Dynamic magnetic information storage or retrieval – General processing of a digital signal – In specific code or form
Reexamination Certificate
1999-06-01
2001-07-03
Faber, Alan T. (Department: 2651)
Dynamic magnetic information storage or retrieval
General processing of a digital signal
In specific code or form
C360S046000, C360S053000
Reexamination Certificate
active
06256159
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to improvements in signal processing techniques and circuits, and more particularly to improvements in signal processing techniques and circuits of the type used in data processing channels, such as may be used in mass data storage devices, or the like.
2. Relevant Background
This invention relates, in general, to mass data storage devices, or the like. Recently there have been many advances in such devices, resulting in increased speed and decreased cost. However, in order to achieve these advances, many compensating considerations need to be implemented. For example, recently magneto-resistive (M/R) read/write transducers, or heads, have been proposed to serve the function of detecting the magnetic fluctuations of the spinning disk data media and transducing them to electrical signals for processing. The M/R heads are particularly popular, primarily because of their reliability and inexpensive construction. One of the tradeoffs, however, for the reliability and inexpensive construction is that M/R heads produce an asymmetrical output waveform. The asymmetry of the output waveform requires special filtering or compensation for accurate decoding.
One of the engineering goals in the design of a mass data storage device is to achieve as low a bit-error-rate as possible in order to achieve a read signal that is as reliable and accurate as possible. One technique that has been employed is to detect a “di-bit”, rather than a single pulse output from the head. A di-bit, generally refrerred to simply as a “dibit”is a pair of sequential bits, the first bit being of a first polarity, the subsequent bit being of an opposite polarity. For example, if the first bit is negative, the dibit has a positive bit next immediately following. Alternatively, if the first bit is positive, the dibit has a negative bit next immediately following.
When a dibit is detected, it may be stored for subsequent decoding of the data it represents. For instance, in modern mass data storage devices, data is recorded onto a magnetic data disk. The data records are arranged in concentric rings from the interior of the disk radially outwardly to the periphery of the disk. The data tracks are numbered, usually using a Gray code scheme that enables particular tracks to be rapidly located. Within each track are a number of sectors, which also have identification indicators to identify the location of particular user data that may be written on that particular track. The track and sector data are written to specific portions of each respective track for ease in detection; however, again, a significant amounts of data processing and concomitant circuitry is required for increased data access speed.
The data pulses are recorded onto the magnetic media originally by a pulse pattern that includes first a positive (or negative) signal pulse, followed by a negative (or positive) signal pulse. When the data pulse is read back from the disk, the resulting data pattern very closely approximates a sine wave, except, as noted above, when M/R heads are employed as the data transducer, the symmetry of the read back signal is not perfect. Often the positive portion of the read back signal is distorted and is larger than the negative portion of the signal.
In order to detect the existence of the read back sine wave, usually a peak detection technique was employed that employed a sign reversal detection method. The theory is that if the sign of the read back signal changed, the peak of the read back sine wave had occurred. However, often noise or other signal disturbances would introduce signal levels that could be misinterpreted for a sign reversal, thereby falsely detecting the signal, i.e., the dibit. It should be appreciated that the read back signal is digitized relatively early in the signal-processing path. As a result, slope reversal detection in the past has been determined by comparing the signs of adjacent data. This makes the sign reversal technique even more “brittle”, since a sign change between two adjacent signal values may not represent the maximum value of the overall signal, at all.
Other types of dibit detectors have been used, as well. For example, dibit detectors have been used that employ a analog peak detector that detects an initial signal peak and when the peak is detected, the circuit triggers a timer, such as a one-shot multivibrator. The one-shot multivibrator establishes a time during which the peak of a reverse signal must occur for a dibit to be detected. This type of dibit detector also has several disadvantages. For example, if a phase shift occurs in the input signal after the first peak has been detected, the dibit may be missed. Also, if the time period of the one-shot multivibrator is not well defined, the time period may expire before the occurrence of the peak of the second portion of the dibit. These deficiencies, of course, lead to increased bit detection errors.
In order to digitize the read back signal, generally sampling techniques are employed. Typically, the signal is oversampled at a rate consistent with the bandwidth of the signal processing circuitry. It can be seen, however, that if the bandwidth of the processing circuitry is relatively low, the sampling rate also must be proportionally low to enable the read back signal to be properly interpreted.
Finally, in the past, the various data processes that are applied to the read back signal have been performed serially, or sequentially. This has the disadvantage in slowing down, or more accurately, limiting the speed at which the signal can be processed to achieve a given bit error rate. This is because the AGC loop performance is adversely impacted due to increased latency, resulting in decreased bit error rate. This results, moreover, in limiting the disk access time, since the read back data has to be read completely before the heads can be properly positioned to read the user data needed.
SUMMARY OF THE INVENTION
In light of the above, it is an object of the invention to provide an improved dibit detector that performs various signal processing which, among other things, processes in parallel for increased signal processing speed. This is particularly useful in increasing performance for Gray code detection in mass data storage devices, or the like.
These and other objects, features and advantages of the invention will be apparent to those skilled in the art from the following detailed description of the invention, when read in conjunction with the accompanying drawings and appended claims.
According to a broad aspect of the invention, a digital dibit detector is presented in which a band pass filter is connected to receive a sampled signal to produce a filter output signal. A magnitude qualification circuit compares an absolute value of the filter output signal to a programmable qualification threshold value to produce a qualified magnitude output signal. A polarity qualification circuit compares a sample polarity of the filter output signal to a programmable dibit polarity to produce a qualified polarity output signal and a local peak qualification circuit qualifies a largest sample magnitude of the filter output signal within a window as a peak to produce a peak output signal. A circuit receives the qualified magnitude output signal, the qualified polarity output signal, and the peak output signal, to produce a dibit detection signal when the qualified magnitude output signal, the qualified polarity output signal, and the peak output signal are of the same state.
According to another broad aspect of the invention, a method is provided for asynchronous digital dibit detection. The method includes concurrently band pass filtering a sampled signal to produce a filter output signal, magnitude qualifying the filter output signal by comparing an absolute value of the of the filter output signal to a programmable qualification threshold value to produce a qualified magnitude output signal, polarity qualifying the filter output signal by comparing a sample polar
Brady W. James
Faber Alan T.
Swayze, Jr. W. Daniel
Telecky , Jr. Frederick J.
Texas Instruments Incorporated
LandOfFree
Method and circuit for dibit detection does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and circuit for dibit detection, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and circuit for dibit detection will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2493760