WO2001099108A1

WO2001099108A1 - Optical disk unit

Info

Publication number: WO2001099108A1
Application number: PCT/JP2001/005368
Authority: WO
Inventors: Youichi Ogura; Shinichiro Sato; Kouichi Urita
Original assignee: Matsushita Electric Industrial Co., Ltd.
Priority date: 2000-06-22
Filing date: 2001-06-22
Publication date: 2001-12-27
Also published as: JP2002008315A

Abstract

An optical disk unit used for reproducing digital data recorded on an optical disk, the unit comprising a transversal filter (6) for performing partial response equalization on a signal reproduced from an optical recording medium (52) where address information-recorded address blocks and data blocks as recordable regions exist, and a by-block filter coefficient learning means (7) for suitably learning a filter coefficient of a transversal filter (6) on a block basis so as to minimize its equalizing error on the basis of an output signal of partial response equalizing, thereby reducing difference in reproducing characteristics among reproduction blocks and improving readability with respect to an optical disk medium.

Description

Description Optical technology

The present invention relates to an optical disk apparatus for reproducing digital data recorded on an optical disk medium, and particularly to a PRML (Partial Response Maximum Likelihood) signal processing method which is effective for high-density recording and reproduction in the linear direction. By improving the digital adaptive automatic equalization means, media with multiple recording areas with different reproduction characteristics, reproduction waveform quality degradation due to tilt, reproduction under poor signal-to-noise ratio conditions, and effects, etc. The present invention relates to a device capable of improving the quality of reproduced digital data and improving playability under adverse conditions such as the occurrence of frequent occurrences. Background art

As a method of recording digital data on an optical disk medium, there are many methods of making the recording density on an optical recording medium uniform by keeping the linear velocity constant as seen in a compact disc (Compact Disc) or a DVD (Digital Versatile Disc). Used. When reproducing digital data from a reproduction signal of an optical disc recorded digitally by mark-width modulation so that the linear recording density is constant. First, a waveform equalization means is applied to the reproduction waveform from the optical disk medium. High-frequency components attenuated by the effect of high-density recording are emphasized, and the equalized waveform that emphasizes the high frequencies is binarized based on the center level of the reproduced waveform, and the clock that the reproduced signal has By detecting the phase of the component and forming a phase-locked loop, phase lock-in was performed and digital data was demodulated.

At this time, in CAV (Constant Angular Velocity) playback, which is a system that keeps the spindle motor rotation speed constant during disc playback, the frequency of the clock component of the playback data changes depending on the radial direction of the disc. As a result, the optimum equalization characteristic of the waveform equalization means also changes. As described above, the jitter component of the equalized waveform was detected, and the equalization characteristic of the waveform equalizing means was adaptively controlled so as to minimize the jitter component.

For example, there is a disk reproducing system as shown in FIG. A digital recording code as shown in FIG. 30 (a) is recorded on the optical disk 52 so that the linear recording density is constant. It is assumed that the recorded data is data regulated such that the number of continuous "0" or "1" is 3 or more and 14 or less, for example, in the 8--16 modulation method. As shown in Fig. 30 (a), the signal obtained by reproduction by the reproducing means 53 such as an optical pickup, as shown in FIG. Since the amplitude decreases as the frequency of the component increases, the signal is amplified by a preamplifier (not shown), and correction is performed by the waveform equalizing means 2 so as to emphasize high frequency components. The waveform equalization means 2 is realized by a high-order ripple filter having a boost characteristic. This is, for example, a 7th-order ripple filter having frequency characteristics as shown in Fig. 31 and having the function of arbitrarily setting the boost amount and high-frequency cutoff frequency according to the characteristics of the reproduced waveform. is there. .

Then, as shown in FIG. 30 (b), the reproduced waveform emphasized by the high-frequency band by the waveform equalizing means 2 is binarized at a predetermined slice level by the binarizing means 54, and ) Convert to a binary digital signal as shown in the figure.

Next, in the cycle detecting means 55, the cycle of a specific pattern included in the digital signal which has been binarized by the binarizing means 54 is counted by the high-frequency clock generated by the oscillator 56. The count information of the specific pattern obtained by the cycle detection means 55 is inversely proportional to the linear velocity, and has the cycle information of the clock component of the playback signal. Control is performed so that the free-running frequency of the clock substantially matches the frequency of the click component of the reproduced signal. Here, the phase locked loop circuit 57 includes a phase comparator 58, a charge pump 59, a loop filter 60, and a voltage controlled oscillator 61. This voltage-controlled oscillator 61 adaptively changes the center frequency based on the frequency information obtained by the period detecting means 55, and the oscillation frequency of the digital signal obtained from the binarizing means 54 is By performing control so that the frequency substantially coincides with the frequency, the phase synchronization after the seek of the reproducing means 53 and the position dependent on the reproducing radius position at the time of CAV reproduction are controlled. Phase synchronization control can be performed. A digital demodulation signal is generated by the demodulation circuit 62 using the binarized signal output from the binarization means 54 and the reproduced clock generated from the phase locked loop circuit 57.

After phase synchronization, the disc reproduction system averages and detects the jitter, which is the phase difference between the binarized signal output from the binarization means 54 and the reproduction turlock controlled by the phase locked loop circuit 57. The waveform equalizer 2 has a function of learning the boost amount and the high-frequency cutoff frequency so that the output signal of the jitter detector 6 3 is minimized. The reproduction characteristics can be improved in response to changes in the data frequency and the high-frequency attenuation characteristic of the reproduction waveform.

Through such a series of operations, waveform equalization and phase-synchronous reproduction can be performed in accordance with the frequency and high-frequency attenuation characteristics of the clock component of the reproduced waveform, and digital data recorded on an optical disk medium can be stably and accurately reproduced. It becomes possible to reproduce.

However, according to the conventional optical disk apparatus described above, an address block and a random data, which are embossed areas in which address information is recorded in advance, such as a DVD-RAM (DVD—random access memory) which is a writable disk, are used. Differences in the reproduction characteristics of writable data blocks, writable optical disks on which digital data is recorded by multiple writing devices with different write characteristics, and the vertical axis and the entrance axis of the laser beam to the recording surface of the optical disk medium If there is deterioration in the reproduction characteristics due to the tilt defined by the angle, or local deterioration in the reproduction characteristics depending on the effects of scratches, dirt, fingerprints, etc. on the disk surface, the boost characteristics described above must be used. Waveform of optically regenerated waveform using only high-order ripple filter In the method of performing reduction, it is impossible to have an optimal high frequency emphasis characteristic for the high-frequency attenuation characteristic of the light reproduced waveform varies.

Also, when the boost amount of the waveform equalization means and the high cutoff frequency are optimized by the jitter detection means, the jitter detection means alone can simultaneously and continuously apply to both the boost amount and the high cutoff frequency. It is difficult to perform automatic control. Therefore, the above-mentioned method cannot sufficiently exhibit the equalization performance, and thus cannot be a useful means for improving the reproduction characteristics.

The present invention has been made in view of such circumstances, and has been made in consideration of a writable disc or the like. In addition to the differences in playback characteristics due to playback position and playback block, local degradation in playback characteristics due to tilt and differential is absorbed, and leadership is improved for optical disc media with various recording and playback characteristics. It is an object of the present invention to provide a possible optical disk device. Disclosure of the invention

An optical disc device according to the present invention includes: an equalization filter that performs partial response equalization on a signal reproduced from an optical recording medium in which a data block that is a recordable area and an address block in which the address information is recorded exist; A block for adaptively learning the filter coefficient of the equalization filter for each block based on the output signal of the partial response equalization so that the equalization error of the equalization filter is minimized. And a separate finoleta coefficient learning means. As a result, in a writable disc or the like having different recording / reproducing characteristics between an address block in which address information is recorded as an embossed area and a data block in which random data can be written, optimal reproduction characteristics can be obtained for each block. Therefore, readability is improved.

Further, in the optical disc apparatus according to the present invention, the block-by-block filter coefficient learning means has a gate signal generating means for generating a gate signal for separating the address block and the data block. Things. As a result, in the data reproduction of a writable disk or the like having different recording / reproduction characteristics between the address block and the data block, the division of the gate signal becomes clear, so that the initial value of the filter coefficient can be loaded for each block. In addition, the hold control that holds the filter coefficient learned for each block and takes over to the next control can be realized, so that the stability and continuity of the digital adaptive automatic equalization control are improved. Also, an optical disc apparatus according to the present invention includes: a preamplifier for enhancing an output amplitude of a signal reproduced from the optical recording medium; a waveform equalizing unit for enhancing a predetermined frequency band of the enhanced signal; Analog-to-digital conversion means for sampling the sampled signal with a reproduction clock, and phase synchronization for controlling the oscillation frequency of the reproduction clock so as to synchronize with the phase of the clock component included in the sampled signal Loop and before Offset correction means for performing a correction to reduce an offset component from the sampled signal, and inputting the corrected signal to the equalization filter, and a type of a partial response equalized by the equalization filter. And a maximum likelihood decoder for performing data demodulation by performing maximum likelihood decoding. This makes it possible to obtain optimum reproduction characteristics for each block in a writable disk or the like having different recording / reproduction characteristics between an address block in which address information is recorded and a data block in which random data can be written. Not only that, a series of signal processing operations improve leadership.

Further, in the optical disc device according to the present invention, the block-by-block filter coefficient learning means includes an envelope detecting means for extracting an envelope of the reproduction signal, and a DC fluctuation of the envelope to detect the address block. And a gate signal generating means for generating a gate signal for identifying the data block. Thus, when a gate signal for identifying an address block and a data block in a writable disk or the like is generated, the gate signal can be read directly from the reproduced signal, so that the generation of the gate signal is stabilized.

Also, the optical disc device according to the present invention, wherein the envelope detecting means has peak holding means for holding a peak value of the waveform of the reproduction signal, and smoothing means for smoothing the held output signal. In this embodiment, the envelope of the reproduced waveform is detected. Thereby, the envelope detection from the reproduced waveform is stabilized, and the learning accuracy of the filter coefficient is improved.

Also, in the optical disc apparatus according to the present invention, the block-by-block filter coefficient learning means includes: a tracking error signal generating means for generating a tracking error signal for a tracking servo; and detecting a DC variation of the tracking error signal. Thus, there is provided a gate signal generating means for generating a gate signal for identifying the address block and the data block. Thus, when generating a gate signal for distinguishing between an address block and a data block in a writable disc or the like, the address block can be further distinguished into the first half and the second half based on the polarity of the tracking error signal. Since it is possible, it is possible to delicately control the digital adaptive automatic equalization in the address block. as a result, Since the reading accuracy of the address information is improved in both the first half and the second half, the address information can be read stably immediately after the seek, which leads to an improvement in seek performance.

Further, in the optical disc apparatus according to the present invention, the block-specific filter coefficient learning means detects an address position of the address block, and reproduces in synchronization with a phase of a clock component included in a reproduction signal output from the phase locked loop. A counter for counting the number of recording data by a clock; and a gate for identifying the address block and the data block by estimating a position where an address block appears next based on the value of the counter. And a gate signal generating means for generating a signal. Thus, when a gate signal for identifying an address block and a data block in a writable disk or the like is generated, an accurate position of the next address block can be estimated by a playback clock, so that the diff. The timing of the gate signal can be accurately controlled without being affected by the disturbance of the tracking servo, and the equalization accuracy and stability of digital adaptive automatic equalization control for each block are improved.

Also, in the optical disc device according to the present invention, the block-by-block filter coefficient learning means includes: a gut signal generating means for identifying the address block and the data block; and a single frequency pattern existing in a reproduction signal. A single pattern detecting means for detecting, and when learning the filter coefficients by classifying each block, in each block according to an output signal of the single pattern detecting means. The learning of the filter coefficients is performed at random pattern positions other than the single frequency pattern. Thus, when generating a gate signal for distinguishing an address block from a data block in a writable disc or the like, it avoids learning in a single frequency pattern which is disadvantageous for learning a finoleta coefficient. Therefore, the control speed and equalization accuracy of digital adaptive automatic equalization control are improved.

Further, in the optical disc apparatus according to the present invention, the block-by-block filter coefficient learning means detects an address position of the address block, and detects a phase of a click component included in a reproduction signal output from the phase locked loop. The playback clock synchronized with A counter for counting the number of recording data, and a gate signal for identifying the address block and the data block is generated by estimating a position where the next address block appears based on the value of the counter. And a pattern identification gate signal generating means for separating the positions of the single frequency pattern and the random pattern based on the value of the counter, and learning the filter coefficient by separating each block. At this time, the filter coefficients are learned at random pattern positions other than the single frequency pattern in each block according to the pattern identification gate signal. As a result, when generating a gate signal for discriminating between an address block and a data block in a writable disk or the like, an accurate position of a single frequency pattern that is disadvantageous for learning of a filter coefficient is estimated by a reproduction clock. As a result, the probability of learning the filter coefficients using only random patterns increases, so that the control speed and equalization accuracy of digital adaptive automatic equalization control can be further improved.

Also, in the optical disc device according to the present invention, the block-by-block filter coefficient learning means may select a position where a random pattern including address information in the address block and a single frequency pattern are alternately present. There is provided a learning position control means for determining whether to perform the learning of the filter coefficient, and the learning of the filter coefficient is performed based on the output signal of the learning position control means. As a result, when reproducing a writable disk or the like, even if the reproduction characteristics and the sampling phase change due to the servo control conditions and the phase synchronization pull-in characteristic of the phase locked loop, etc. Since filter coefficients can be learned using only data that can be reproduced stably and satisfactorily, it is possible to learn filter coefficients that are suitable for the servo control state and phase synchronization performance. Address information can be read stably.

Also, an optical disc device according to the present invention includes a maximum likelihood decoder that performs data demodulation by performing maximum likelihood decoding according to a type of a partial response on a signal obtained by equalization with the equalization filter; EDC determination means for detecting EDC (Error Detection Code) for determining correctness of address information from a demodulated signal demodulated into binary data by the maximum likelihood decoder, and a result of EDC determination obtained from the EDC determination means And a filter coefficient learning control means for controlling the block-specific filter coefficient learning means so that the learning of the finoleta coefficient for the address block corresponding to the address information is stabilized. As a result, in the case of reproduction of a writable disc or the like, if the reproduction characteristics are degraded due to a defect or the like, the learning of the filter coefficient at that position in the address block is stopped, and only those at the position where the reproduction can be performed well are performed. Since the filter coefficient can be learned by using the filter coefficient, the learning of the filter coefficient in the address block is stable in the learning of the filter coefficient in the abnormal state, and the address information can be read stably.

Further, in the optical disc device according to the present invention, the block-by-block filter coefficient learning means, when correct address information is obtained by the determination by the EDC determination means, adaptively learns filter coefficients in the address block. When the correct address information cannot be obtained, the learning of the filter coefficient is stopped and the filter coefficient is reset to the initial value. As a result, in the case of reproduction of a writable disc or the like, if a reproduction abnormality occurs due to a defect or the like, it is possible to recover quickly, and thus it is possible to shorten a recovery time from the abnormality control.

Further, in the optical disc device according to the present invention, when the block-by-block filter coefficient learning means performs initial learning for determining an initial value of a filter coefficient of the address block, the determination by the EDC determination means is continuously correct. An EDC counter that counts the number of address information that has been obtained, and a filter coefficient when the count value of the EDC counter indicates that the address information is correct for more than an arbitrary number is stored as the initial value of the filter coefficient. And initial value storage means for loading the information of the initial value storage means at the start of the learning of the filter coefficients. As a result, in the reproduction of a writable disc, etc., it is confirmed that the reproduction is normally performed, and the learning result of the filter coefficient is stored as the initial value. Reliability is improved.

Further, the optical disc device according to the present invention includes an equalizing filter for performing partial response equalization when reproducing digital data from an optical recording medium, and an equalizing filter based on the equalized output signal. Set the equalization filter so that the equalization error is minimized. Filter coefficient learning means for adaptively learning the filter coefficient of the filter, and equalization error averaging means for averaging the absolute value of the equalization error amount at each level of the partial response equalization level in the equalized signal An equalization performance detecting means for detecting equalization performance based on an output of the equalization error averaging means; a self-determining filter coefficient learning control means for controlling a learning function of a filter coefficient of the equalization filter; It is intended to be equipped with. As a result, based on the equalization error information, the current digital equalization state is self-analyzed, and adaptively digitally responds to the difference in playback characteristics depending on the playback position, which appears due to a writable disc or tilt, etc. Since the automatic equalization control can be performed, the readability of an optical disk medium having various recording / reproducing characteristics is improved.

Also, the optical disk device according to the present invention is a preamplifier for enhancing an output amplitude of a reproduction signal from the optical recording medium, and a waveform equalizing means for enhancing a predetermined frequency band of the signal in which the output screw width is enhanced. Analog-to-digital conversion means for sampling the waveform-equalized signal into digital data using a reproduced clock; and controlling the oscillation frequency of the reproduced clock so as to synchronize with the phase of the clock component of the sampled signal. Phase-locked loop, digital data correction means for correcting offset components and amplitude from the sampled signal, and inputting the corrected data to the equalization filter, and equalization obtained from the equalization filter Equalization error averaging means for averaging the absolute value of the equalization error amount at each level of the partial response equalization level in the output signal; and An equalization performance detection unit that detects equalization performance based on an output of the equalization error averaging unit; a self-determination filter coefficient learning control unit that controls a learning function of a filter coefficient of the equalization filter; A maximum likelihood decoder that performs data demodulation from the equalized output signal according to the type of the partial response applied in the equalizing filter is provided. This enables the current digital equalization state to be self-analyzed based on the equalization error information, and adaptively adjusts the digital automatic response in response to the difference in playback characteristics depending on the playback position, which appears due to a writable disc or tilt. Not only is it possible to perform read / write control, but also by a series of signal processing operations, the leadership of optical disc media having various recording and reproduction characteristics is improved.

Also, in the optical disc device according to the present invention, the equalization error averaging means may include a predetermined chip. Channel bits (minimum unit of recording data) During the N (N is a positive integer) period, add the absolute value of the equalization error amount at each level of the partial response equalization level in the equalization filter, and set N as the mother If the equalization error output result averaged by the equalization error averaging means is smaller than a predetermined level X (X is a positive integer), an equalization state is obtained. Is determined to be normal, and when the level is equal to or higher than the level X, a determination signal indicating that the equalization state is abnormal is generated. This improves the accuracy of detecting the digital equalization performance, stabilizes the automatic equalization performance, and easily realizes any reproduction performance without operating the circuit for learning the filter coefficients more than necessary. Power consumption can be reduced.

Also, in the optical disc device according to the present invention, the equalization error averaging means may include: a partial response in the equalization filter during a predetermined number of channel bits (minimum unit of recording data) (N is a positive integer); The absolute value of the equalization error amount is added for each level of the equalization level, and division is performed with N as a parameter.The equalization performance detection means performs the equalization of the specified partial response equalization level. If the equalized error output result averaged by the error averaging means is smaller than a predetermined level X (X is a positive integer), a judgment signal indicating that the equalization state is normal is generated, and the predetermined level X (X Is a positive integer). If the value is equal to or greater than the above, a determination signal that the equalization state is abnormal is generated.

As a result, digital adaptive automatic emphasis, with emphasis on reproduction performance, is achieved by detecting digital equalization performance only for the partial response equalization level that is significantly related to the performance of PRML signal processing, rather than the overall equalization performance. This makes it possible to improve leadership, especially when the reproduced waveform has asymmetry characteristics.

Also, in the optical disc device according to the present invention, the self-determination type filter coefficient learning control means performs an initial learning for determining an initial value of a filter coefficient of the equalization filter, the output signal of the equalization performance detection means. Has an initial value storage means for storing the result of the filter coefficient learning when the equalization state is normal as the initial value of the filter coefficient. Load the value of the storage means and Control is performed with the filter coefficient fixed. If the output signal of the equalization performance detection means indicates that the equalization state is abnormal during the fixed control, learning of the filter coefficient is started. Learning is continued until the output signal of the equalization performance detection means indicates that the equalization state is normal, and the filter coefficient at the time when the equalization state becomes normal is stored in the initial value storage means. The fixed control is performed. As a result, the reliability of the initial value is improved in the initial value learning of the filter coefficient, and digital equalization is usually performed by fixed control, and the filter coefficient is relearned only when the equalization state deteriorates. , It is possible to realize the required minimum power consumption. Also, in the optical disc device according to the present invention, the self-determination type filter coefficient learning control means performs an initial learning for determining an initial value of a filter coefficient of the equalization filter, the output signal of the equalization performance detection means. Has initial value storage means for storing the result of the filter coefficient learning when the equalization state is normal as the initial value of the filter coefficient. The adaptive automatic control is performed by loading the value of the storage means and learning the continuous filter coefficient. During the adaptive automatic control, the output signal of the equalization performance detecting means determines that the equalization state is abnormal. In the case shown, learning of the filter coefficient is stopped, the value of the initial value storage means is loaded, and then adaptive automatic control is performed again by continuous learning of the filter coefficient. As a result, when the control is disturbed by self-healing due to, for example, digital adaptive automatic equalization, the equalization state is judged to be abnormal by the equalization performance detection means. It has the self-healing ability from the abnormal state of resetting to the value and restoring to the normal control state.

Also, in the optical disc device according to the present invention, the self-determination type filter coefficient learning control means may include a plurality of optical discs as optical recording media in a radial direction when determining an initial value of a filter coefficient of the equalization filter. It has a zone-specific initial value storage means for performing initial learning by dividing the data into zones and storing each learning result as an initial value of the filter coefficient. It loads the value of the initial value storage means for each zone and performs adaptive automatic control. As a result, when the setting resolution of the high-frequency cutoff frequency of analog waveform equalization for the change in the frequency of the clock component of the reproduced data that depends on the radial direction during CAV reproduction is coarse. However, by keeping the digital equalization filter coefficients divided into zones, the read performance immediately after the seek is improved.

Further, in the optical disc apparatus according to the present invention, when the self-determination type filter coefficient learning control means determines an initial value of a filter coefficient of the equalization filter, the self-determination type filter coefficient learning control means moves the optical disc as an optical recording medium in a radial direction and a circumferential direction. It has a zone-specific initial value storage means for performing initial learning by dividing it into a plurality of zones in the direction and storing each learning result as an initial value of a filter coefficient. The value of the initial value storage means for each zone is loaded, and adaptive automatic control is performed. As a result, not only the difference in the reproduction characteristics depending on the reproduction position and the reproduction block but also the local reproduction characteristics due to the tilt and the difference appear on a writable disk or the like on which digital data is recorded by a plurality of writing devices having different write characteristics. By absorbing the deterioration, it is possible to improve the reproducibility of an optical disk medium having various recording / reproducing characteristics.

Further, in the optical disc device according to the present invention, the self-determination type filter coefficient learning control means may perform, during the initial learning of the filter coefficient, the equalization of the plurality of zones on the circular optical disc as an optical recording medium. For a zone in which the output signal of the performance detection means indicates that the equalization state is particularly poor, the apparatus has a poor zone storage means for recording the zone, and stores the storage result of the bad zone storage means. Based on this, the control gain in filter coefficient learning of adaptive automatic control is adjusted for each zone to be reproduced. This enables stable control and improvement of leadership by switching the equalization control gain and the control method in advance in the zone where the local reproduction characteristic deterioration due to the effect is confirmed. S becomes possible. BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 is a block diagram showing a configuration of an optical disc device according to Embodiment 1 of the present invention.

FIG. 2 is a block diagram showing a configuration of an offset correction unit according to the first embodiment. FIG. 3 is an explanatory diagram of a difference between a PR (a, b, b, a) equalization method realized by a transversal filter and a general binarization determination method according to the first embodiment. .

FIG. 4 is a diagram showing frequency characteristics of various partial response systems realized by the transversal filter in the first embodiment.

FIG. 5 is an explanatory diagram of a recording format of DVD-RAM in the first embodiment.

FIG. 6 is a block diagram showing a configuration of the transversal filter according to the first embodiment.

FIG. 7 is a block diagram showing a configuration of a block-by-block filter coefficient learning means according to the first embodiment.

FIG. 8 is an explanatory diagram of the principle of a Viterbi decoder which is one of the maximum likelihood decoders according to the first embodiment.

FIG. 9 is a block diagram showing a configuration of a gate signal generating means according to the first embodiment.

FIG. 10 is an explanatory diagram for explaining a principle of generating a gate signal in the first embodiment.

FIG. 11 is a block diagram illustrating a principle of generating a gate signal according to the first embodiment.

FIG. 12 is a block diagram showing a configuration of a gate signal generating means according to the first embodiment.

FIG. 13 is a block diagram showing a configuration of a gate signal generating means according to the first embodiment.

FIG. 14 is a block diagram showing a configuration of an optical disk device according to Embodiment 2 of the present invention. '

FIG. 15 is a diagram showing a configuration of an ID block in a recording format of DVD-RAM according to the second embodiment.

FIG. 16 is a diagram showing a flowchart of a control method of a block-by-block filter coefficient learning means according to the second embodiment. FIG. 17 is a block diagram showing a configuration of a block-by-block filter coefficient learning means according to the second embodiment.

FIG. 18 is a block diagram showing a configuration of an optical disc device according to Embodiment 3 of the present invention.

FIG. 19 is a block diagram showing a configuration of a self-determination type filter coefficient learning control means according to the third embodiment.

FIG. 20 is a block diagram showing the configuration of an equalization performance self-determination block in the self-determination filter coefficient learning control means according to the third embodiment.

FIG. 21 is an explanatory diagram of the operation principle of the equalization performance self-determination block in the self-determination filter coefficient learning control means according to the third embodiment.

FIG. 22 is a diagram showing a flowchart of a control method of the self-determination filter coefficient learning control means in the third embodiment.

FIG. 23 is a diagram showing a flowchart of a control method of the self-determination filter coefficient learning control means in the third embodiment.

FIG. 24 is a block diagram showing a configuration of an optical disc device according to Embodiment 4 of the present invention.

FIG. 25 is an explanatory diagram of a zone division method in the zone-based filter coefficient learning means according to the fourth embodiment.

FIG. 26 is a block diagram showing a configuration of zone-based filter coefficient learning means according to the fourth embodiment.

FIG. 27 is an explanatory diagram of a zone division method in the zone-based filter coefficient learning means according to the fourth embodiment.

FIG. 28 is a diagram showing a flowchart of a control method of the zone-based filter coefficient learning means in the fourth embodiment.

FIG. 29 is a block diagram showing a configuration of a conventional optical disk device.

FIG. 30 shows the recording data of the conventional optical disk device and the output signal waveforms in each functional block.

FIG. 31 is an explanatory diagram of a frequency characteristic of a high-order ripple filter. BEST MODE FOR CARRYING OUT THE INVENTION

Hereinafter, before describing the embodiments of the present invention, the outline of the present invention will be described first.

The present invention applies a PRML signal processing method, which is advantageous for high-density recording in the linear direction, when reproducing digital data recorded on an optical disk medium, and realizes a response response suitable for reproduction characteristics. The combination of equalization and maximum likelihood decoding effective for white noise, which estimates a likely sequence using the correlation of partial response intersymbol interference, aims to improve digital data reproduction characteristics, By applying adaptive automatic type filter coefficient learning means that can always learn to minimize the equalization error due to partial response equalization as the shape equalization means and optimizing the initial filter coefficient learning and control means, Not only differences in playback characteristics depending on the playback position and playback block that appear on writable discs, etc., but also stations due to tilt and effect By absorbing the deterioration of the regeneration characteristics, improved Ridapiriti becomes possible for the optical disc medium to have a variety of recording and reproducing characteristics, it is intended to'll provide an optical disc equipment.

(Embodiment 1)

In the first embodiment, for example, an optical disc having an embossed area called CAPA in which address information is written in advance, and an area having different characteristics, such as a data section in which actual data is written, is used in a DVD-RAM disc. This makes it possible to solve the problem that sufficient reproduction performance cannot be exhibited when equalization and reproduction are performed by the same waveform equalization means.

In FIG. 1, after the output amplitude of the optical disk reproduction signal reproduced by the reproduction means 53 is emphasized by the bias 1, correction is made by the waveform equalization means 2 to emphasize the high frequency range. The waveform equalizing means 2 is composed of a finoletor capable of arbitrarily setting a boost amount and a cutoff frequency. This is, for example, a higher-order equal ripple filter having a frequency characteristic as shown in FIG. The output signal of the waveform equalizing means 2 is sampled into a multi-bit digital signal by an analog-to-digital converter 3 which converts an analog signal into a digital signal using a reproduced clock generated by the oscillator 10. This sampled multi-bit digital signal is input to the offset correction means 4. Thus, the offset component included in the reproduced digital signal is corrected.

The offset correction means 4 may have a configuration as shown in FIG. 2, for example. This includes an offset detecting means 13 for detecting an offset component of the reproduced digital signal, a smoothing means 14 for smoothing the offset signal detected thereby, and an output signal of the smoothing means 14 for converting the output signal of the reproduced digital signal to a reproduced digital signal. It comprises a subtraction means 15 for further subtracting.

Then, the output signal of the offset correction means 4 is input to the auto gain controller 5 to adjust the amplitude of the reproduced digital signal to an arbitrary value. The auto gain control 5 may be, for example, a type that detects an envelope of a signal waveform and controls the difference between an arbitrary set value and an envelope signal to be zero. Next, the output signal of the auto gain control 5 is input to the transversal finole letter 6 to perform partial response equalization. Here, the partial response equalization is performed, for example, in a DVD-RAM capable of digital recording of 4.7 Gbytes on one layer on one side, as shown in FIG. 3 (c), the waveform amplitude after the equalization is The PR (a, b, b, a) method is used so that it can be divided into five values.

Conventionally, in a read channel of an optical disc reproducing system, digital data demodulation is performed by binarization discrimination using a slice level (center level) from a waveform equalized output signal as shown in FIG. 3 (a). I was Also in the case of sampling, binarization was performed by sampling as shown in FIG. 3 (b) and discriminating the multi-bit digital signal based on the slice level.

On the other hand, the PR (a, b, b, a) method is different from the PR (a, b, b, a) method in that four different sampled data are added at the ratio of a: b: b: a. b * D ² + a * D ³ ), and adds the characteristics of a low-pass filter as shown in FIG. 4 to the reproduced signal.

In Fig. 4, the MTF indicates the optical reproduction characteristics of DVD-RAM (4.7 Gbytes). The closer this frequency characteristic is, the more advantageous the partial response method is. In addition to the method shown in Fig. 4, there are various types of partial response other than the PR (a, b, b, a) method, but not only for a specific method, but also for performance. If something is possible, use another method There is no problem with using it. This is one of the partial response method that adds a correlation in the time direction of these reproduced data and one of the maximum likelihood decoding methods (maximum likelihood decoding) described later. By combining a Viterbi decoder that demodulates the data, a PRML (Partial Response Maximum Likelihood) signal processing that is advantageous for high recording density reproduction in the linear recording direction is realized.

As described above, the PMLL signal processing method has various combinations depending on the characteristics of the reproduction waveform and the modulation code. Therefore, it is necessary to select an appropriate method for various recording and reproduction systems.

In addition, as shown in Fig. 5, the DVD-RAM recording format is such that an address block AB in which address information of a sector SC is recorded for each sector as an emboss area, and random data can be rewritten. It is divided into a possible data block DB. The VFO in FIG. 5 is an area in which 4 T patterns (T is the minimum recording unit time) are continuously recorded, and is used for phase synchronization pull-in and the like. AM is an address mark, and includes a synchronization pattern represented by 14 T + 4 T in DVD. The ID is an area where the address information of the sector is recorded. There are four locations in the address block (ID 1 to ID 4) and one location in the data block. The PS in the data block is called a pre-sink and contains a synchronization pattern. DATA is a data recording area that can be freely rewritten by the user. Here, the characteristics of the reproduced waveform of the data block often vary depending on the characteristics of the device used for writing. Therefore, the reproduction characteristics of the address block and the data block often differ. For example, the transversal filter 6 in FIG. 1 may use a FIR filter (Finite Impulse response Filter) composed of finite taps. The equalization characteristics of the FIR filter are realized by changing the filter coefficients. The FIR filter includes a delay element 16a or a delay element 16f connected in series for sequentially delaying an input signal as shown in FIG. 6, and a multiplication element 17 for multiplying the input signal by a filter coefficient S1. a and the outputs of delay elements 16 a to 16 f and the multiplication elements 17 b to 17 g for multiplying the filter coefficients S 2 to S 7 with the outputs of multiplication elements 17 a to 17 g And the addition means ¹⁸ It may be something. The filter coefficients S 1 to S 7 of the FIR filter are set by the block-specific filter coefficient learning means 7. The block-specific filter coefficient learning means 7 uses an LMS (Least Mean Square) algorithm that adaptively controls the equalization error present in the partial response equalization output signal by the transversal filter 6 to be minimized. It has a function to learn the filter coefficient, and separates the address block and the data block by the gate signal, and after the separation, learns the filter coefficient of the filter for each block in parallel. The block-specific filter coefficient learning means 7 may have a configuration as shown in FIG. 7, for example.

This is because the provisional decision circuit 19 detects an equalization target value corresponding to the partial response method from the equalization output signal of the transversal filter 6, and converts the output signal of the transversal filter 6 from the equalization target value. An equalization error detection means 20 for detecting an equalization error by subtraction, a correlator 21 for calculating a correlation between an output signal of the equalization error detection means 20 and an input signal of the transversal filter 6, A feedback gain adjuster 22 as gain adjusting means for adjusting the feedback gain by multiplying the output of the compensator 21 by a gain, and a means for adding the output to the filter coefficient of each tap and updating the filter coefficient Filter coefficient updating unit 23a to 23g, and a filter unit generated by switching between the initial value of the address block filter coefficient and the initial value of the data block filter coefficient. Generator 2 4 a through is formed using a 2 4 g.

The filter coefficient generators 24a to 24g are address block filter coefficient initial value storage means 2411a to 2411 for holding address and data block learning results for each address block and data block. g, data block filter coefficient initial value storage means 24 2 a to 24 2 g, address block filter coefficient initial value storage means 24 1 a to 24 1 g output, and data block A selector block for selecting the output of the filter coefficient initial value storage means 242 a to 242 g, and an address block generated by the gate signal generation means 266. It has a function of selecting an initial value of an address block filter coefficient and an initial value of a data block filter coefficient in accordance with a signal for identifying the data block and the data block. Also, the filter coefficient update units 23a to 23g are adders 2311a to 2311 that add the output of the feedback gain adjuster 22 and the output of the delay elements 2332a to 2332g. g and a selector 2 3 3 a to 2 3 3 g to select the output of the 2 3 3 a to 2 4 3 g and an output of the adder 2 3 1 a to 2 3 1 g according to the load signal, and a selector 2 3 It has a delay element 2 32 a to 23 2 g that delays the output of 3 a to 23 3 g by a unit time, and uses the initial value of the filter coefficient loaded for each block to automatically adjust the filter coefficient. It has a function of performing equalization control.

On the other hand, as shown in FIG. 1, the optical disc device according to the first embodiment has a phase comparator 8 for detecting a phase error from an output signal of the offset correction means 4, and a phase output from the phase comparator 8. It has a loop filter 9 for smoothing the error signal, and an oscillator 10 for oscillating the reproduction clock based on the output signal of the loop filter 9. The phase of the reproduction clock output from the oscillator 10 is A phase-locked loop 11 that controls the phase of the clock component of the reproduced signal so that it is synchronized with the phase of the reproduced signal is provided.The analog-to-digital converter 3 uses the reproduced clock output from the oscillator 10 to sample the reproduced waveform. By performing the conversion, a multi-bit sampled signal synchronized with the phase of the clock component of the reproduced signal is generated, so that it is possible to realize the PRML signal processing.

As described above, data demodulation is performed through the maximum likelihood decoder 12 that performs decoding according to the type of the partial response using the partial response equalized signal output by a series of operations. Here, the maximum likelihood decoder 12 may be, for example, a Viterbi decoder. The Viterbi decoder calculates probabilities in accordance with the law of correlation of intentionally added codes, which depends on the type of partial response, and reproduces a likely sequence. For example, when the applied partial response type is the PR (a, b, b, a) method, the state changes based on the state transition diagram as shown in FIG. 8 (a). This takes into account, in particular, the 8-16 modulation code used in DVDs, and is related to the fact that the run length is limited to "2". It can be expressed by the six state transitions. In Fig. 8 (a), X / Y indicates the transition of the recording code, and Y indicates the signal amplitude at that time. You. One state is represented by three different time codes. For example, in the state transition from S 4 “1 1 0” to S 3 “100”, the code “0” is added to “1 10” to the left. Means that the leftmost "1" disappears and the state S3 becomes "100".

The change over time is represented by a trellis diagram as shown in Fig. 8 (b). Therefore, the probabilistic length of each path 1 _k ^ab (hereinafter referred to as the branch metric) is calculated, and the branch metric is added when transitioning to each state. Here, k represents a temporal transition, and ab represents a branch metric at the transition from the state Sa to Sb. The added value in each state of the branch metric is called a metric, and the path with the minimum metric is used as a surviving path and sequentially output to be demodulated into binary digital data. is there. That is, if demodulation is performed according to the recording code in FIG. 8 (b), the path indicated by the solid line is the surviving path.

As can be seen on a writable disc or the like having different recording and reproduction characteristics between an address block as an embossed area in which address information is recorded and a data block to which random data can be written, blocks having different reproduction characteristics are used. By using an optical disc device that has a means for learning the filter coefficient in partial response equalization, it is possible to obtain the most appropriate playback characteristics for each block, and to clarify the positional separation between the two by the gate signal. Therefore, not only can the initial value of the filter coefficient be loaded for each block, but also the hold control that retains the filter coefficient learned for each block and takes over to the next control can be realized. Improves stability and continuity of adaptive automatic equalization control, Above is possible.

The gate signal generating means 26 (see FIG. 7) in the block-specific filter coefficient learning means 7 shown in FIG. 1 may have, for example, a configuration as shown in FIG. For example, on the recording surface of a rewritable optical disk, the address block and the data block force exist spirally as shown in FIG. ₁₀ ( _a ), and the address block A and the address block B force the data block. Address block and data block if they are shifted by half in the radial direction, respectively. As shown in Fig. 10 (b), at the switching position of, the DC (DC component) level of the output waveforms corresponding to the address splock and the data block differs greatly, as shown in Fig. 10 (b). Utilizing this characteristic, as shown in FIG. 9, a peak hold means 27 and a low-pass filter 28 for smoothing the output signal of this reproduced waveform are used to obtain the waveform shown in FIG. 10 (c). The envelope of such a reproduced waveform is extracted, an arbitrary level is set, and the obtained envelope signal is binarized by the binarizing means 29, as shown in FIG. 10 (d). This is to generate a gate signal for identifying a block and a data block.

By using such gate generation means, it is possible to directly read the positional relationship between the address block and the data block from the reproduced waveform, leading to stabilization of the generation of the gate signal. The accuracy of coefficient learning is improved.

The gate signal generation means 26 (see FIG. 7) in the block-by-block filter coefficient learning means 7 shown in FIG. 1 may be based on the principle shown in FIG. 11, for example. For example, on the recording surface of a rewritable optical disk, address blocks and data blocks exist spirally as shown in FIG. 10 (a), and address blocks A and address blocks B become data blocks. On the other hand, if each shifts by half in the radial direction, it is added to the four-segment photodetector 530, which separates into A, B, C, and D areas and detects reflected light as shown in Fig. 11 (a). 535, 536 and subtracter 537 detect the difference between the sum signal (A + C) of the diagonal output of photodetectors 531, 533 and the sum signal (B + D) of the outputs of photodetectors 532, 534. When the tracking error signal is detected by using the phase difference method, the tracking error signal at the position where the address block and the data block are switched becomes as shown in FIG. 11 (b). In A and address block B, it becomes opposite polarity, the further data block, in a state where the tiger Tsu King servo is operating normally, with a feature that the zero level near. By using this characteristic, two decision levels P and N are provided for the tracking error signal shown in Fig. 11 (b), and the two decision levels are binarized. Equivalent to block A and address block B In addition, a gate signal as shown in FIG. 11 (c) can be generated. By using these gate signals, two types of address block filter coefficient initial value storage means shown in FIG. 7 can be used (24 1 a to 24 1 g and 24 22 a to 24 2 g). g) It is possible to store and load the filter initial values of the first half address block A and the second half address block B respectively. By using the gate generation means capable of distinguishing the address block into the first half and the second half by using the polarity of the tracking error signal, the first half of the address block due to the servo state or the like is used. It is possible to control the digital adaptive automatic equalization in detail for the difference in amplitude and characteristics between the first half and the second half. Address information can be read stably, leading to an improvement in seek performance. The gut signal generation means 26 (see FIG. 7) in the block-by-block filter coefficient learning means 7 shown in FIG. 1 may have, for example, the configuration shown in FIG. This is because the address position detecting means 30 for detecting the address position based on the demodulated signal of the maximum likelihood decoder 12 and the phase of the clock component of the reproduced signal are synchronized with the detected address position as a reference. A sector counter 31 that counts a section corresponding to one sector by using a playback clock, and a predicted gate generation means 32 that generates a gate at the position of the next end address information predicted by the sector counter. , For more accurate identification of address blocks and data blocks.

By using such a gate generation means, the accurate position of the next address block can be estimated from the reproduced clock without being affected by the disturbance of the differential or the track servo, so the timing of the gate signal can be accurately controlled. Therefore, the equalization accuracy and stability of digital adaptive automatic equalization control for each block are improved. Further, the gate signal generating means 26 (see FIG. 7) in the block-specific filter coefficient learning means 7 shown in FIG. 1 may have, for example, a configuration as shown in FIG. This means that an address position detecting means 30 for detecting an address position based on the demodulated signal of the maximum likelihood decoder 12 and a phase of a clock component included in the reproduced signal with respect to the detected address position are used as a reference. Using a synchronized playback clock, a sector counter 31 that counts a section corresponding to one sector and a sector counter 31 that is predicted Predicted gate generating means 32 for generating a gate at the position of the next address information, and single pattern detecting means 33 for detecting a single frequency pattern included in the reproduced signal. And learning stop signal generating means 34 for stopping the learning of the filter coefficient when the detected time is equal to or more than an arbitrary section N (N is a positive integer).

By providing such a function that avoids learning in a single frequency pattern that is disadvantageous for learning filter coefficients, the efficiency of learning filter coefficients is improved, and the control speed of digital adaptive automatic equalization control is improved. Accuracy is improved.

Further, the gate signal generation means 26 in the block-by-block finole coefficient coefficient learning means 7 according to the first embodiment may have the following configuration, for example. This is based on the demodulated signal of the maximum likelihood decoder 12, the address position detecting means 30 for detecting the address position shown in FIG. 12 and the reproduced signal based on the detected address position. Using a recovered clock synchronized with the phase of the clock component, a sector counter 31 that counts the interval corresponding to one sector, and a gate generated at the next address information position predicted by the sector counter 31 In accordance with the DVD-RAM recording format shown in FIG. 5, a random gate position other than a single frequency pattern, such as AM (address mark) and ID (key), is provided. Dress information) and the position of a random data area are estimated by the sector counter, and a learning permission signal for permitting learning of filter coefficients is generated. Here, a 4T pattern (T is the minimum recording unit) is usually continuously written in VFO or the like in FIG.

By avoiding such a learning with a single frequency pattern that is disadvantageous for the learning of filter coefficients, the random pattern position can be accurately determined based on the reproduction clock synchronized with the phase of the click component of the reproduction signal. By providing a function to learn the filter coefficients in the filter, the stability is improved in addition to the control speed and equalization accuracy of the digital adaptive automatic equalization control.

As for the learning permission signal, when there are a plurality of ID portions having address information as in the address block shown in FIG. 5, the learning of the filter coefficient is performed for the ID of any position. Learning position control means to determine It may have a function of generating a learning permission signal based on an output signal of the position control means and learning a filter coefficient. Thereby, for example, the learning position control means can learn the filter coefficient of the address block by giving a learning permission signal to ID3 and ID4 shown in FIG.

By providing such a function of learning the filter coefficient by limiting the position, the reproduction characteristics and the sampling phase can be adjusted according to the servo control conditions and the phase synchronization pull-in characteristics of the phase-locked loop. Even if the value fluctuates, the filter coefficient can be learned using only the address data that can be reproduced stably and satisfactorily from a plurality of address data existing in the address block, so it is suitable for the servo control state and phase synchronization performance. The learning of the filtered filter coefficients becomes possible, and there is an effect that the address information can be read stably.

As described above, according to the first embodiment, the optical disk device for reproducing the DVD-RAM disk is provided with the digital filter based on the PRML signal processing method, and the signal is equalized by the digital FIR filter at the subsequent stage. In this process, the emboss area (CAP A area) of the DVD-RAM disc and the data area are separately subjected to adaptive automatic learning of FIR filter coefficients, and optimal equalization is performed for each. Since the MS (least squares method) algorithm is used, it is possible to improve the error rate characteristics of the CAP A area and the data portion, and to improve the reproduction characteristics of digital data.

Also, according to the first embodiment, the emboss area (CAPA area) and the data part are separated by the gate signal, and after the separation, the adaptive automatic learning of the FIR filter coefficients is performed in parallel. Learning can be speeded up.

Furthermore, according to the first embodiment, when learning the filter coefficients in the CAP A region, random data having a high learning efficiency is recorded by avoiding the VFO section which is constituted by a single frequency and hinders the learning effect. Since the learning is performed using the ID part, the learning efficiency can be improved and the learning period can be shortened.

(Embodiment 2)

In the second embodiment, for example, in a DVD-RAM disk, adaptive automatic learning of FIR filter coefficients is performed separately for the CAP A area and the data section, and When performing the optimal equalization, it is possible to solve the problems caused by the fact that the VFO (single frequency) occupies a large proportion in the CAPA region and it is difficult to derive the optimal equalization characteristics.

In FIG. 14, after the output amplitude of the optical disc reproduction signal is emphasized by the preamplifier 1, correction is made by the waveform equalizing means 2 to emphasize the high frequency range. The waveform equalizing means 2 is composed of a filter capable of arbitrarily setting a boost amount and a cutoff frequency. For example, it is a high-order equiripple filter having frequency characteristics as shown in FIG. An output signal of the waveform equalizing means 2 is sampled into a multi-bit digital signal by an analog / digital converter 3 which converts an analog signal into a digital signal using a reproduced clock generated by the oscillator 10. By inputting the sampled multi-bit digital signal to the offset correcting means 4, the offset component included in the reproduced digital signal is corrected.

The offset correction means 4 may have a configuration as shown in FIG. 2, for example. This is because the output signal of the offset detection means 13 for detecting the offset component of the reproduced digital signal, the smoothing means 14 for smoothing the offset signal detected thereby, and the output signal of the smoothing means 14 It comprises a subtraction means 15 for subtracting from a reproduced digital signal.

The output signal of the offset correction means 4 is input to the auto gain control 5, whereby the amplitude of the reproduced digital signal is adjusted to an arbitrary value. For example, the auto gain control 5 may be one that detects the envelope of the signal waveform and controls the difference between an arbitrary set value and the envelope signal to be zero.

Next, the output signal of the auto gain control 5 is input to the transversal filter 6, and partial response equalization is performed. Here, partial response equalization is performed, for example, in a DVD-RAM (Random Access Memory) capable of digital recording of 4.7 Gbytes on one side, as shown in Fig. 3 (c). The PR (a, b, b, a) method is used so that the waveform amplitude after equalization is divided into five values. As shown in Fig. 5, the DVD-RAM recording format consists of an address block AB in which the address information of the sector SC is recorded in advance as an emboss area for each sector, and data in which random data can be rewritten. Block DB And divided into Here, the characteristics of the reproduced waveform of the random data block often vary depending on the characteristics of the device used for writing. Therefore, the reproduction characteristics of the address block and the data block are often different.

The transversal filter 6 in FIG. 14 is composed of finite taps, for example, a FIR finoleta (Finite Impulse response Filter). The equalization characteristic of the FIR filter is realized by changing the filter coefficient. The FIR filter may be configured by delay elements 16a to 16f, multiplication elements 17a to 17g, and addition means 18 as shown in FIG. The filter coefficients S 1 to S 7 of the FIR filter are based on an LMS (Least Mean Square) algorithm that adaptively controls the Eigen error present in the partial response equalized output signal by the transversal filter 6 to be minimized. This is set by the block-by-block filter coefficient learning means 7 which has a used filter coefficient learning function and learns filter coefficients of the filters for each address block and data block block. The block-by-block filter coefficient learning means 7 may have, for example, a configuration as shown in FIG.

As described in the first embodiment, the temporary decision circuit 19 detects an equalization target value corresponding to the partial response method from the equalization output signal of the transversal filter 6, and compares the equalization target value with the transformer. calculating the equalization error detector 2 0 for detecting the equalization error by subtracting the output signal of the transversal filter ^6, the output signal of the equalization error detecting means 2 0, the correlation between the input signal of the transversal filter 6 Correlator 21 and feedback gain adjuster 22 as means for adjusting the feedback gain by multiplying the output of correlator 21 by gain, and adding the output to the filter coefficient of each tap to update the filter coefficient And a filter coefficient updating unit 23 a to 23 g as a means for performing the learning of the filter coefficients for the address block and the data block. Address block filter coefficient initial value storage means 2411 a to 241 g for storing the result, and data block filter coefficient initial value storage means 2424 a to 24 2 g for the address block. The gate signal generation means 26 generates a signal for identifying the data block, selects the initial value of the filter coefficient for each block, and loads it to the filter coefficient update section 23 a to 23 g Thus, it has a function of performing adaptive automatic equalization control of filter coefficients.

On the other hand, as shown in FIG. 14, the optical disc device according to the second embodiment has a phase comparator 8 for detecting a phase error from an output signal of the offset correcting means 4, and an output from the phase comparator 8. A loop filter 9 for smoothing the phase error signal to be obtained, and an oscillator 10 for oscillating a reproduction clock based on the output signal of the nore-pass filter 9. The phase of the reproduction clock output from the oscillator 10 is And a phase locked loop 11 for controlling the phase of the clock component of the reproduced signal so as to be synchronized with the phase of the clock component. The analog / digital converter 3 uses the reproduced clock output from the oscillator 10 to By sampling the reproduced waveform, a multi-bit sampled signal synchronized with the phase of the peak component of the reproduced signal is generated, so that PMLL signal processing can be realized.

As described above, data demodulation is performed through the maximum likelihood decoder 12 that performs decoding according to the type of the partial response using the partial response equalized signal output by a series of operations. Here, the maximum likelihood decoder 12 may be, for example, a Viterbi decoder. The Viterbi decoder calculates probabilities in accordance with the law of correlation of intentionally added codes, which depends on the type of partial response, and reproduces a likely sequence.

Further, as shown in FIG. 14, the optical disk device according to the second embodiment has an EDC (Electromagnetic Compatibility) added to address information included in a demodulated signal demodulated by the maximum likelihood decoder 12. Error detection code) to determine the correctness of the demodulated address information.Based on the output results of the EDC determination means 35 and the EDC determination means 35, the address block in the filter coefficient learning means 7 for each block is determined. It has filter coefficient learning control means 36 for controlling the hold, reset, etc. of the learning of filter coefficients, so that learning in address blocks etc. containing many single frequency (VFO) parts becomes stable. It is provided with learning stabilization control means 37 for performing control as described above.

Here, for example, in the DVD-RAM address block, as shown in FIG. 15, EDC has 2 bytes for 4 bytes of ID, and by applying EDC, Whether the address information demodulated as ID is correct It may have a feature that makes it possible to determine whether or not. Here, the address information C1 to C4 and the EDC, E1 and E2 shown in FIG. 15 are codes in byte units. Since it is an 8-16 modulation code, one byte may correspond to 16 bits.

By providing such a filter coefficient learning means having a function of determining whether the demodulated address information is correct or incorrect by the EDC determination means 35, the reproduction characteristics are degraded due to a defect or the like in the reproduction of a writable disc or the like. In this case, the learning of the filter coefficient at that position of the address block is stopped, and the learning of the filter coefficient can be performed using only the position that can be satisfactorily reproduced. The learning of the filter coefficient in the block is stabilized, and the possibility of reading the address information stably is improved.

The learning of the filter coefficients in the block-by-block filter coefficient learning means 7 using the EDC determination means 35 described in the second embodiment is controlled by, for example, a procedure shown in a flowchart of FIG. It can be something. This is because when address block filter coefficient learning is started (step 101), when correct address information is obtained by the EDC determination means (step 102) 35, the filter coefficient of the address block is obtained. The adaptive learning is continued (step 103). If correct address information cannot be obtained, the learning of the filter coefficients is stopped and the initial values are loaded from the initial value holding means (step 104). ), And is realized by filter coefficient learning control means 36 having a reset function.

By performing the learning control of the filter coefficient using the EDC determination means 35 as described above, in the case of reproducing a writable disc or the like, if a reproduction abnormality occurs due to a defect or the like, an early recovery can be performed. Therefore, it is possible to shorten the recovery time from abnormal control.

The learning of the filter coefficients in the block-by-block filter coefficient learning means 7 using the EDC determination means 35 described in the second embodiment may have, for example, a configuration as shown in FIG. This is because, in the initial learning of the filter coefficient, the EDC determination means 35 determines that the correct address information is continuously obtained by the determination.

3 8 and the count value of the EDC counter 38 EDC determination initial value storage means 38 a to 38 g for storing filter coefficients as initial values when the address information indicates that the address information is correct for more than an arbitrary number. It has a function of loading information of judgment initial value storage means 38a to 38g.

By providing such a function for performing initial learning of the filter coefficient by the EDC determination means 35, it is confirmed that the reproduction is normally performed in the reproduction of the writable disk or the like, and the filter coefficient is determined. Since the learning result is stored as the initial value, the reliability of the initial value of the filter coefficient in the address block is improved. As described above, the result of learning the filter coefficient by the EDC determination is more accurate and stable than the equalization performance obtained from the normal control state.

As described above, according to the second embodiment, measures are taken when learning the filter coefficient of the CAPA area of the DVD-RAM disk, or when control breaks down during adaptive control or an abnormal state occurs. As a result, the EDC (Error Detection Code) result of the address in the CAPA area in the demodulated signal is constantly monitored, and if the address cannot be detected normally, the control is stopped or reset again, so The return is quick, and the reliability of the filter coefficient can be improved.

(Embodiment 3)

In the third embodiment, for example, in a DVD-RAM disk, when performing adaptive automatic learning of FIR filter coefficients separately for a CAPA area and a data section, stable reproduction of filter coefficients and improvement of control performance are aimed at. It is a thing.

In FIG. 18, after the output amplitude of the raw optical disk signal is enhanced by the preamplifier 1, correction is made by the waveform equalizing means 2 to enhance the high frequency range. 'The waveform equalizing means 2 is constituted by a filter capable of arbitrarily setting a boost amount and a cutoff frequency. For example, it is a high-order equiripple filter having frequency characteristics as shown in FIG. The output signal of the waveform equalizing means 2 is sampled into a multi-bit digital signal by an analog-to-digital converter 3 which converts an analog signal into a digital signal using a reproduced clock generated by the oscillator 10. This sampled multi-bit digital signal is The offset component contained in the reproduced digital signal is corrected by inputting to the offset correction means 4.

The offset correction means 4 may have a configuration as shown in FIG. 2, for example. This means that the output signal of the offset detection means 13 for detecting the offset component of the reproduced digital signal, the smoothing means 14 for smoothing the offset signal detected thereby, and the output signal of the smoothing means 14 And a subtraction means 15 for subtracting from the reproduced digital signal.

The output signal of the offset correction means 4 is input to an auto gain control knob 5, whereby the amplitude of the reproduced digital signal is adjusted to an arbitrary value. For example, the auto gain control 5 may be one that detects the envelope of the signal waveform and controls the difference between an arbitrary set value and the envelope signal to be zero.

Next, the output signal of the auto gain control 5 is input to the transversal filter 6, and partial response equalization is performed. Here, partial response equalization is performed, for example, in a DVD-RAM (Random Access Memory) capable of digital recording of 4.7 Gbytes on one side, as shown in Fig. 3 (c). The PR (a, b, b, a) method is used so that the waveform amplitude after equalization is divided into five values. As shown in Fig. 5, the DVD-RAM recording format is an address block in which sector address information is recorded in advance for each sector as an emboss area, and data in which random data can be rewritten. It is divided into blocks. Here, the characteristics of the reproduced waveform of the random data block often vary depending on the characteristics of the device used for writing. Therefore, the reproduction characteristics of the address block and the data block often differ. The transversal filter 6 is composed of finite taps, for example, a FIR (Finite Impulse response Filter).

The equalization characteristic of this FIR filter is realized by changing the filter coefficient. The FIR finoletor comprises a delay element 16a or 16f, multiplication elements 17a to 17g, and-. Addition means 18 as shown in FIG. You can. The filter coefficients S 1 to S 7 of the FIR filter are present in the partial response equalized output signal of the transversal filter 6. Has a filter coefficient learning function using an LMS (Least Mean Square) algorithm that adaptively controls so that the equalization error to be minimized.Each of the partial response equalization levels in the output signal of the transversal filter 6 is provided. Equalization error averaging means 39 for averaging the absolute value of the equalization error amount at each level, and equalization performance detection means 4 for detecting equalization performance based on the output of the equalization error averaging means 39 0, and a self-determination type filter coefficient learning control means 4 for controlling a method of learning a filter coefficient of the transversal filter by a filter coefficient learning means 42 for learning a filter coefficient according to the detected equalization performance. Set by 1.

The self-determination filter coefficient learning control means 41 may have, for example, a configuration as shown in FIG. This is because the provisional decision circuit 19 detects an equalization target value corresponding to the partial response method from the equalization output signal of the transversal filter 6, and subtracts the equalization target value from the output signal of the transversal filter 6 by subtracting it. An equalization error detector 20 for detecting an equalization error, a correlator 21 for calculating a correlation between an output signal of the equalization error detector 20 and an input signal of the transversal filter 6, and a correlator 21 A feedback gain adjuster 22 as a means for adjusting the feedback gain by multiplying the output of 1 by a gain, and a filter coefficient updating unit as a means for adding the output to the filter coefficient of each tap and updating the filter coefficient Filter coefficient learning means comprising filter coefficient initial value storage means 4 2 a to 42 g for holding learning results of the filter coefficients. Block 43, absolute value conversion means 44 for converting the equalization error amount at each level of the partial response equalization level in the output signal of the transversal filter 6 to an absolute value, and equalization for averaging the output of each. Based on the output result of the error averaging means 39 and the equalization error averaging means 39, an equalization performance self-judgment block 45 composed of equalization performance detection means 40 for detecting equalization performance is provided. Then, based on the self-judgment result of the equalization performance obtained from the equalization performance self-judgment block 45, it has a function of controlling the filter coefficient learning block 43 to perform adaptive automatic equalization control.

On the other hand, as shown in FIG. 18, the optical disc device according to the third embodiment includes a phase comparator 8 for detecting a phase error from an output signal of the offset correcting means 4 and a phase comparator 8 for detecting a phase error. A loop filter 9 for smoothing the output phase error signal; An oscillator 10 that oscillates a reproduction signal based on the output signal of 9 and a phase-locked loop that controls the phase of a reproduction clock output from the oscillator 10 to be synchronized with the phase of a clock component of the reproduction signal. 11 The analog-to-digital converter 3 uses the recovered clock output from the oscillator 10 to sample the reproduced waveform to synchronize the phase of the reproduced signal with the phase component of the reproduced signal. Since a multi-bit sampled signal is generated, PRML signal processing can be realized.

Self-determining filter coefficient learning control that self-analyzes the current digital equalization state and equalization performance based on such equalization error information and self-adjusts adaptive control to satisfy the target performance. The provision of the means 41 makes it possible to adaptively perform digital automatic equalization control with respect to the difference in reproduction characteristics depending on the reproduction position, which appears due to a writable disk or the like or tilt. Leadership is improved for media having

The equalization performance self-judgment block 45 in the self-judgment type filter coefficient learning control means 41 according to the third embodiment may have, for example, a configuration as shown in FIG. This is based on a playback clock that is synchronized with the phase of the clock component of the playback data. Error detection that counts N (N is a positive integer) arbitrary channel bits (the minimum unit of recording data) In reproducing a writable optical disc such as a DVD-ROM (DVD-Read Only Memory), a CD-ROM, and a DVD-RAM, for example, a PRML signal processing method PR (a, b, b, a) When ML is applied, the equalization level is as shown in Fig. 21 (a) due to the fact that the same record code is always repeated three or more times and the run length is limited by "2". As shown, it can be divided into 5 values of level a to level e. Therefore, the absolute value conversion means 47 a for converting the equalization error amount for each level into an absolute value and the absolute value in the N-bit period determined by the error detection period counter 46 are used. Value conversion means 47 Equalization error for adding output signals of 47 a to 47 e respectively Addition means 48 a to 48 e and equalization error averaging means for dividing the output using N as a parameter 49 a to 49 e, and if the equalized error output result of the equalization error averaging means 49 a to 49 e is smaller than a predetermined level X (X is a positive integer), Equalization performance detection means for generating a judgment signal indicating that the equalization state is normal and generating a judgment signal indicating that the equalization state is abnormal when the level is equal to or higher than an arbitrary (predetermined) level X (X is a positive integer). 40.

By having the function of self-determining the equalization performance for each partial response equalization level, the accuracy of detecting the digital equalization performance is improved, and not only is the automatic equalization performance stabilized, but also more than necessary. Since the circuit relating to the learning of the filter coefficients does not need to be operated at the same time, any reproduction performance can be easily realized, so that the power consumption can be reduced.

In the equalization performance self-determination block 45 as shown in FIG. 20, when the equalization performance is determined by using the equalization performance detection means 40, for example, FIG. 21 (a) Among the partial response equalization levels shown in Fig. 21, the evaluation of the equalization performance is valid only for the three levels near the center, level b, level c, and revenor d, which are indicated by black circles in Fig. 21 (b). As an alternative, learning of the filter coefficient may be controlled.

By having the function of self-determining the equalization performance only for an arbitrary selected partial response level, it is significantly related to the performance of the PMLL signal processing, not the overall equalization performance. Detecting digital equalization performance only for the partial response equalization level enables digital adaptive automatic equalization with emphasis on playback performance, and especially readability when the playback waveform has asymmetry characteristics Can be improved. '

The self-determination filter coefficient learning control means 41 according to the third embodiment may perform control as shown in the flowchart of FIG. 22, for example. First, the initial values of the filter coefficients of the transversal filter 6 for absorbing variations in the reproduction characteristics depending on the disk medium immediately after the optical disk medium is inserted into the device. When the output signal of the equalization performance detection means 40 indicates that the equalization state is normal when determining the initial value, the initial value storage for storing the learning result of the filter coefficient as the initial value of the filter coefficient. It is stored in the means 42 a to 42 g (see step 201), and the equalization performance is determined (see step 202). If the equalization state is abnormal, continue learning the filter coefficient until it becomes normal. Then, the filter coefficients are stored in the initial value storage means (see step 203). At the start of data reproduction (see step 204), the values of the initial value storage means 42a to 42g are loaded ( Control is performed with the filter coefficient fixed (see step 206), and the equalization performance is determined (see step 205). If the equalization state is abnormal, adaptive automatic equalization control is started (see step 208). Then, the equalization 4 ability is determined (see step 209). If the output signal of the equalization performance detection means 40 indicates that the equalization state is abnormal, the learning of the filter coefficient is started, and the output signal of the equalization performance detection means 40 becomes Learning is continued until it shows that it is normal, and the filter coefficient at the time when it becomes normal is stored in the initial value storage means 42a to 42g, and that value is held and fixed control is performed. .

The learning of the filter coefficients is performed until the target equalization performance is obtained, and during the period in which the equalization performance is obtained, the control based on the fixed control is performed. In the initial value learning, the reliability of the initial value is improved, and the learning of the filter coefficient is performed again only when the equalization state has deteriorated, so that it is possible to realize the minimum necessary power consumption.

The self-determination filter coefficient learning control means 41 according to the third embodiment may be one that performs control like a flowchart shown in FIG. 23, for example.

First, when determining the initial value of the filter coefficient of the transversal filter 6 to absorb the variation of the reproduction characteristics depending on the disk medium immediately after the optical disk medium is inserted into the apparatus (see step 301). The learning result of the filter coefficient when the output signal of the equalization performance detection means 40 (see step 302) indicates that the equalization state is normal is set as the initial value of the filter coefficient. The value is stored in the value storage means _42a to 42g (see step 303). If the equalization state is abnormal, the learning of the filter coefficient is continued until the value becomes normal. At the beginning of the data world (Step 30 4), the values of the initial value storage means 4 2 a to 4 2 g are loaded (see step ³ 05), and adaptive automatic control is performed by learning continuous filter coefficients (see step 3 0 6). If, during the adaptive automatic control, the output signal of the equalization performance detecting means 40 (see step 300) indicates that the equalization state is abnormal, the learning of the filter coefficient is stopped and the initialization is stopped. After the values of the value storage means 42 a to 42 g are loaded, adaptive automatic control is performed again by continuously learning the filter coefficients.

By having such a self-recovery function, if the control is disturbed due to a difference during digital adaptive automatic equalization control and self-healing becomes impossible with the adaptive equalization function alone, the equalization performance detection is performed. Since the equalization state is determined to be abnormal by the means 40, the filter coefficient is reset to the initial value, and the normal control state can be quickly restored.

As described above, according to the third embodiment, when the optimum filter coefficient is learned for each of the CAPA area and the data part of the DVD-RAM disc, and the optimum equalization is performed for each of them. When constant control is performed, the filter coefficient is returned to the initial value when an abnormal state is detected, and when control is performed with the learned coefficient fixed, the equalization performance is detected, and when the equalization error is large, Since the filter coefficient is learned and the initial value is rewritten, the stability of the filter coefficient and the control performance can be improved. In addition, control is started with the filter coefficient fixed, and when the equalization error is equal to or larger than a predetermined value, learning is started again. When the equalization error is equal to or smaller than the predetermined value, the coefficient is held. By repeating the process of shifting to fixed control with fixed filter coefficients, adaptive control is performed only in the required range, and fixed control is performed in the area where the required characteristics are obtained. Reproduction can be performed.

(Embodiment 4)

The reproduction characteristics of an optical disc may be degraded by a specific error due to not only the difference between the inner and outer circumferences but also the radius and tilt of the disc.

In the fourth embodiment, it is possible to improve an adverse effect due to bending or tilting of a disk.

In FIG. 24, after the output amplitude of the optical disk reproduction signal is emphasized by the preamplifier 1, correction is made by the waveform equalizing means 2 to emphasize the high frequency range. Waveform equalization means 2 It is composed of a filter that can set the amount of cut and cut-off frequency arbitrarily. For example, it is a high-order equiripple filter having frequency characteristics as shown in FIG. The output signal of the waveform equalizing means 2 is sampled into a multi-bit digital signal by an analog / digital converter 3 as means for converting an analog signal into a digital signal using a reproduced clock generated by the oscillator 10. . By inputting the sampled multi-bit digital signal to the offset correcting means 4, the offset component included in the reproduced digital signal is corrected.

The offset correction means 4 may have a configuration as shown in FIG. 2, for example. This is because the output signal of the offset detection means 13 for detecting the offset component of the reproduced digital signal, the smoothing means 14 for smoothing the offset signal detected thereby, and the output signal of the smoothing means 14 And a subtracting means 15 for subtracting from the reproduced digital signal. The output signal of the offset correction means 4 is input to the auto gain control 5, whereby the amplitude of the reproduced digital signal is adjusted to an arbitrary value. For example, the auto gain control 5 may be one that detects an envelope of a signal waveform and controls the difference between an arbitrary set value and an envelope signal to be zero.

Next, the output signal of the auto gain control 5 is input to the transversal filter 6 to perform partial response equalization. Here, the partial response equalization is performed, for example, as shown in Fig. 3 (c) for a DVD-RAM (Random Access Memory) that is capable of digital recording of 4.7 Gbytes with one layer on one side. The PR (a, b, b, a) method is used so that the waveform amplitude after equalization is divided into five values. As shown in Fig. 5, the DVD-RAM recording format is an address block in which sector address information is recorded in advance for each sector as an emboss area, and a data block that can rewrite random data. It is divided into

Here, the characteristics of the reproduced waveform of the random data block often vary depending on the characteristics of the device used for writing. Therefore, the playback characteristics of the address block and the data block often differ.

The transversal filter 6 is composed of finite taps, for example, an FIR filter. Filter (Finite Impulse response Filter). The equalization characteristic by the FIR finoleta is realized by changing the filter coefficient. As shown in FIG. 6, the FIR finalizer is composed of delay elements 16a to 16f, multiplication elements 17a to 17g, and addition means 18. Well. The filter coefficients S 1 to S 7 of the FIR filter are based on an LMS (Least Mean Square) algorithm that adaptively controls the equalization error existing in the partial response equalization output signal by the transversal filter 6 to be minimized. If the optical recording medium is a circular optical disk with the filter coefficient learning function used, as shown in Fig. 25, it is divided into a plurality of zones in the radial direction and the initial learning of the filter coefficient is performed. This is set by the zone-specific filter coefficient learning means 50 having means for storing the result of learning for each zone as the initial value of the filter coefficient. The zone-specific filter coefficient learning means 50 may have a configuration as shown in FIG. 26, for example.

This is because the provisional decision circuit 19 detects the equalization target value corresponding to the partial response method from the equalization output signal of the transversal filter 6 and subtracts the equalization target value and the output signal of the transversal filter 6. Error detector 20 for detecting an equalization error by performing a correlation operation, a correlator 21 for calculating a correlation between an output signal of the equalization error detector 20 and an input signal of the transversal filter 6, and a correlator 21 A feedback gain adjuster 22 as a means for adjusting the feedback gain by multiplying the output of the filter by a gain, and a filter coefficient update unit 23 a to 23 f as a means for adding the output to the filter coefficient of each tap and updating the filter coefficient 2 3 g, and zone-based initial value storage means 5 1 (1,..., N ) a to 5 1 (1,..., N) g, and at the start of data reproduction, initial value storage means for each zone corresponding to the zone to be reproduced 5 1 (1,..., N) a to 5 1 Loads the value of (1,..., N) g into the filter coefficient update unit 23 a to 23 g as the initial value of the filter coefficient, and performs a function to perform adaptive automatic equalization control of the filter coefficient. Have

On the other hand, as shown in FIG. 24, the optical disc device according to the fourth embodiment comprises a phase comparator 8 for detecting a phase error from an output signal of the offset correcting means 4, and a phase ratio A loop filter 9 for smoothing the phase error signal output from the comparator 8, and an oscillator 10 for oscillating a reproduction clock based on the output signal of the noise filter 9. A phase-locked loop 11 that controls the phase of the recovered clock to be synchronized with the phase of the clock component of the recovered signal.The analog-to-digital converter uses the recovered clock output from the oscillator 10. In step 3, by sampling the reproduced waveform, a multi-bit sampled signal that is synchronized with the phase of the peak component of the reproduced signal is generated, so that PRML signal processing can be realized. Become.

By dividing the optical disc medium into a plurality of zones in the radial direction and learning the filter coefficients for each zone, the frequency of the clock component of the reproduced data depending on the radial direction during the CAV reproduction can be reduced. Even if the setting resolution of the high-frequency cutoff frequency of analog waveform equalization for changes is coarse, the readout performance immediately after seek is improved by storing digital equalization filter coefficients in zones, and seek performance is improved. Can be improved.

The zone-based filter coefficient learning means 50 according to the fourth embodiment is, for example, divided into a plurality of zones in a radial direction and a circumferential direction as shown in FIG. And zone-based initial value storage means 5 1 (1,..., N) a to 51 (1,..., N) g for storing each learning result as an initial value of the filter coefficient. At the start of playback, means to store the initial value for each zone corresponding to the zone to be played 5 1 (1, ..., N) a to 51 (1, ..., N) It may perform equalization control.

By dividing the optical disk medium into a plurality of zones in the radial direction and the circumferential direction and learning the filter coefficient for each zone, the write characteristics can be improved. Absorb not only differences in playback characteristics due to playback position and playback block, but also local degradation in playback characteristics due to tilt effects, which appear on writable discs or the like on which digital data has been recorded by different writing devices. allows improvement in Li one Dabiriti an optical recording medium having various recording reproducing JP I ¹ production. The zone-based filter coefficient learning means 50 according to the fourth embodiment may perform adaptive automatic equalization control according to a flowchart as shown in FIG. 28, for example.

First, when the initial learning of the filter coefficient is performed for each zone (Step 401), the output signal (Step 402) of the equalization performance detecting means 40 described in Embodiment 3 is equal to the power equalization state. In particular, a zone indicating bad is provided with a bad zone storage means for recording the zone (step 403), and at the start of data reproduction (step 404), the zone to be reproduced is stored. In step 4, the inferior zone storage means determines the inferior zone (step 406), and if it indicates that the characteristics of the zone are inferior, the control gain in learning the filter coefficient is reduced to reduce the stability. And if the poor zone storage means indicates that the characteristics of the zone are good, the adaptive automatic equalization control () in which the control gain in the learning of the filter coefficient is increased and the equalization performance is prioritized. S Tsu may be a flop 4 0 5).

By switching the equalization control gain or switching the control method in advance in such zones where local degradation of the reproduction characteristics due to the effects is confirmed, stable control and improvement in leadership can be performed. .

As described above, according to the fourth embodiment, when learning the filter coefficient, the optical recording medium is divided into a plurality of zones in the radial direction and the circumferential direction, and the learning of the filter coefficient is performed for each zone. And control is performed using the filter coefficient corresponding to each of these zones as the initial value, so that the error characteristics immediately after the start of reproduction are improved and the tilt due to the tilt on the disc varies depending on the area on the disc. Can be improved. Industrial applicability

As described above, the optical disk device according to the present invention is useful as an optical disk device for reading data from an optical disk medium. It is suitable as an optical disk device that reads data from a recordable optical disk medium such as a disc drive or a CD-R (compact disk-recordable) or CD-RW (co- rpact disk-rewritable) playback device.

Claims

The scope of the claims

1. an equalizing filter that performs partial response equalization on a signal reproduced from an optical recording medium having a data block that is a recordable area and an address block in which the address information is recorded;

Based on the output signal of the partial response equalization, the filter coefficient of the equalization filter is divided into blocks for each block to be adaptively learned so that the equalization error of the equalization filter is minimized for each block. An optical disc device comprising: a filter coefficient learning unit.

2. In the optical disk device according to claim 1,

The block-specific filter coefficient learning means,

An optical disk device comprising a gate signal generating means for generating a gate signal for separating the address block and the data block.

3. The optical disk device according to claim 1,

A preamplifier that emphasizes an output amplitude of a signal reproduced from the optical recording medium; a waveform equalizer that emphasizes a predetermined frequency band of the emphasized signal;

Analog-to-digital conversion means for sampling the waveform-equalized signal using a reproduction clock;

A phase locked loop for controlling the oscillation frequency of the reproduction clock so as to synchronize with the phase of the clock component included in the sampled signal;

Offset correction means for performing a correction to reduce an offset component from the sampled signal, and inputting the corrected signal to the equalization filter;

An optical disc device comprising: a maximum likelihood decoder that performs data demodulation by performing maximum likelihood decoding according to a type of a partial response equalized by the equalization filter.

4. In the optical disk device according to claim 1,

The block-specific filter coefficient learning means,

Envelope detection means for extracting an envelope of the reproduction signal;

By detecting the DC fluctuation of the envelope, An optical disk device comprising: a gate signal generating means for generating a gate signal for identifying the data block.

5. The optical disk device according to claim 4, wherein

The envelope detection means,

A peak holding means for holding a peak value of the waveform of the reproduction signal; and a smoothing means for smoothing the held output signal,

An optical disk device, wherein an envelope of the reproduced waveform is detected.

6. The optical disk device according to claim 1,

The block-by-block filter coefficient learning means,

Tracking error signal generating means for generating a tracking error signal for tracking servo;

An optical disk apparatus comprising: a gate signal generating unit configured to generate a gate signal for identifying the address block and the data block by detecting a DC variation of the tracking error signal.

7. The optical disk device according to claim 1,

The block-specific filter coefficient learning means,

A counter that detects an address position of the address block and counts the number of recording data by a reproduction clock synchronized with a phase of a click component included in a reproduction signal output from the phase locked loop;

A gut signal generating means for generating a gate signal for identifying the address block and the data block by estimating a position where the next address block appears based on the value of the counter. Light

8. The optical disk device according to claim 1,

The block-specific filter coefficient learning means,

Gate signal generating means for distinguishing the address block from the data block;

A single-pattern detection means for detecting a single-frequency pattern present in the reproduced signal, When classifying each block and learning the filter coefficient, the learning of the filter coefficient is performed at a random pattern position other than the single frequency pattern in each block according to the output signal of the single pattern detection means. An optical disc device characterized by the following.

9. In the optical disk device according to claim 1,

The block-specific filter coefficient learning means,

A counter that detects an address position of the address block, counts the number of recording data by a reproduction clock synchronized with a phase of a click component included in a reproduction signal output from the phase locked loop,

Gate signal generation means for generating a gate signal for identifying the address block and the data block by estimating a position where the next address block appears based on the value of the counter;

Pattern discriminating gate signal comfort means for discriminating between a single frequency pattern and a random pattern position based on the value of the counter,

When discriminating each block and learning a filter coefficient, an optical disk device is characterized in that the filter coefficient is learned at a random pattern position other than a single frequency pattern in each block according to the pattern identification good signal. .

10. The optical disk device according to claim 9, wherein:

The block-by-block filter coefficient learning means,

When a plurality of random patterns including the address information in the address block and a single frequency pattern are alternately present, a learning position control means for determining which position is to be used for learning the filter coefficient,

An optical disk device, wherein learning of a filter coefficient is performed based on an output signal of the learning position control means.

11. The optical disc device according to claim 1, wherein:

A maximum likelihood decoder that performs data demodulation by performing maximum likelihood decoding in accordance with a partial response type on a signal obtained by equalization with the equalization filter, and demodulates into binary data by the maximum likelihood decoder Of the address information from the demodulated signal EDC determination means for detecting an EDC (Error Detection Code) for determining correctness and error, and learning of a filter coefficient for an address block corresponding to the address information is stabilized in accordance with an EDC determination result obtained from the EDC determination means. An optical disc device comprising: a filter coefficient learning control means for controlling the block-specific filter coefficient learning means.

1 2. In the optical disk device according to claim 11,

The block-specific filter coefficient learning means,

If the correct address information is obtained in the determination by the EDC determination means, adaptive learning of the filter coefficient in the address block is continued,

An optical disc device characterized in that when correct address information cannot be obtained, learning of the filter coefficient is stopped and the filter coefficient is reset to an initial value.

1 3. The optical disk device according to claim 11,

The block-specific filter coefficient learning means,

At the time of initial learning for determining the initial value of the filter coefficient of the address block, an EDC counter that counts the number of continuous correct address information obtained by the EDC determination means,

Initial value storing means for storing a filter coefficient as an initial value of the filter coefficient when the count value of the EDC counter indicates that the address information is correct for an arbitrary number or more,

An optical disc device characterized by loading information of said initial value storage means at the start of learning of filter coefficients.

1 4. When reproducing digital data from an optical recording medium, an equalizing filter that performs partial response equalization,

Filter coefficient learning means for adaptively learning filter coefficients of the equalization filter so that an equalization error of the equalization filter is minimized based on the equalization output signal; and the equalization is performed. Equalization error averaging means for averaging the absolute value of the equalization error amount at each level of the partial response equalization level in the signal,

Equalization performance detection means for detecting equalization performance based on the output of the equalization error averaging means, An optical disc apparatus comprising: a self-determining filter coefficient learning control unit that controls a learning function of a filter coefficient of the equalization filter.

15. The optical disk device according to claim 14, wherein:

A preamplifier for emphasizing the output amplitude of the reproduction signal from the optical recording medium; a waveform equalizing means for emphasizing a predetermined frequency band of the signal whose output amplitude is emphasized; and a reproduction clock for the waveform-equalized signal. Analog-to-digital conversion means for sampling into digital data;-a phase locked loop for controlling the oscillation frequency of the recovered clock so as to synchronize with the phase of the clock component of the sampled signal;

Digital data correction means for correcting an offset component and an amplitude from the sampled signal and inputting the corrected data to the equalization filter;

Equalization error averaging means for averaging the absolute value of the equalization error amount at each level of the partial response equalization level in the equalization output signal obtained from the equalization filter,

Equalization performance detection means for detecting equalization performance based on the output of the equalization error averaging means,

Self-determining filter coefficient learning control means for controlling a learning function of a filter coefficient of the equalization filter;

An optical disc device comprising: a maximum likelihood decoder that demodulates data from the equalized output signal in accordance with a type of a partial response applied in the equalizing filter.

16. The optical disk device according to claim 14, wherein:

The equalization error averaging means,

Adds the absolute value of the equalization error amount in each of the partial response equalization levels in the equalization filter during a predetermined channel bit (minimum unit of recording data) N (N is a positive integer) period , N as a parameter and perform division.

The equalization performance detection means,

If the equalization error output result averaged by the equalization error averaging means is smaller than a predetermined level X (X is a positive integer), a determination signal indicating that the equalization state is normal is generated, An optical disc device, wherein when the level is equal to or higher than the level X, a determination signal indicating that the equalization state is abnormal is generated.

17. The optical disk device according to claim 14, wherein:

The equalization error averaging means,

The absolute value of the equalization error amount is added for each of the partial response levels in the equalization filter during a predetermined channel bit (the minimum unit of recording data) N (N is a positive integer). , N as a parameter and perform division.

The equalization performance detection means is configured to determine whether the equalization error output result of the specified partial response equalization level averaged by the equalization error averaging means is smaller than a predetermined level X (X is a positive integer). An optical disc characterized by generating a determination signal indicating that the equalization state is normal, and generating a determination signal indicating that the equalization state is abnormal when the equalization state is equal to or higher than a predetermined level X (X is a positive integer). apparatus.

18. The optical disk device according to claim 14, wherein:

The self-determination filter coefficient learning control means,

In the initial learning for determining the initial values of the filter coefficients of the equalization filter, the filter coefficient learning is performed when the output signal of the equalization performance detection means indicates that the equalization state is normal. Initial value storage means for storing the result of

At the start of data reproduction, the value of the initial value storage means is loaded, the filter coefficient is fixed, and control is performed. During the fixed control, the output signal of the equalization bear detection means indicates that the equalization state is abnormal. , The learning of the filter coefficient is started, and the learning is continued until the output signal of the equalization performance detecting means indicates that the equalization state is normal. An optical disk device, wherein a filter coefficient at a point in time is stored in said initial value storage means, and fixed control is performed based on the value.

1 9. The optical disk device according to claim 14, wherein:

The self-determination filter coefficient learning control means,

At the time of initial learning for determining the initial value of the filter coefficient of the equalization filter, the result of the coefficient learning when the output signal of the equalization performance detection means indicates that the equalization state is normal. , Initial value storage means for storing as initial values of filter coefficients Having a step,

At the start of data reproduction, the value of the initial value storage means is loaded, and adaptive automatic control is performed by continuous learning of filter coefficients. During the adaptive automatic control, the output signal of the equalization performance detection means becomes: If it is indicated that the equalization state is abnormal, the learning of the filter coefficient is stopped, the value of the initial value storage unit is loaded, and the adaptive automatic control is performed again by continuously learning the filter coefficient. An optical disk device characterized by performing:

20. The optical disk device according to claim 14, wherein

The self-determination filter coefficient learning control means,

When determining the initial value of the filter coefficient of the equalization filter, the optical disk as an optical recording medium is divided into a plurality of zones in the radial direction and the initial learning is performed, and each learning result is stored as the initial value of the filter coefficient. Zone-based initial value storage means

An optical disc apparatus characterized in that at the start of data reproduction, the value of said zone-specific initial value storage means corresponding to the zone to be reproduced is loaded and adaptive automatic control is performed.

21. The optical disk device according to claim 14, wherein:

The self-determination filter coefficient learning control means,

When determining the initial value of the filter coefficient of the equalization filter, the optical disc as an optical recording medium is divided into a plurality of zones in a radial direction and a circumferential direction, and initial learning is performed. A zone-based initial value storage means for storing as a value,

An optical disk device, wherein at the start of data reproduction, a value of said zone-specific initial value storage means corresponding to a zone to be reproduced is loaded and adaptive automatic control is performed.

22. In the optical disk device according to claim 21,

The self-determination filter coefficient learning control means,

At the time of the initial learning of the filter coefficient, of the plurality of zones on the circular optical disc as the optical recording medium, the zone where the output signal of the equalization performance detecting means indicates that the equalization state is particularly poor. An optical disk characterized in that it has a poor zone storage means for recording the control gain in the adaptive automatic control filter coefficient learning for each zone to be reproduced based on the storage result of the bad zone storage means.

8

CS0 / T0df / X3d 80166/10 OAV