Detecting servo data and servo bursts from discrete time samples of an analog read signal in a sampled amplitude read channel

Description Submit all comments and votes

This application is related to U.S. patents, namely U.S. Pat. No. 5,424,881 entitled "Synchronous Read Channel," U.S. Pat. No. 5,359,631 entitled "Timing Recovery Circuit for Synchronous Waveform Sampling," U.S. Pat. No. 5,291,499 entitled "Method and Apparatus for Reduced-Complexity Viterbi-Type Sequence Detectors," U.S. Pat. No. 5,297,184 entitled "Gain Control Circuit for Synchronous Waveform Sampling," and U.S. Pat. No. 5,329,554 entitled "Digital Pulse Detector." All of the above-named patents are assigned to the same entity, and all are incorporated herein by reference.

FIELD OF INVENTION

The present invention relates to computer technology and, more specifically, to servo burst digital demodulation in a sampled amplitude read channel for positioning a read head to retrieve digitized data from a magnetic storage medium.

BACKGROUND OF THE INVENTION

In magnetic storage systems, a transducing head writes digital data onto a magnetic storage medium. The digital data serve to modulate the current in a read/write head coil so that a sequence of corresponding magnetic flux transitions are written onto the magnetic medium in concentric tracks. To read this recorded data, the read/write head passes over the magnetic medium and transduces the magnetic transitions into pulses in an analog signal. These pulses are then decoded by the read channel circuitry to reproduce the digital data.

Decoding the pulses into a digital sequence can be performed by a simple pulse detector in a conventional analog read channel or, as in more recent designs, by using a discrete time sequence detector in a sampled amplitude read channel. Discrete time sequence detectors are preferred over simple analog pulse detectors because discrete time detectors compensate for intersymbol interference (ISI), thereby decreasing the necessary bandwidth. Thus, more data can be stored on the storage medium. There are several types of well known discrete time sequence detection methods including discrete time pulse detection (DSP), maximum likelihood sequence detection (MLSD), decision-feedback equalization (DFE), enhanced decision-feedback equalization (EDFE), and fixed-delay tree-search with decision-feedback (FDTS/DF).

The application of sampled amplitude techniques to digital communication channels is well documented. See Y. Kabel and S. Pasupathy, "Partial Response Signaling", IEEE Trans. Commun. Tech., Vol. COM-23, pp.921-934, Sept. 1975; and Edward A. Lee and David G. Messerschmitt, "Digital Communication", Kluwer Academic Publishers, Boston, 1990; and G. D. Forney, Jr., "The Viterbi Algorithm", Proc. IEEE, Vol. 61, .pp. 268-278, March 1973. Applying sampled amplitude techniques to magnetic storage systems is also well documented. See Roy D. Cideciyan, Francois Dolivo, Walter Hirt, and Wolfgang Schott, "A PRML System for Digital Magnetic Recording", IEEE Journal on Selected Areas in Communications, Vol. 10 No. 1, January 1992, pp.38-56; and Wood et el, "Viterbi Detection of Class IV Partial Response on a Magnetic Recording Channel", IEEE Trans. Commun., Vol. Com-34, No. 5, pp. 454-461, May 1986; and Coker Et al, "Implementation of PRML in a Rigid Disk Drive", IEEE Trans. on Magnetics, Vol. 27, No. 6, Nov. 1991; and Carley et al, "Adaptive Continous-Time Equalization Followed By FDTS/DF Sequence Detection", Digest of The Magnetic Recording Conference, August 15-17, 1994, pp. C3; and Moon et al, "Constrained-Complexity Equalizer Design for Fixed Delay Tree Search with Decision Feedback", IEEE Trans. on Magnetics, Vol. 30, No. 5, Sept. 1994; and Abbott et al, "Timing Recovery For Adaptive Decision Feedback Equalization of The Magnetic Storage Channel", Globecom '90 IEEE Global Telecommunications Conference 1990, San Diego, Calif., Nov. 1990, pp.1794-1799; and Abbott et al, "Performance of Digital Magnetic Recording with Equalization and Offtrack Interference", IEEE Transactions on Magnetics, Vol. 27, No. 1, Jan. 1991; and Cioffi et al, "Adaptive Equalization in Magnetic-Disk Storage Channels", IEEE Communication Magazine, Feb. 1990; and Roger Wood, "Enhanced Decision Feedback Equalization", Intermag '90.

In disk drives utilizing either analog or sampled amplitude read channels, the read/write head is normally mounted on an actuator arm which is positioned by means of a voice coil motor ("VCM"). The VCM moves the head and actuator arm assembly across the disk surface at a very high speed to perform seek operations in which the head is positioned over a selected data track. The VCM also maintains the head over a selected track while reading or writing information. A servo system controller is the subsystem of the disk drive which is responsible for providing the head positioning necessary for reading and writing information in response to requests from a computer to which the disk drive is connected.

Along each track, the magnetic data are arranged consecutively about a centerline of the tracks. The data are generally organized into sectors or fields of predetermined length. A field of information is often preceded by a field of control information that may be used to verify the position of the head before a subsequent read or write operation. The data information fields may also include an error correction code ("ECC") which aids in correcting errors that may occur when information is read. In embedded servo disk drives, position verification and control information is contained in a servo field which is recorded on the tracks at the time of manufacture, utilizing a high precision servo writer or other techniques. The servo field information is used to perform continuous on-track positioning of the head with respect to the centerline of the track by reading and responding to the control information contained within the servo fields. The servo fields are interspersed with data fields in which the data information is recorded.

The servo control information typically includes a preamble which demarks the beginning of a servo field, a servo address mark ("SAM") which indicates that a valid servo field has been detected, a servo synch mark ("SSM") which is utilized to establish and maintain synchronization over reading and writing operations, an index mark which indicates a single reference point common to all the tracks or a band of tracks on the disk and a track number code, which is a Gray coded integer value of the track currently spanned by the read/write head.

The embedded servo field also typically includes off-track burst information which is written on the track when the disk drive is manufactured. The off-track bursts, which also comprise magnetic pulses, are physically positioned at precise intervals and locations with respect to the various track centerlines to provide the servo system controller with information relative to the fractional track-to-track displacement of the head with respect to a given track centerline. Normally, there are four off-track bursts, and the information obtained by reading the burst is sometimes referred to as quadrature signals, quadrature information or quadrature data. In the typical disk drive, the quadrature data are utilized by a data processor associated with a servo system controller to generate, calculate and provide control signals to the VCM to accurately position the head over the track centerline.

The servo control information in the servo field is commonly extracted from the head's signal, in conventional analog read channels, by an analog circuit which detects the presence of individual pulses. For example, U.S. Pat. No. 4,783,705 discloses an analog pulse detector circuit which detects peaks in the analog signal from the head (whether positive or negative in amplitude). These amplitude signals are then converted to digital signals and then passed to a servo controller. This technique is susceptible to noise in the channel and can erroneously detect two consecutive positive or negative pulses when, in magnetic recording, the pulses normally occur with alternating polarity.

Also in conventional analog read channels, the servo burst information in the servo field is typically extracted using an analog circuit that measures the servo burst amplitudes. These servo burst amplitudes are then processed by a motion control processor which generates control signals for positioning the read/write head. Typically, the amplitude of the off-track bursts is measured with analog peak detectors, which respond to the maximum of the head signal. Alternatively, the off-track bursts may be measured by analog area detectors (as in U.S. Pat. No. 4,783,705) which respond to the integrated amplitude of the head signal. In either case, the conventional burst amplitude measurement is generated by analog circuits and passed as an analog signal from a read channel integrated circuit ("IC") to an additional ADC in a separate servo controller.

Such conventional analog techniques for servo demodulation are inefficient for use in sampled amplitude read channels such as PRML read channels. Sampled amplitude read channels operate with discrete time circuits (and commonly digital circuits) which, being programmable, are highly configurable and adaptable. It is inefficient to incorporate the conventional analog servo demodulation circuits into a sampled amplitude read channel when programmable discrete time techniques can be implemented instead. Further, the discrete time circuitry already incorporated within a sampled amplitude read channel, such as an analog-to-digital converter and discrete time pulse detector, can also be used to implement demodulation of the servo data. Sharing the discrete time circuitry is a more cost effective implementation of servo demodulation since it requires less die area and less power. Finally, the prior art analog servo demodulation systems incorporated within a read channel cannot be programmably adapted to operate according to the various disk drives, data densities, and magnetic media found in the market. Nor can the prior art demodulation systems be programmably adapted to compensate for changes in the disk drive that occur over time.

Although for sampled amplitude read channels it is more economical to implement servo demodulation using discrete time circuitry, there are drawbacks which are overcome by the present invention. For example, the discrete time burst amplitude measurements are subject to inaccuracies due to variations in the sampling phase. Also, the burst amplitude measurement is subject to inaccuracies due to inconsistent timing of the burst detection signal. Further, the resolution of the channel ADC is inadequate for that required for servo demodulation.

Thus, a general object of the present invention is to demodulate servo control data in a magnetic storage system utilizing discrete time circuitry. Specifically, it is an object to provide discrete time servo demodulation in a sampled amplitude read channel IC. A further object is to share the discrete time circuitry already incorporated within a sampled amplitude read channel IC with the discrete time servo demodulation technique of the present invention, thereby minimizing the integrated circuitry and associated cost. Still another object is to transfer the servo field data to a servo controller through a wholly digital interface, thereby obviating the servo controller analog-to-digital converter. Yet another object is to implement servo demodulation utilizing programmable circuitry in order to adapt its operation to a particular disk drive system. Still a further object is to prevent the detection of two consecutive positive or negative pulse in the servo data. Another object is to increase the effective resolution of the channel ADC through various digital signal processing techniques. Still another object is to overcome inaccuracies in the burst amplitude discrete time measurement caused by variations in the sapling phase. A final object is to control the timing of the burst amplitude measurement so that the head signal is sampled over an integer number of servo burst cycles.

SUMMARY OF THE INVENTION

The objects of the present invention are achieved by incorporating, within a sampled amplitude read channel IC, a unique apparatus and method for processing servo fields in discrete time. Rather than incorporate analog circuitry into the read channel IC to demodulate the servo fields using conventional analog methods, the servo fields are demodulated in discrete time. The discrete time circuitry of the read channel, such as the analog-to-digital converter and pulse detector, is advantageously shared with the discrete time demodulation technique of the present invention. The analog signal from the read head is converted into a sequence of discrete time sample values and optionally converted to digital values by the ADC of the read channel. The pulse detector of the read channel detects the servo control data of the servo field, and a discrete time area detect circuit measures the amplitude of the servo bursts. These signals are then transferred digitally to a servo controller for positioning the read/write head in response thereto. Thus, the analog-to-digital converter found in a conventional servo controller is obviated.

In the preferred embodiment of the present invention, a digital pulse detection circuit receives an input signal from a discrete time sample generator and supplies signals indicative of a presence and polarity of servo data pulses of a servo information field recorded on a rotating storage medium. The digital pulse detection circuit includes a discrete time peak detector for detecting peaks in the analog head signal represented by the discrete time sample values. The discrete time peak detector includes a discrete time slope detection circuit for detecting a change of slope in the analog head signal. Also, a pulse may be detected only if its polarity is opposite in sign from the polarity of the previous pulse. This makes the pulse detection circuit less susceptible to noise since, in magnetic recording, the pulses always occur with alternating polarity. The pulse detection circuit may be a simple peak detector, or a complex sequence detector, such as a Viterbi sequence detector, for detecting both servo data and user data.

The preferred embodiment of the present invention also includes a discrete time area detection circuit that receives the discrete time sample values and supplies signals indicative of a magnitude of individual ones of a number of bursts in a servo burst segment of a servo information field recorded on the rotating storage medium. The discrete time area detection circuit comprises: a rectifier circuit to rectify the sample values, an accumulator circuit to accumulate a predetermined number of the rectified samples, and a plurality of storage .registers corresponding to the servo bursts to store the accumulated samples.

The discrete time servo demodulator circuit of the present invention overcomes servo burst amplitude measurement sensitivity to the timing of the analog signal sampling through digital processing and control. Such otherwise inherent sensitivity is overcome by a combination of novel signal processing techniques. The head signal is sampled over an integer number of servo burst pulses to increase the accuracy of the burst amplitude measurement. This results in high accuracy of the servo burst amplitude measurements by eliminating the variation in area due to mismatch between the area timing and cycle period of the servo burst signal. Additionally, a sweep of the sampling frequency around a nominal frequency further reduces the sensitivity to the sampling phase. Finally, the signal samples are rectified or squared, and interpolated in order to further increase the effective resolution of the channel ADC and to achieve even higher accuracy in the discrete time burst amplitude measurement.

The foregoing, and other features and objects of the present invention and the manner of attaining them will become more apparent, and the invention itself will be best understood by reference to the following description of a preferred embodiment taken in conjunction with the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a simplified illustration of the organization of alternating servo control information and data information fields, or sectors, disposed along a track centerline of a disk drive and schematically illustrating a read/write head displaced slightly off-track prior to the arrival of a servo data field beneath it.

FIG. 2 is a simplified illustration of a typical servo field shown in FIG. 1, Showing preamble, SAM, SSM, index mark and track number code portions thereof, as well as off-track bursts for providing fractional off-track servo control information.

FIG. 3 is a simplified and enlarged illustration of a read/write head (shown in phantom) along a track centerline prior to the off-track servo burst shown in FIG. 2 passing underneath the head due to the rotation of the disk.

FIG. 4 illustrates a conventional analog read channel and associated servo controller, the former incorporating an analog pulse detector and analog servo burst measurement circuit for providing digital servo data and analog burst signals, respectively, to the servo controller.

FIG. 5 illustrates a sampled amplitude read channel and associated servo controller in accordance with the discrete time servo demodulator circuit of the present invention wherein the read channel comprises a discrete time area detection circuit and discrete time pulse detection circuit for providing the servo controller with digital information representative of the servo burst and servo data, respectively, and wherein the servo controller need not incorporate an additional ADC circuit.

FIG. 6 is a functional block diagram of a peak detector circuit for use in conjunction with the discrete time servo demodulator of the present invention.

FIG. 7 is a more detailed logic block diagram of the peak detector operable to produce output data in one of four modes of operation.

FIG. 8 illustrates the operation of the peak detector when processing the sampled analog signal from the magnetic read head.

FIG. 9 is a simplified logic block diagram of a frequency dither circuit which allows the sampling frequency to be changed over a set of frequencies within a small fraction of a nominal sampling rate.

FIGS. 10A-10D illustrate a sampled servo burst signal and corresponding rectified burst samples, squared burst samples and interpolated burst samples respectively, useful for understanding the principles of the discrete time area detect circuit of the discrete time servo demodulator circuit of the present invention.

FIG. 11 is a simplified logic block diagram of a discrete time area detect circuit wherein the signal samples are first passed through a non-linearity block and then summed in an accumulator prior to being stored in a corresponding register.

FIG. 12 is a logic block diagram of a portion of the discrete time area circuit of the present invention wherein all samples and interpolated samples are rectified and accumulated to produce a servo burst amplitude measurement with the signal samples being processed in parallel.

FIG. 13 is a simplified logic block diagram of a servo burst accumulator in accordance with the present invention for use in conjunction with the discrete time area detection circuit.

DESCRIPTION OF A PREFERRED EMBODIMENT

The features of the present invention interact with and respond to the servo control information contained in a servo field 10 which is embedded or otherwise present in a track 12, as shown in FIG. 1. A plurality of servo fields 10 are located along the track 12, interspersed with a plurality of data fields 14 as shown. The pattern of a servo field 10 and a data field 14 repeats in the direction of rotation of the disk, with the preceding servo field 10 usually being associated with the following data field 14. The adjacent servo fields 10 and data fields 14 may be separated by gaps or transitions 16 as shown, or they may be contiguous to one another and have no transitions delimiting the field boundaries (a situation which is not shown in FIG. 1).

The servo fields 10 and data fields 14 lie along a track center line 18. The dibits which define the information on the track are located linearly along the center line 18. The magnetic reversal or the absence of a magnetic reversal at each physical interval along the length of the track 12 signals the presence of a logical one or a logical zero, respectively.

A conventional transducer or head 20 (shown schematically) reads the dibits from the servo field 10 and the data field 14 on the track 12 over which the head is positioned or writes information to the data fields 14. Generally, the information contained in the servo fields 20 should not be overwritten during use of the disk, thereby ensuring the preservation of the servo field information. A conventional actuator arm (not shown) is connected to the head to suspend and position it over the surface of the disk upon which the tracks 12 of information are located. A conventional voice coil motor ("VCM", not shown) is connected to the actuator arm to move the arm and the attached head 20 to locate the head over a selected track 12.

If the information in the servo and data fields is to be written and read in a reliable manner, the head 20 should be positioned over the center line 18 of the track 12. When the head is not located on the center line 18 of the track, there is a greater risk that the information will not be read or written in a reliable manner, and the risk increases with increasing displacement of the head from the center line 18.

The servo control operation is illustrated in FIG. 1. At time t.sub.0, a valid measurement of a position error is obtained by reading information from the servo field 10. At time t.sub.1, a control signal in response to the position error is applied to the VCM in order to stabilize movement of the actuator arm and head 20 to the desired position without overshoot or hunting.

More details concerning one possible implementation of the control information contained in the servo field 10 is shown in FIG. 2. The servo field 10 includes a number of subfields 24, 26, 28, 30, 32, and 34. These subfields define a preamble 24, a servo address mark ("SAM") 26, a servo synch mark ("SSM") 28, an index mark 30, a track number code 32, and the servo bursts 34. The preamble 24 (commonly comprised of 2T data: 0,0,1,1,0,0,1,1,0,0, . . . ) delimits the beginning of the servo field 10 and facilitates automatic gain control and timing recovery. The SAM 26 follows the preamble 24 and comprises one or two servo address fields ("SAFs", none of which are specifically shown). The SAM 26 serves to indicate that a valid servo field 10 has been detected and to signal that the SSM 28 will follow. The SSM 28 is an unique pattern of dibits which is used to establish and maintain synchronization between sequential data and servo fields. The SSM 28 is followed by the index mark 30 which indicates a specific position on the track as a whole, usually with respect to a single defined radial position on the disk. The track number code 32 follows and its magnetic transitions form Gray coded information indicative of an integer value representation of the number of the track currently spanned by the read write head 20.

The information in the subfields 24, 26, 28, 30, and 32 of the servo field 10 is the dibit magnetic transitions recorded in the track 12. The position of these dibits is at the center line 18 of the track 12 and results in the creation of information which is binary in nature when read by the head 20, due to the fact that the magnetic transitions are centered about the center line 18. However, the off-track burst subfield 34 generates an analog component of the control signal derived by the head 20 as it passes over a first or "A" off-track burst 36, a second or "B" off-track burst 38, a third or "C" off-track burst 40 and a fourth or "D" off-track burst 42. The dibits of the B, C and D bursts 38-42 are located at positions off of or to the side of the track center line 18. When the dibits of the A, B, C and D bursts 36-42 are read by the head 20, four different analog signals result, depending on the physical position of the head 20 relative to the bursts. The analog nature of the signals derived by the bursts 36, 38, 40 and 42 is represented by different heights of the bursts in the subfield 34 shown in FIG. 2.

The dibit patterns of the off-track bursts 36, 38, 40 and 42 are very accurately positioned or written to the disk surface using a laser interferometer, laser positioning system or other suitable technique. The dibit off-track bursts 36, 38, 40 and 42 are commonly located at predetermined locations with respect to the track center line 18, as shown in FIG. 3. In this example, each track includes a C burst 40 and a D burst 42 positioned adjacent to but on opposite sides of the track center line 18. Each track also includes either an A burst 36 or a B burst 38. For example, each track having an even track number may have an A burst 36, while the odd numbered tracks on each side of the even numbered track include a B burst 38. Track center lines 18a and 18b respectively represent the track numbers N-1 and N+1 of the track number N represented by the center line 18. With the alternating occurrence of the A and B bursts 36, 38 on adjacent tracks 18, and the consistent positional relationship of the C and D bursts 40, 42 on each track 18, there is no overlap or conflict in the position of the bursts on the tracks.

The derivation of the different magnitude analog off-track signals by the head 20 reading the bursts 36, 38, 40 and 42 can be understood by reference to FIG. 3. The head 20 is shown positioned directly above the track center line 18 as the off-track servo bursts 36, 38, 40 and 42 approach due to the rotation of the disk. At time t.sub.A, the A burst 36 will pass directly beneath head 20. At time t.sub.B, the B bursts 38 from the adjacent track center lines 18a and 18b will pass substantially to sides of the head 20. At time t.sub.C, the C burst 40 will pass under approximately one-half of head 20, while at time t.sub.D, the D burst 42 wi