Dynamic magnetic information storage or retrieval – General processing of a digital signal – Data in specific format
Reexamination Certificate
2000-08-11
2002-05-21
Neal, Regina Y. (Department: 2651)
Dynamic magnetic information storage or retrieval
General processing of a digital signal
Data in specific format
C360S046000, C360S051000
Reexamination Certificate
active
06392830
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Technical Field
The present invention relates generally to the retrieval of data from a storage medium and more particularly to a system and method for enabling synchronization of a clock to read data from the medium when the primary synchronization byte has been corrupted.
2. Description of the Background Art
Magnetic disk drives utilized for the storage of information are well-known. Data is stored in the form of magnetic transitions created on the surfaces of one or more magnetically coated storage disks stacked on a rotatable spindle motor assembly. The data is organized into a plurality of annular rings or tracks, and each group of tracks having a same location on the disks is referred to as a cylinder. Data tracks are further subdivided into one or more blocks or sectors of data. The actual format of a data track depends upon the particular design of the disk drive system. Regardless of format, the electronics controlling the storage and retrieval operations of a disk drive must have the means to precisely and reliably determine the start of user data in each data sector so that data may be accurately reproduced.
Data is retrieved or “read” from a disk surface with a magnetic transducer or read “head” which is positioned in close proximity to the rotating disk for sensing the magnetic transitions and converting them into an electrical signal. For high track densities, magnetoresistive (MR) read heads are desirable because of their high degree of sensitivity. Each head is electrically coupled to arm electronics (AE) including a preamplifier, which in turn is coupled to a data channel preferably of the PRML type. The PRML channel includes, among other things, an automatic gain control circuit (AGC), a variable frequency oscillator (VFO) circuit, and sync byte detection circuitry. Descriptions of PRML channels are provided in commonly assigned U.S. Pat. Nos. 5,220,466 and 5,255,131.
The heads are mounted to a linear or rotary actuator assembly for selective positioning of the heads over desired tracks. Movement of the actuator assembly is controlled by the servo electronics and servo microcode, which regulate a control signal to a voice coil motor. Closed loop servo systems utilize feedback information from the disk to find and maintain a position over a target track. This feedback information may be located on a single, dedicated disk surface (i.e., dedicated servo) or embedded on data tracks between portions of data (embedded servo). Numerous servo systems are known in the art for providing a disk drive with the means for seeking to a desired track and following the track during reading or writing. Embedded servo disk drives are described, for example, in U.S. Pat. Nos. 5,285,327 and 5,369,535.
Reading and writing of data is accomplished through the data channel under the direct control of a disk controller, which includes a sequencer for executing microcode control sequences. Data to be written to a storage disk is received by the disk drive in binary form. Before writing, the incoming data stream is first encoded and clocked for enhanced readback reliability. Encoding assures that magnetic transitions recorded around a track are spaced sufficiently far apart to prevent interference between adjacent transitions which may corrupt the readback signal. For a more detailed discussion of encoding schemes, the reader is referred to commonly assigned U.S. Pat. Nos. 4,707,681 and 5,461,631. Clocking assures constant spacing of transitions to obtain a desired track bit density and a constant readback signal frequency. Data frequency may be constant from track to track, or may vary, e.g., as in banded recording schemes. See, for example, U.S. Pat. No. 5,440,474. In sector servo systems, data sectors are usually recorded at frequencies and amplitudes different from those of the servo sectors. In addition, some of the data sectors may be “split” by servo fields.
A data track for any of the. preceding formats typically comprises “splices” of data. That is, portions of data will be written to the track at different times, and consequently, the frequency of each portion will not be synchronous with other portions on the same track. It is therefore crucial to data retrieval that the disk drive's read channel can adjust to the proper clock phase of each data splice. Additionally, amplitude may vary among the heads and frequency may vary from band to band. Thus the channel must also be able to adjust to the particular amplitude and frequency of the data splice being read. Finally, it is important to correctly identify the starting location of the first frame of user data in the data splice.
To facilitate the preceding requirements, a preamble is annexed to each data splice at the time of writing. The preamble normally includes a plurality of repeating patterns having the same amplitude, frequency and phase as the data, to be used by the channel in preparation for reading. It also includes an identifier for use in locating a particular data block. Some disk drives employ headerless or ID-less data block formats for enhanced data capacity, i.e., data blocks in which the preamble does not contain identification or “ID” information. See, for example, commonly owned U.S. Pat. No. 5,438,559, and application Ser. No. 08/082,826, filed on Jun. 23, 1993 and Ser. No. 08/218,546, filed on Mar. 28, 1994 for a description of disk drives employing a “NoID” (™) format.
When a read head encounters a desired data splice, it first passes over an amplitude adjustment portion of the preamble, often called the automatic gain control or AGC field. The AGC field contains a repeating pattern for producing a corresponding repetitive electrical signal of the same amplitude as the data to be read. The repeating signal is used by the AGC circuit in the data channel to adjust a variable gain amplifier (VGA) and thereby amplify the read signal to a predetermined normalized level. The AGC field must be long enough to accommodate fluctuations in spindle speed, transients occurring in a write to read switch, and the actual amplitude adjustment operation. Determination of the appropriate length or number of bytes is made by studying the response of a particular disk drive configuration.
The head next passes over a data synchronization or VFO portion of the preamble comprising a repeating pattern that for simplicity is identical to the repeating pattern of the AGC field. The VFO pattern produces a repetitive electrical signal of the same frequency and phase as the data to be read. It is used by the VFO circuit in the data channel for tuning a variable frequency oscillator, e.g., a voltage controlled oscillator (VCO), to match the frequency and phase of the signal. The VFO field must be long enough to accommodate this synchronizing operation.
The VFO field is followed by a pattern or group of adjacent patterns, generally referred to as “sync bytes”, that mark the beginning of the data field and provide a frame of reference for correctly distinguishing data bytes. Sync bytes are detected by sync byte detection logic in the data channel that looks for one or more predetermined sync byte patterns during a certain window of time. Once the sync byte is identified, the data bytes that follow can be properly decoded.
Portions of the information on a magnetic disk are known to become defective over time for a variety of reasons. Bits, bytes, and even large areas of dropout, e.g., 15-20 bytes in length, occasionally occur and may be the result of phenomena such as contamination or thermal asperities. Data loss is avoided in some cases with error correction that is used to detect errors and reconstruct lost bits or bytes of data. Two examples of error correction code are provided in commonly assigned U.S. Pat. Nos. 4,494,234 and 4,706,250. Despite this protection, disk defects which corrupt the sync byte field are catastrophic to data retrieval. Thus a sync byte detection scheme is required that can tolerate both small and large dropout.
A number of schemes have been implemented to enhance the rob
International Business Machines - Corporation
Neal Regina Y.
Raissinia Abdy
LandOfFree
System and method for providing nonadjacent redundancy... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System and method for providing nonadjacent redundancy..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System and method for providing nonadjacent redundancy... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2836505