japanese Patent Application Laid-Open No. 8-221910, when obtaining a sample value from a read signal, the amplitude is limited so that a predicted value at the time of Viterbi decoding can be reduced. Further, in Japanese Patent Application Laid-Open No. 8-263943, a sample value smaller than a predetermined value is extracted from sample values of a reproduced signal,
The phase of the sampling clock pulse for A / D conversion is corrected based on the polarity and the slope, and the level of the sample value is automatically corrected to an optimum state.

However, these methods simply improve the decoding performance by reducing the error between the predicted value and the sample value,
Because of the attempt to improve the error rate, there are many cases where the error rate is not so much improved in a reproduced signal including asymmetry.

Accordingly, an object of the present invention is to provide a reproducing apparatus capable of greatly improving an error rate even for a reproduced signal including asymmetry.

For this reason, the inventors of the present application have studied and found that there is a large problem in equalizing the waveform of a reproduced signal including asymmetry. In view of the above, an object of the present invention is to provide a waveform equalizer that can significantly reduce an error rate when performing Viterbi decoding in a waveform equalizer that equalizes a reproduced signal including asymmetry. I have.

[0015]

The present inventors first studied how an original waveform including asymmetry was converted by a waveform equalizer. 1 and 2
Indicates an equalized waveform obtained when a 4T waveform including positive asymmetry is input as an original waveform. The equalized waveforms shown in FIG. 1 and FIG.
In the equalizer 11 shown in FIG. 5, a value determined so as to equalize a waveform without asymmetry is used as an equalization coefficient of the multiplier 17. The upwardly convex original waveform shown in FIG. 1 has almost no narrower skirt even if there is asymmetry. This is because the upwardly convex waveform does not become narrow even if the pit becomes narrow because the spot diameter of the laser beam is dominant in determining the waveform. As a result, even if equalization is performed using an equalization coefficient that does not consider asymmetry, the predicted value (target value)
Can be obtained. Therefore, the voltage distribution of the equalized waveform at this time is PR (122
The voltage distribution based on the 4T digital signal “011110” of 1), that is, each value of “0, 1, 3, 5, 6” is obtained, and the voltage histogram as shown in FIG. 3A is obtained.

On the other hand, as shown in FIG. 2, the original waveform which is convex downward has a wider base and an increased amplitude. Therefore, if this original waveform is equalized with an equalization coefficient that does not consider asymmetry, the equalized waveform deviates from the target value, and the value at the detection point deviates downward as a whole. This voltage distribution (the value at the detection point) is shown by a voltage histogram in FIG.
The voltage distribution is shifted downward from the values of 0, 1, 3, 5 and 6 as a whole. In particular, since the level 5 and the level 3 are largely shifted, the interval between the level 6 and the level 5 and the level 3 is widened. In addition, the interval between level 3 and level 1 and level 0 becomes narrower.

Thus, 4T with positive asymmetry
The voltage distribution when the waveform is equalized is a distribution as shown in FIG. 3C in which FIGS. 3A and 3B are superimposed. As can be seen from this figure, if there is asymmetry,
Not only does the voltage value at the detection points shift, but the variance at each detection point also increases. This tendency is not limited to the 4T solitary wave, but also applies to a reproduced signal in which 3T to 11T of 8/16 modulation adopted in DVD is randomly. As described above, according to the studies by the inventors of the present application, when a reproduced signal having asymmetry is subjected to Viterbi decoding of a waveform-equalized signal, the error rate can be significantly improved only by reducing the error from the predicted value. I knew I couldn't.

The same can be said for a reproduced signal having negative asymmetry. First, FIGS. 4 and 5 show equalized waveforms obtained when a 4T waveform including negative asymmetry is input as an original waveform. The equalized waveforms shown in these figures also use coefficients for equalizing an original waveform without asymmetry. When there is negative asymmetry, the upwardly convex original waveform shown in FIG. 5 has a wider tail and an increased amplitude. Therefore, if this original waveform is equalized with an equalization coefficient that does not take into account asymmetry, the equalized waveform will deviate from the target value, and the value at the detection point will be deviated upward as a whole. This voltage distribution (the value at the detection point) is shown by a voltage histogram in FIG. The voltage distribution is shifted upward from the values of 0, 1, 3, 5 and 6 as a whole. In particular, since the level 1 and the level 3 are largely shifted, the interval between the level 6 and the level 5 and the level 3 is narrowed, and In addition, the intervals between level 3 and level 1 and level 0 are widened.

On the other hand, if there is a negative asymmetry, the waveform of the downward convex has almost no narrower skirt of the waveform as shown in FIG. In this case, since the downwardly convex waveform is dominant in determining the laser beam spot diameter, the waveform does not become narrow even if the pit becomes narrow. Therefore, even if equalization is performed using an equalization coefficient that does not consider asymmetry, an equalized waveform that can sample a predicted value (target value) is obtained.
The result is as shown in FIG. Voltage histogram is obtained.

Therefore, 4T with negative asymmetry
The voltage distribution when the waveform is equalized is a distribution as shown in FIG. 6C in which FIGS. 6A and 6B are superimposed, and not only does the voltage value at the detection point shift, but also The variance at each detection point increases. Since this tendency is the same in a reproduced signal having random values from 3T to 11T, it can be seen that even in a reproduced signal having negative asymmetry, the error rate cannot be significantly improved merely by reducing the error from the predicted value.

As described above, when there is asymmetry, one of the vertically convex waveforms obtains a value substantially coincident with the target value at the detection point, while the other has a value at the detection point larger than the target value. Because of the deviation, the values at the detection points are dispersed, which is a major cause of an increase in the error rate. FIG. 7 schematically shows this state. As shown in FIGS. 7A and 7B, when an equalization coefficient is determined so as to match a predetermined PR characteristic, for example, the PR (1221) characteristic as described above, for an original waveform without asymmetry. The waveform after the equalization matches the target value at the detection point. In contrast, a waveform with asymmetry is
As shown in FIGS. 7C and 7D, a thick waveform (a wide waveform) is obtained. Therefore, when such an original waveform is equalized by the equalization coefficients shown in FIGS. 7A and 7B, the value at the detection point of the upwardly convex waveform greatly deviates from the target value, and the downwardly convex waveform At this point, it is greatly shifted below the target value. For this reason, for example, in a reproduced signal having positive asymmetry, FIG. 7A and FIG. 7D overlap each other, and not only the value at the detection point is shifted, but also the variance of the value is increased. In a reproduced signal having negative asymmetry, the tendency is reversed, but a similar phenomenon appears.

As described above, the inventors of the present application have found that in order to reduce the error rate in Viterbi decoding after equalization, it is first necessary to reduce the variance of values at the detection points. Was done. Therefore, the present inventors, when equalizing a reproduced signal having asymmetry,
We considered changing the equalization factor. Then, they have found that the variance can be reduced by using an equalization coefficient determined to equalize a waveform with asymmetry slightly wider (wider) than a waveform with asymmetry and an intermediate waveform substantially intermediate with the waveform without asymmetry.

A description will be given with reference to FIG. First, FIG.
As shown in (b) and (b), the original waveform of the bold
The equalization coefficient is determined so as to match the characteristics. When the original waveform thinner than the middle-thick waveform is equalized using the equalization coefficient, FIG.
As shown in (c) and (d), at the detection point, the upwardly convex waveform shifts slightly below the target value, and the downwardly convex waveform shifts slightly above the target value. On the other hand, when the original waveform that is thicker than the medium-thick waveform is equalized, as shown in FIGS.
In the case of a downwardly convex waveform, the waveform deviates slightly below the target value.

As described above, a waveform having positive asymmetry has a thin upwardly convex waveform and a thick downwardly convex waveform. Therefore, original waveforms corresponding to FIGS. 8C and 8F are obtained, and when the original waveforms are equalized by the equalization coefficients shown in FIG. 8, both the upwardly convex and downwardly convex waveforms are detected at the detection points. , A value slightly lower than the target value is obtained. Therefore, at the detection point, a value slightly deviated from the target value but small in variance can be obtained. For this reason, by supplying this equalized waveform to the maximum likelihood decoding device of the Viterbi decoding method capable of appropriately correcting the predicted value, the error rate can be extremely reduced.

The same applies to waveforms having negative asymmetry. That is, a waveform having negative asymmetry has a thick upward convex waveform and a narrow downward convex waveform. Therefore,
Since the original waveforms shown in FIGS. 8D and 8E are obtained, when equalization is performed, the values at the detection points are both slightly shifted upward, and the variance is reduced. Therefore, the error rate can be reduced by using the Viterbi decoding method that can similarly correct the predicted value.

Since the positive or negative asymmetry is almost uniquely determined by the combination of the optical disk or the pickup, the presence or absence of the asymmetry of the reproduced signal is detected in advance, and when the asymmetry exists, the equalization coefficient is changed so that the detection point at the detection point is determined. Dispersion can be reduced. Thus, the error rate can be greatly improved only by changing the equalization coefficient depending on the presence or absence of asymmetry. Of course, it is possible to improve the variance of the voltage distribution by further adjusting the equalization coefficient depending on the amount of asymmetry. For example, when the amount of asymmetry is not so large, the waveform of the middle bold (FIG. 8A and FIG. (B)) and an intermediate waveform (average waveform) between the narrow waveforms (FIGS. 8 (c) and (d)) can be used, and the equalization coefficient obtained as the original waveform can be used.

That is, according to the present invention, there is provided a waveform equalizing apparatus for equalizing a waveform of a reproduced signal obtained from a recording medium and outputting the equalized signal, and a detecting means for detecting asymmetry of the reproduced signal, and at least detecting the asymmetry of the reproduced signal. Control means for changing the characteristics of the device. Further, in the method of controlling a waveform equalizer according to the present invention for equalizing a waveform of a reproduced signal obtained from a recording medium and outputting the same, the step of detecting the asymmetry of the reproduced signal includes the steps of: Changing the characteristics of the device.

A plurality of delay elements connected in series in order to sample a reproduced signal in time series, and a plurality of sample values obtained by these delay elements are multiplied by a predetermined equalization coefficient and then added to each other to form a waveform. In the waveform equalizer having an output unit that outputs a converted output signal, the control unit can obtain the above-described effect by changing the equalization coefficient at least according to the presence or absence of asymmetry,
An equalized waveform with small variance at the detection point can be output. The equalization coefficient selected when there is asymmetry by the control means is based on the original waveform of the reproduced signal without asymmetry, the waveform that is wider in both the positive and negative directions of asymmetry, and the above-mentioned middle-thick waveform. It is desirable that the equalization coefficient is assumed to be equalized as a waveform.

In a DVD employing 8/16 modulation, a signal which does not normally appear as a sync signal is used.
The signal of T is adopted as a sync signal. For this reason,
The detecting means can determine the presence or absence of asymmetry based on the amplitude intensity of the signal following the sync signal.

[0030] A reproduction apparatus having a waveform equalizer of the present invention and maximum likelihood decoding means for maximum likelihood decoding of an output signal output from the waveform equalizer using a plurality of predicted values and reproducing digital data. In the apparatus, even when a reproduced signal with asymmetry is input, decoding can be performed using an equalized waveform with small variance. For this reason, it is possible to provide a reproducing apparatus having a low error rate and high decoding ability. further,
As described above, it is desirable for the maximum likelihood decoding means to be able to correct the predicted value by the value at the detection point.

[0031]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 9 shows a configuration of the equalizer 11 according to the embodiment of the present invention. This equalizer 11
Constitutes the optical disk device 1 described above with reference to FIG. 14, and a detailed description of the optical disk device 1 will be omitted below.

The equalizer 11 of the present embodiment has a plurality of delay elements 15 connected in series and a plurality of sampling taps 16 for outputting values before and after these delay elements 15. An output unit 19 is provided which multiplies each of the values obtained in step 16 by a weighting coefficient (equalization coefficient) C according to the equalization method and adds the multiplied values to output an output signal φs.
The output unit 19 generally includes each sampling tap 1
6 and an adder 18 that adds the multiplication results of the multipliers 17 and outputs a sample value. The equalizer 11 shown in FIG.
However, it is a matter of course that the number may be three or more.

The equalizer 11 of this embodiment further includes a reproduction signal φ
An asymmetry detector 31 capable of detecting the presence or absence of asymmetry from 1 and a coefficient selection circuit 32 for changing the equalization coefficient C of each multiplier 17 when the reproduced signal φ1 has asymmetry. The coefficient selection circuit 32 of the present example selects the asymmetry equalization coefficient C from a look-up table prepared in advance in the memory 33 and changes the setting of the multiplier 17, but has different variables in advance. Of course, the method and circuit for changing the equalization coefficient C, such as changing the setting of the multiplier, are not limited to this example. Then, instead of using the equalization coefficient for equalizing the reproduced signal φ1 without asymmetry, the equalizer 1 is used by using the equalization coefficient prepared for the reproduced signal φ1 with asymmetry.
1 can be changed, and as described above, by using an equalization coefficient for equalizing an original waveform slightly wider than the original waveform without asymmetry, a reproduced signal with asymmetry can be obtained. Can output an equalized waveform having a small value variance at the detection point.

In selecting an equalization coefficient, it is important to determine the presence or absence of asymmetry of the reproduced signal φ1. In this example, the asymmetry detector 31
The presence or absence of asymmetry is determined by comparing the amplitude W14 of the 14T sync signal (Sync Pattern) with the amplitude W4 of the subsequent 4T waveform. As shown in FIG. 10A, when there is no asymmetry, the reproduced signal of the sync signal has the center line M of the amplitude of the 14T signal.
14 coincides with the center line M4 of the amplitude of the 4T signal that follows. On the other hand, if there is positive asymmetry,
0 (b), the center line M of the amplitude of the 14T signal
14, the center line M4 of the amplitude of the 4T signal shifts upward. As a result, looking at the upwardly convex signal, the ratio of the 4T amplitude W4 to the 14T signal amplitude W14 increases. On the other hand, when there is negative asymmetry, as shown in FIG. 10C, the center line M4 of the amplitude of the 4T signal is shifted downward with respect to the center line M14 of the amplitude of the 14T signal. Therefore, looking at the upwardly convex signal, the ratio of the 4T amplitude W4 to the 14T signal amplitude W14 becomes smaller. Although the polarity of the sync signal may be reversed, the relationship between asymmetry and amplitude can be similarly grasped.

The asymmetry detector 31 shown in FIG.
Is an example of determining the presence or absence of asymmetry by calculating the ratio of the amplitude of such a 14T sync signal to the amplitude of the subsequent 4T signal. The end point of the 14T sync signal is detected by supplying the reproduced signal φ1 to the 14T signal detector 41. Then, the timing value is held by the timing circuit 42 and the hold circuit 44.
On the other hand, the delay circuit 43 detects the peak timing of the 4T signal from the end point of the 14T sync signal, and the hold circuit 45 holds the value of the timing. Although the values held by the hold circuits 44 and 45 change depending on the polarity of the sync signal, the amplitude W4 of the 4T signal can be obtained by calculating these values by the subtractor 46.

The reproduced signal φ1 is also supplied to a peak value detection circuit 47 and a bottom value detection circuit 48.
By calculating these values by the subtractor 47, 14
The amplitude W14 of the T signal can be obtained. Therefore,
By calculating the outputs of the subtracters 46 and 49 by the divider 50, the ratio of the amplitude can be output, and based on this, the presence or absence of asymmetry can be determined. Of course, the present invention is not limited to such a circuit, and the asymmetry of the reproduced signal can be calculated using the sync signal by various methods such as calculating and comparing the center value M14 of the 14T signal and the center value M4 of the 4T signal. I can judge. If the conditions of the optical disk, the optical pickup, and the like are constant, the presence or absence of the asymmetry or the direction thereof is almost uniquely determined. For this reason, the coefficient selection circuit 32 of the equalizer 11 can appropriately change the equalization coefficient according to the result of the asymmetry determination based on the sync signal.

FIGS. 12 and 13 show the equalizer 11 of this embodiment.
The value at the detection point of the waveform obtained by equalizing the reproduced signal with asymmetry using is shown by a voltage histogram. FIG. 12 is a voltage histogram when a reproduced signal having positive asymmetry is equalized, and FIG.
It is a figure contrasted with. In the equalizer 11 of the present example,
If it is determined that the reproduced signal has positive asymmetry, the equalization coefficient C is switched to a coefficient for equalizing the medium-sized original waveform in consideration of the asymmetry described with reference to FIG. As a result, FIG.
As described above, the original waveform convex above the positive asymmetry corresponds to FIG. 8C, and the value at the detection point shifts slightly downward. Therefore, a voltage histogram shown in FIG.

The downwardly convex original waveform corresponds to FIG. 8 (f), and the value at the detection point also shifts slightly downward.
Therefore, a voltage histogram shown in FIG. 12B is obtained. As a result, a voltage histogram of a reproduced waveform having positive asymmetry is as shown in FIG. 12C in which FIGS. 12A and 12B are superimposed. That is, since the values at the detection points of the upwardly convex and downwardly convex waveforms are both slightly shifted downward, the values are shifted downward as a whole, but the variance of the values at the respective detection points is reduced.

FIG. 13 is a voltage histogram when a reproduced signal having negative asymmetry is equalized, and is a diagram to be compared with FIG. 6 described above. In the equalizer 11 of this example, even when it is determined that the reproduced signal has a negative asymmetry, the equalization coefficient C is set to a coefficient for equalizing the medium-thick original waveform in consideration of the asymmetry described in FIG. Switch. As a result, as described with reference to FIG. 8, the original waveform convex above the negative asymmetry corresponds to FIG. 8E, and the value at the detection point shifts slightly upward. Therefore, FIG.
The voltage histogram shown in FIG.

The original waveform having a downward convexity corresponds to FIG. 8D, and the value at the detection point is also slightly shifted upward.
Therefore, a voltage histogram shown in FIG. 13B is obtained. As a result, the voltage histogram of the reproduced waveform having negative asymmetry is as shown in FIG. 13C in which FIGS. 13A and 13B are superimposed. That is, since the values at the detection points of the upwardly convex and downwardly convex waveforms are both slightly shifted upward, the values are shifted upward as a whole, but the variance of the values at the respective detection points is reduced.

As described above, in the waveform equalizer 11 of this embodiment, when the reproduced signal having asymmetry in either the positive or negative direction is equalized, the value at the detection point slightly shifts downward or upward, but the variance does not increase. Can be smaller.
In the case of the conventional waveform equalizer, when there is asymmetry, the value at the detection point is shifted and the variance is widened, which causes the error rate to increase. Is one of the factors,
Further, it is possible to remove a factor that cannot be processed by the maximum likelihood decoding circuit 20. Therefore, by employing the waveform equalizer 11 of the present example, it is possible to provide a reproducing apparatus with few decoding errors, high decoding speed, and high decoding ability. In particular, by combining with a maximum likelihood decoding circuit having a function of correcting a predicted value at the time of maximum likelihood decoding according to the value of a detection point, it becomes possible to remove the influence of a value shift at a detection point, It is possible to provide a reproducing apparatus having a very high decoding ability.

In the above description, the case where the asymmetry detector 31 determines the presence or absence of asymmetry is described as an example. However, it is of course possible to determine the amount of asymmetry using the asymmetry detector 31. is there. Then, an equalization coefficient corresponding to the amount of asymmetry is prepared in the memory 33 of the equalizer 11, and the equalization coefficient can be optimally set according to the amount of asymmetry. As described above, the equalization coefficient corresponding to the amount of asymmetry can be determined as appropriate by changing the thickness of the original waveform at the time of obtaining the equalization coefficient. The equivalent coefficient can be calculated in advance at a pitch of about a degree, formed into a look-up table, and stored in the memory 33.

Also, in this specification, as shown in FIG. 14, the reproduced signal is A / D-converted and then the waveform is equalized by the equalizer. Of course it is good. Further, in the above description, an example is shown in which the number of sampling taps of the waveform equalizer is four.
Needless to say, the number of sampling taps is not limited to this. Further, in this example, the equalization method is PR (122
Although the reproduction device of 1) is described as an example, the present invention can be applied to an equalization device and a reproduction device that employ other equalization characteristics.

[0044]

As described above, according to the present invention, when equalizing a reproduced signal having asymmetry, not only the value at the detection point shifts but also the variance of the value at the detection point increases. It is disclosed that changing the equalization coefficient in the waveform equalizer depending on the presence or absence of asymmetry is very important in improving the error rate. By changing the equalization coefficient depending on the presence or absence of asymmetry, even when there is asymmetry,
It is possible to output an equalized waveform having a small variance of the voltage distribution at the detection point. Therefore, by performing maximum likelihood decoding (Viterbi decoding) on an output signal whose waveform has been equalized using the waveform equalizer of the present invention, an optical drive such as a DVD or MOD, a magnetic disk drive such as an HDD or a ZIP, etc. In the signal reproducing device, a reproducing device having high decoding ability can be provided.

[Brief description of the drawings]

FIG. 1 is a diagram showing a waveform obtained by equalizing an upwardly convex original waveform having positive asymmetry by a conventional waveform equalizer.

FIG. 2 is a diagram showing a waveform obtained by equalizing a downward convex original waveform having positive asymmetry by a conventional waveform equalizer.

FIG. 3 is a voltage histogram when an original waveform having positive asymmetry is equalized by a conventional waveform equalizer.

FIG. 4 is a diagram showing a waveform obtained by equalizing an upwardly convex original waveform having negative asymmetry by a conventional waveform equalizer.

FIG. 5 is a diagram illustrating a waveform obtained by equalizing a downward convex original waveform having negative asymmetry by a conventional waveform equalizer.

FIG. 6 is a voltage histogram when an original waveform having negative asymmetry is equalized by a conventional waveform equalizer.

FIG. 7 shows that the value at the detection point shifts due to asymmetry,
It is a figure which shows a mode that it disperses typically.

FIG. 8 is a diagram schematically showing that variance at a detection point is reduced when an equalization coefficient is changed in the present invention.

FIG. 9 is a block diagram showing an equalizer according to the embodiment of the present invention.

FIG. 10 is a diagram schematically illustrating a state in which the presence or absence of asymmetry is detected based on a sync signal.

FIG. 11 is a block diagram illustrating an example of an asymmetry detector illustrated in FIG. 9;

12 is a voltage histogram when an original waveform having positive asymmetry is equalized by the waveform equalizer illustrated in FIG. 9;

FIG. 13 is a voltage histogram when an original waveform having negative asymmetry is equalized by the waveform equalizer shown in FIG. 9;

FIG. 14 is a block diagram illustrating a general configuration of a PRML-type playback device.

FIG. 15 is a block diagram showing a configuration of a conventional equalizer.

FIG. 16 is a graph showing an error rate when 8/16 modulation data is reproduced by various PRML methods.

[Explanation of symbols]

 Reference Signs List 1 optical disk device 2 optical disk 3 optical pickup 4 preamplifier 5 waveform shaping circuit 6 A / D conversion circuit 10 system showing PR equalization characteristics 11 equalizer 20 maximum likelihood decoding circuit 16 sampling tap 15 delay element 17 multiplier 18 adder 19 Output unit 31 Asymmetry detector 32 Coefficient selection circuit 33 Memory

Claims (6)

[Claims] 1. A waveform equalizing apparatus for equalizing a waveform of a reproduced signal obtained from a recording medium and outputting the equalized waveform, comprising: detecting means for detecting asymmetry of the reproduced signal; A waveform equalizer having control means for changing characteristics. 2. A method according to claim 1, wherein a plurality of delay elements connected in series to sample the reproduction signal in time series, and a predetermined equalization coefficient is respectively assigned to a plurality of sample values obtained by these delay elements. An output unit that outputs an output signal that has been multiplied and then added and waveform-equalized, wherein the control unit changes the equalization coefficient according to at least the presence or absence of asymmetry. 3. The equalization coefficient according to claim 2, wherein the equalization coefficient selected when there is asymmetry by the control means is:
A waveform equalizer characterized by an equalization coefficient assuming that a wide waveform in both positive and negative directions of asymmetry is equalized to an original waveform of a reproduced signal having no asymmetry. 4. The waveform equalizer according to claim 1, wherein said detecting means determines the presence or absence of asymmetry based on the amplitude intensity of a signal following the sync signal. 5. A waveform equalizer according to claim 1, and maximum likelihood decoding means for maximum likelihood decoding of an output signal output from the waveform equalizer using a plurality of predicted values to reproduce digital data. A playback device having 6. A method of controlling a waveform equalizer for equalizing a waveform of a reproduced signal obtained from a recording medium and outputting the equalized waveform, comprising the steps of: detecting an asymmetry of the reproduced signal; Changing the characteristics of the device.