Consistent with the present disclosure, various example embodiments are directed to an apparatus including a first circuit configured to present a readback signal and a multiple-track signal correction circuit. The first circuit is configured to present a readback signal representing data stored in at least one of multiple tracks on a data-recording disc. The multiple-track signal correction circuit is communicatively connected to the first circuit and configured to compensate for asymmetry in the readback signal using different correction coefficients for respective ones of the multiple tracks.
In an exemplary environment in which a readback signal is processed in a disc drive apparatus having a disc with multiple tracks and a magneto-resistive transducer reading targeted data stored in one of the multiple tracks, various aspects of the present disclosure are directed toward a signal correction circuit that processes the readback signal by removing or reducing the asymmetry aspects of the noise and distortion.
In other embodiments also consistent with the present disclosure, certain aspects are directed to compensating for asymmetry in the readback signal of the read data stored in at least one of multiple tracks based on different correction coefficients for respective ones of the multiple tracks.
More specific embodiments are directed to a disc drive including a disc, a transducer assembly and circuitry configured and arranged for each of an associated one of the multiple tracks of the disc. The transducer assembly is configured to read magnetically-stored data stored in multiple tracks of the disc by providing a readback signal from each of the multiple tracks. The circuitry for each of an associated one of the multiple tracks is configured with a front-end signal processing circuit that includes an amplifier for providing gain to the readback signal and an analog filter for providing a version of the readback signal in which noise, present in an output signal from the amplifier, is removed. The circuitry for each of an associated one of the multiple tracks is also configured with a signal estimation circuit that includes an analog-to-digital conversion circuit for converting an analog version of the readback signal to a digital signal, and a digital filter for shaping the digital signal and for providing estimated values of the respective readback signals based on the shaped digital signal. Further, a multiple-track signal correction circuit is included to compensate for asymmetry in the readback signal of the read data stored in at least one of multiple tracks, based on the estimated values and different correction coefficients for respective ones of the multiple tracks.
The above discussion/summary is not intended to describe each embodiment or every implementation of the present disclosure. The figures and detailed description that follow also exemplify various embodiments.
Various example embodiments may be more completely understood in consideration of the following detailed description in connection with the accompanying drawings, in which:
While the disclosure is amenable to various modifications and alternative forms, specifics thereof have been shown by way of example in the drawings and will be described in detail. It should be understood, however, that the intention is not to limit the disclosure to the particular embodiments described. On the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the scope of the disclosure including aspects defined in the claims. In addition, the term “example” as used throughout this application is only by way of illustration, and not limitation.
Aspects of the present disclosure are believed to be applicable to a variety of different types of methods, devices, and systems that store data in magnetic media which is susceptible to asymmetric noise or interference from adjacently-stored data such as data that is stored in tracks of discs and read by magneto-resistive transducers. Specific embodiments are believed to be particularly beneficial to a disc drive or other type of apparatus where data (e.g., stored in another tracked form with the data adjacently-aligned) might be susceptible to asymmetric noise or interference. While the present disclosure is not necessarily so limited, various aspects thereof may be appreciated through a discussion of examples using this context.
In certain specific embodiments, a circuit provides a readback signal to a multiple-track signal correction circuit. The readback signal represents data stored in at least one of multiple tracks on a disc of a disc drive. In this exemplary context, the circuit for presenting the readback signal can be an active or passive circuit (e.g., an amplifier or another transistor-based signal-drive circuit, or a passive signal-coupling electrode) that is part of or connected to the magneto-resistive transducer assembly. Using the readback signal, the multiple-track signal correction circuit is configured and arranged to compensate for asymmetry in the readback signal of the read data stored in at least one of the multiple tracks based on the estimated values, and different correction coefficients for respective ones of the multiple tracks.
The following describes various approaches and aspects useful for correcting (or compensating) for asymmetry in the readback signal using different correction coefficients for respective ones of the multiple tracks. These approaches and aspects may be carried out and implemented, alone or together in various combinations thereof, using one or more of the components as shown in
Other aspects are directed toward a disc drive apparatus including a disc having multiple tracks, a magneto-resistive transducer assembly, and multiple-track signal correction circuitry. The magneto-resistive transducer assembly includes a magneto-resistive transducer (or a read/write head) that reads and/or writes (magnetically-stored) data from/to the multiple tracks. The multiple-track signal correction circuitry can be implemented to compensate for asymmetry in a readback signal of read data stored in at least one of multiple tracks based on different correction coefficients for respective ones of the multiple tracks. As discussed further below, in certain specific embodiments, this correction circuitry is configured and arranged to compensate for asymmetry in the signal readback from the disc by the transducer. Depending on the specific embodiment, the (so-called) readback signal can be presented in various forms including an analog form that is raw or cleansed (e.g., smoothed or otherwise filtered to some degree) and including a digital form (e.g., using high-resolution samples) that might also be raw or somewhat cleansed.
In certain embodiments, the analog front-end signal processing circuit has data-processing channels for the respective multiple tracks and a signal estimation circuit providing estimated values of the respective readback signals for the read data stored in the multiple tracks. The signal estimation circuit includes an analog-to-digital conversion circuit and a digital filter for cleaning/shaping the readback signal. The analog-to-digital conversion circuit can be implemented to convert an analog version of the readback signal to a digital signal that, in turn, is next processed for filtering (e.g., using a detection algorithm) and presenting a version of the signal into a digital sequence.
In a typical disc drive application, the transducer (read head) presents such a readback signal by detecting, in a time sequence, the alterations on the medium and generating a sequence of corresponding pulses in an analog form. These pulses are then detected and decoded by read channel circuitry in order to reproduce the digital sequence corresponding to the digital data as stored in the track of the disc. Discrete time sequence detectors can be advantageous because they compensate for intersymbol interference (ISI) and are less susceptible to channel noise. Commonly-used discrete time sequence detection methods include, among many others, discrete time pulse detection, maximum likelihood sequence detection, Viterbi detection, decision-feedback equalization, and variations of these including combinations thereof. Accordingly, Viterbi filtering is one such implementation for presenting a somewhat cleansed and shaped digital version of the readback signal.
In various disc drive embodiments, signal processing is performed for each of an associated one of the multiple tracks, using an analog front-end signal processing circuit, and a signal estimation circuit. The analog front-end signal processing circuit includes a variable gain amplifier providing (e.g., fixed) gain to the readback signal, and a continuous time filter providing a version of the readback signal in which noise, present in an output signal from the variable-gain amplifier, is removed. The signal estimation circuit includes an analog-to-digital conversion circuit for converting an analog version of the readback signal to a digital signal, and a digital filter as discussed above for shaping the digital signal and providing estimated values of the respective feedback signals based on the shaped digital signal.
Additionally, certain embodiments are implemented with the multiple-track signal correction circuitry being configured to cancel such asymmetry noise in the readback signals by executing a mathematical process that operates on data from each of multiple tracks. A mathematical process can be used for such correction/cancellation by operating on inputs including the signal expected to be read from the target track (or “the expected signal”), a correction coefficient and the number of the multiple tracks involved. In more specific embodiments, a multiple-track signal correction circuit compensates for asymmetry noise in the readback signal of the read data stored in at least one of the multiple tracks, relative to an expected signal based on estimated values (as computed using a predictive/adaptive algorithm), and different correction coefficients for respective ones of the multiple tracks.
Consistent with the above, methods are also disclosed that can include the steps of reading data stored in multiple tracks, and compensating for asymmetry in the readback signal. Specifically, the step of reading data is accomplished by using a magneto-resistive transducer assembly, reading data stored in the multiple tracks and providing therefrom a readback signal from each of the multiple tracks. In some embodiments, the method includes the step of compensating for asymmetry in the readback signal of the read data stored in at least one of multiple tracks, relative to an expected signal as processed by the discrete time sequence detection/filtering, based on different correction coefficients for respective ones of the multiple tracks.
Certain aspects of the present disclosure are related to correction for such asymmetry based on the recognition or discovery that related distortion of the signal being read back by a transducer is at least in part due to the effect of the multiple tracks being adjacently situated. In connection with these aspects of the present disclosure, the above-discussed signal correction circuitry can be configured and arranged for effecting correction of the readback signal based on the assumptions or belief that such asymmetry is attributable to significant distortions relating (or corresponding) to magnetic data patterns stored on both the track targeted to be accessed and its adjacent track(s).
Turning now to the figures, various examples of the present disclosure are presented by way of the illustrations provided beginning with
As depicted in the graph 104, the apparatus 100 provides a relatively noise-free output signal as graphically shown in the form of an extracted dipulse (or dipulse signals). More specifically, the graph 104 shows the output of the multi-track asymmetry correction circuitry 102 as a time-sequence signal that is corrected for distortions caused by asymmetry noise/echoes picked up and carried from the MR reader 106. The correction would be evident in the dipulse extraction as very low or no amplitude signal (echo) at the sample index corresponding to distortions.
In certain embodiments in which such an apparatus 100 includes a disc drive, the disc drive can include a multi-track asymmetry correction circuit 102, a disc with data stored in multiple tracks thereof, and a magneto-resistive transducer assembly (MR reader). In many applications, the MR reader presents readback signal to the multi-track asymmetry correction circuitry 102 for processing within the disc drive but it will be appreciated that the readback signal can be transmitted elsewhere for such processing with the multi-track asymmetry correction circuitry 102 being located in a communicatively-coupled network server, cloud computing system, etc.
Such asymmetry has been recognized, in connection with the present disclosure, as being due to the effect of the multiple tracks being adjacently situated. The above-discussed signal correction circuitry is configured and arranged for effecting correction of such readback signal as processed by the transducer. In this context, such asymmetry in the readback signal is believed to be due at least in part to significant distortions relating to magnetic data patterns stored on both the track targeted to be accessed and its adjacent tracks. The data patterns create a plurality of magnetic fields that, when read by the transducer (a read head or the MR reader 106), distort the true signal as represented by the stored data in the transducer-targeted track of the disc. The distortions are also believed to be partially due to manufacturing imperfections of pinning fields in the magneto-resistive transducers. In certain embodiments, the distortions include (but are not necessarily limited to) one or more of the following: background and foreground asymmetry noise, asymmetry echoes, background and foreground saturation, and non-linear mixing of background and foreground signal.
It has also been recognized that by correcting for such asymmetry in the (raw) readback signal, before being processed by a partial response maximum likelihood sequence detection (PRML) channel, a higher mean square error in the detection stage of the PRML channel can be avoided. In certain circumstances, the position of the reader relative to adjacent data tracks may create a readback signal that appears to be devoid of any asymmetry response. However, the readback signal should still be processed for asymmetry because, if left uncompensated, the readback signal will result in a higher error rate when processed later in the channel by signal detection circuitry. By correcting for asymmetry in the readback signal, the adverse effects due to multiple adjacent tracks are significantly mitigated (or removed entirely) as most noticeably apparent when comparing corrected and uncorrected output signal error rates and/or the independent asymmetry echoes/effects as broken down for each track.
Such asymmetry in the readback signal might be better appreciated with reference to
In contrast, the magnetic fields from track 215 cause the MR reader pinning field to be alternatively reduced or increased when the track 215 signal is positive or negative, respectively. This change to the MR pinning introduces a positive asymmetry with respect to the track 215 data, as the positive response of the reader is stronger than the negative response. As discussed above, this results in a sensed signal that is ostensibly more symmetric when analyzed using standard asymmetry correction circuitry. However, the response only appears to be symmetric as the two asymmetric responses of a2 and a1 can have the observed effect of canceling one another.
In
The skilled artisan would appreciate that such dipulse mapping and related background technology is well published in the disc drive literature. This background technology includes dipulse extraction and approaches for characterizing non-linearities of readback signals and characterization of signal parameters such as
Referring now to
After being processed by some form of front-end signal processing circuitry, asymmetric signal issues can be treated for in the readback signal of each track in a number of different ways. As an example, in one embodiment the asymmetric signal issues can be mitigated by signal-error recognition and table look-up compensation. In such implementations, look-up tables could be used to pre-store recognizable signal errors and their corresponding correction signals. Using laboratory testing and/or computations, a discrete number of asymmetry echoes (signal errors) can be developed and stored based on representative distortion-causing patterns. Compensating correction signals can be combined with readback signal as similarly determined by laboratory testing and/or computations. Depending on the range of distortion-causing patterns and the extent to which the asymmetric signal would need to be corrected, more or less look-up memory tables and laboratory evaluation would be needed. Assuming a relatively fast access for the data entries in the look-up memory tables, this approach might be computationally efficient.
From the analog front-end signal processing circuitry 401, the amplified and conditioned signal is then presented to signal estimation circuitry 410 which generates certain signal parameters, from each of the relevant tracks, useful for asymmetric signal correction 403 by computing one of multiple mathematical algorithms as a function of these signal parameters. The signal estimation circuitry 410 performs an analog-to-digital conversion at block 411, which further conditions (or filters) via filter circuit 412 (e.g., finite impulse response filter (FIR), or other digital filter) to further filter and shape the digitized signal to have optimal behavior characteristics for transmission to the initial detector 4131, for example, by implementing a predictive algorithm such as soft-output Viterbi (SOVA). It will be appreciated that such analog-to-digital conversion samples the incoming data (e.g., post-continuous time filtering) at a given data rate for quantizing the processed version of the readback signal for certain N-bit precision.
The initial detector 4131 outputs the detected data bits from the first track (a1), and the filter 4121 outputs the actual digitized signal (y1) from the first track. The duplicative analog front-end circuitry and signal estimation circuitry 2, 3, . . . , N (as shown in
In the explanation above, the expected signal was formed by a convolution of detected data bits with a channel target. In another implementation, the detected data bits weighted by soft-information associated with their detection confidence could be used. In another implementation that may be more appropriate for factory calibration, known fixed patterns could be used throughout the test process when writing to the multiple tracks under consideration, and these known fixed patterns could be used in place of detected bit patterns.
Using any of various methods including, for example, the above-noted dipulse extraction methodology, such a method can be used to characterize and calibrate α1, α2, . . . . In other embodiments, a two-dimensional gradient search or sweep of the potential values of α can provide the best-performing value of the signal parameter denoted as α. In further embodiments and in a manner consistent with the above implementation of look-up tables, the value of the asymmetry correction coefficient α for a given readback signal can be determined based upon a look-up table populated either through regular use of the disc drive apparatus, or during a manufacturing testing process, whereby the block 522 determines α based on the α value generally associated with the target read sector, bit, track, or other method of association between a target read area and α value.
The cancellation block 522 of the multi-track asymmetry correction circuitry 520 cancels/compensates for the multi-track asymmetry in the readback signal, using either a multi-track asymmetry correction summation mode or a multi-track asymmetry correction product mode. Using either mode, the cancellation block 522 produces the corrected signal (yC1, yC2) for readback signals 1 and 2.
As discussed above, the cancellation block 522 cancels asymmetry echoes in the readback signals. In certain embodiments, the cancellation block performs a computation that has the effect of canceling the asymmetry (echoes) in the readback signals by executing a summation-based process from which corrected signals for each of the readback signals are provided. Certain embodiments of the present disclosure are characterized by the summation-based process being a function of, for each of an associated one of the multiple tracks, the expected signal, the correction coefficient and the number of the multiple tracks.
In one summation-based process embodiment, the summation-based process using a computation such as follows:
where, for each of an associated one of the multiple tracks,
In other embodiments of the present disclosure, the cancellation block 522 effects cancelation of asymmetry (echoes) in the readback signals and therefrom produces corrected signals for each of the readback signals by executing a product-based process. Such a process can be implemented with a computation performed as a function of, for each of an associated one of the multiple tracks, the expected signal, the correction coefficient and the number of the multiple tracks. In one type of product-based process embodiment, the product-based process performs a computation as follows:
In Equation 2, for each of an associated one of the multiple tracks, the signal parameter
The multi-track asymmetry correction for MR reader readback signals, as presented in certain experimental/detailed embodiments, has been modeled and tested to evidence significant and surprising results attributed to the use of such multi-track asymmetry correction circuitry. In support,
Various modules or other circuits may be implemented to carry out one or more of the operations and activities described herein and/or shown in the figures. In these contexts, such modules or blocks are understood to be circuitry that carries out one or more of the described operations/activities (e.g., analog front-end signal processing circuitry, signal estimation circuitry, or multi-track asymmetry correction circuitry). For example, in certain of the above-discussed embodiments, such modules or blocks are discrete logic circuits or programmable logic circuits configured and arranged for implementing these operations/activities, as in the circuit modules shown in
Based upon the above discussion and illustrations, those skilled in the art will readily recognize that various modifications and changes may be made to the present disclosure without strictly following the exemplary embodiments and applications illustrated and described herein. For example, the claimed circuits/modules need not be present in the disc drive itself, but instead can be located in a host device that controls a disc drive. Such modifications do not depart from the true spirit and scope of the present disclosure, including that set forth in the following claims.
Number | Name | Date | Kind |
---|---|---|---|
6151177 | Shrinkle et al. | Nov 2000 | A |
6208477 | Cloke et al. | Mar 2001 | B1 |
6400518 | Bhaumik et al. | Jun 2002 | B1 |
6519106 | Goh et al. | Feb 2003 | B1 |
6587292 | Ashley et al. | Jul 2003 | B1 |
7012772 | Vis | Mar 2006 | B1 |
7092180 | Franck | Aug 2006 | B2 |
7126890 | Learned et al. | Oct 2006 | B2 |
7193800 | Coker et al. | Mar 2007 | B2 |
7643238 | DeGroat | Jan 2010 | B2 |
8077427 | Mathew et al. | Dec 2011 | B2 |
8130462 | Marrow | Mar 2012 | B2 |
8149529 | Mathew et al. | Apr 2012 | B2 |
8300339 | Nangare et al. | Oct 2012 | B1 |
8378731 | Tang et al. | Feb 2013 | B1 |
8411383 | Cao et al. | Apr 2013 | B2 |
8441751 | Song et al. | May 2013 | B1 |
8665543 | Bellorado et al. | Mar 2014 | B2 |
20070041115 | Lee | Feb 2007 | A1 |
20070076313 | Hutchins et al. | Apr 2007 | A1 |
20080151704 | Harada | Jun 2008 | A1 |
Number | Date | Country |
---|---|---|
0684484 | Nov 1995 | EP |
0830669 | Mar 2004 | EP |
1739674 | Jan 2007 | EP |
WO0048173 | Aug 2000 | WO |
Entry |
---|
Palmer, et al., “Identification of Nonlinear Write Effects Using Pseudorandom Sequences,” IEEE Transactions on Magnetics, vol. MAG-23, No. 5, Sep. 1987. |
Palmer, et al. “Characterization of the Read/Write Process for Magnetic Recording,” IEEE Transactions on Magnetics. vol. 31. No. 2. Mar. 1995. |
R. Hermann, “Volterra Modeling of Digital Magnetic Saturation Recording Channels,” IEEE Transactions on Magnetics, vol. 26, No. 5, Sep. 1990. |
Robert D. Nowak, “Volterra Filter Identification Using Penalized Least Squares,” Acoustics, Speech, and Signal Processing, May 1996. ICASSP-96. Conference Proceedings., 1996 IEEE International Conference on IEEE, 2813-2816, vol. 5. |