Coded data generation or conversion – Digital code to digital code converters – To or from nrz codes
Reexamination Certificate
1999-05-20
2001-01-16
Tokar, Michael (Department: 2819)
Coded data generation or conversion
Digital code to digital code converters
To or from nrz codes
C360S065000
Reexamination Certificate
active
06175319
ABSTRACT:
TECHNICAL FIELD
The present invention is related to a field of encoding a sequential stream of digital data.
BACKGROUND ART
Recording digital data on disks and tapes in accomplished by passing a write current through a magnetic transducer to align the magnetic domain of the medium in one of two opposing directions. To achieve high data densities on the medium, a non-return to zero inverting on the ones (NRZI) form of encoding is applied to the data stream to produce the write current. An NRZI encoded signal will maintain its present polarity for each logical zero bit in the data stream, and invert its polarity at the start of each logical one bit. The magnetic domain of the medium follows the write current by orienting in one direction when the write current has a positive polarity, and in the opposite direction when the write current has a negative polarity.
NRZI encoding has a limitation in tape drive and disk drive applications when presented with long strings of logical zero bits. During these long strings, the write current remains at the same polarity. This causes long regions in the data tracks to be magnetically oriented in the same direction. Long regions of uniform magnetic orientation sometimes cause a magneto-resistive read sensor to saturate, resulting in data errors.
The Full-Cell Write Equalization method was developed to prevent read sensor saturation. Full-Cell Write Equalization method briefly inverts the write current polarity one or more times during long string of zeros. These brief polarity inversions are called equalization pulses. For a string of N consecutive logical zero bits, the Full-Cell Write Equalization method generates N−X equalization pulses. The value of X is typically equal to d, the minimum number of zeros between adjacent ones for a (d,k) run length-limited modulation code. These equalization pulses are typically centered in time during the writing of the logical zero bits, starting with the first logical zero bit in the string. Other encoding methods space the equalization pulses at equal intervals between the NRZI transitions caused by logical one bits that lead and trail the string of consecutive logical zero bits.
Another problem occurs where data stream encoding causes the write current transitions at high frequencies. The net effect of these high frequency transitions is to shift the apparent position of the leading and trailing transitions as seen by the read sensor. These apparent shifts may sometimes cause problems in the read circuitry. The Write Pre-Compensation method compensates for the apparent shifts seen by offsetting the transition write positions in the opposite direction of the apparent shift. This method is well known in the field of disk drives. Shifts in the apparent position of the data on magnetic medium are sometimes caused by slow rise-times in the write transducers. The Write Pre-Emphasis method applies extra power at the start of a transition to produce a faster rise-time in the leading edge of the transitions. While these two methods reduce the apparent shifts, they require special circuitry to generate the complex write currents.
A common practice in the disk drive and tape drive industry is to read the data from the magnetic medium as it is being written to verify that the correct data is being recorded. For space and alignment purposes, the read sensors and write transducers are usually co-located in the same magnetic head. This close physical proximity, and the wiring connecting the magnetic head to circuit cards, results in cross-talk between the write channel and the read channel. Signals in the write channel are on the order of volts. Signals in the read channel are only on the order of millivolts. Consequently, any time power is applied to the write channel it is picked up in the read channel as unwanted noise.
The Uniform Pulse-Write method partially solves the problem of write signal noise in the read channel by reducing the amount of power in the write signal. This is accomplished by modulating the write current with a duty cycle so that part of the time the write current is not being applied to the write transducer. The frequency of the duty cycle is sufficiently high so that the individual pulses overlap and blend in the magnetic medium. As a result, data written using the Uniform Pulse-Write method has the same read characteristics as data written using a non-pulsed method.
The Pulse-On-Transition method requires even less power than the Uniform Pulse-Write method. Pulse-On-Transition pulses the write current at the same time that the Full-Cell Write Equalization method inverts the write current's polarity. The Pulse-On-Transition method returns the write current to zero at the completion of each pulse. Because some pulses start from a zero write current condition, data written using this method has different read characteristics than data written using the Full-Cell Write Equalization method making the two methods incompatible. Another reason the two methods can have different read characteristics is that a portion of the magnetization in the write transducer may switch so slowly that the magnetization continues to increase as long as the transducer is energized. Therefore, the magnetic state before and after the polarity inversion may be very different for the two methods.
As recording densities increase, the signals from the read sensors will decrease causing a lower signal to noise ratio. To maintain a reasonable signal to noise ratio, the write current induced noise needs to be decreased. A new encoding method is required that produces less write current than the Full-Cell Write Equalization method. At the same time, the data written on magnetic medium using the new encoding method should have the same read characteristics as data written using the Full-Cell Write Equalization method to maintain compatibility with existing disk drives and tape drives.
DISCLOSURE OF INVENTION
The present invention is a method of generating an encoded signal from a sequential stream of digital data, where the encoded signal has a non-power carrying null state and a power carrying active state with two opposing polarities. Logical one bits are distinguished from logical zero bits in the data stream by inverting the polarity of the encoded signal at the start of only the logical one bits. The encoded signal is set to the active state during a bit set-up period before, and held in the active state during a bit hold period after each polarity inversion. At other times the encoded signal is set to the null state. The method may include the addition of equalization pulses during strings of consecutive logical zero bits to keep the encoded signal from remaining in the null state for extended periods. Each equalization pulse is preceded by an equalization set-up period, and followed by an equalization hold period during which the encoded signal is in the active state. Where the equalization hold period of one equalization pulse is contiguous with the equalization set-up period of the next equalization pulse, the equalization set-up period may be eliminated. The equalization set-up period may also be eliminated for select equalization pulses in a string of equalization pulses to balance the magnetic state of the write transducer. This approach is used for write transducers that have slow magnetization reversal mechanisms. The slow magnetization reversal mechanisms can result in an imbalance in the magnetization state when the write transducer is energized to one polarity longer than the opposite polarity, averaged over time during of the string of equalization pulses. Finally, the equalization pulse positions may be centered on the logical zero bits, spread uniformly in time between the logical one bit polarity inversions bounding the string, or in other positions appropriate for the write transducer, recording medium and read sensor.
In an application where the encoded signal is a write current used with a magnetic transducer, the preferred method encodes N−1 equalization pulses into each string of N
Boyer Keith Gary
Koren Norman Lee
Schneider Richard Crane
Brooks & Kushman P.C.
Le Don Phu
Storage Technology Corporation
Tokar Michael
LandOfFree
Power-tailored write-current method does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Power-tailored write-current method, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Power-tailored write-current method will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2444961