Abstract: A tape burnishing device presents aligned stretches of burnishing tape at each side of a rotating disk. Air nozzles at each side of the disk urge an elliptical pattern of the respective stretches of tape into contact with the disk. During burnishing the tape is stationary while the device is advanced to burnish the desired surface of the disk. The tape stretches, that are portions of a single tape, are positioned at an angle with respect to the radius of the disk and the device is advanced in a radial direction to optimize the burnishing action relative to the rotating disk surface. The tape is advanced between burnishing operations twice the dimension of use during a burnishing operation with the two operating locations out of phase to enable the single tape to advance past both burnishing locations while presenting an unused surface at each location and using the entire surface of the tape.

Abstract: A magnetic head/support assembly 10 for assembly into the data access mechanism 70 of a magnetic disk file comprises a magnetic head element 12 and a support structure 11, 20, 21 for the head element. The head/support assembly also comprises a head lead locator 27 including a frame 51 for locating and terminating the leads 15 from the head element 12 remotely from the element. The locator is attached to the support structure and the leads extend across the frame and are retained in fixed spaced apart relationship over a sufficient portion of their lengths to permit direct electrical connection to a correspondingly spaced pattern of conductors on external data channel circuitry carried by the access mechanism.

Abstract: A system for clamping and retaining a stack of rigid magnetic disks is shown using disk spacer rings with an annular groove in the outer cylindrical surface. A concentrated axial load is applied to the stack by support surfaces which overlie the grooved portions of the spacers. The concentrated load is only partially distributed over the radial extent of the spacers with respect to the disks at the axial ends of the stack to resist coning or cupping of such end disks, while the force distribution approaches a uniform loading with respect to inner disks of the stack to retain the undistorted, planar orientation of such inner disks. The grooved spacers also provide a surface which can be used to engage and manipulate the disk assembly during manufacture of the drive.

Abstract: A device that is to be isolated from mechanical shock and vibration such as a magnetic disk storage device is provided with a series of four shock mounting members that have a bolt projecting from one axial end and a stiffened grooved portion at the opposite axial end. A frame member with a rigid end portion and cantilevered, resilient frame members extending therefrom carries a series of U-shaped hook portions that are aligned with the shock mounting members. The pair of hooks on the end portion face the same direction and extend in the opposite direction with respect to the cantilevered frame portions. The other pair of hooks are mounted on the resilient cantilevered frame portions and face toward or away from one another so that they may be placed in engagement with the cooperating shock mounting member grooves by forcing the cantilevered frame portions toward or away from one another and thereafter releasing such parts when the hooks are aligned with the respective grooves.

Abstract: Data byte fields and corresponding ECC byte fields to be magnetically recorded as multi-byte records is randomized prior to run-length encoding. The randomized/encoded data is then written. When the randomized/encoded data is read back, the data is first decoded, and then derandomized in a complementary fashion. The randomizing avoids the writing of byte patterns that tend to stress the ability of a head/disk interface to record the data transitions at proper linear track positions. When reading, the complementary derandomizing scheme avoids presenting a repeating pattern of read-errors to the ECC network. Thus, read-errors are not propagated through the read channel. Data is randomized, and subsequently derandomized, in a repeating or pseudo random, complementary sequence for each byte. The complementary randomizing/derandomizing sequence is determined by the count contained in a counter.

Abstract: A pulse generator and a disk sector boundary signal generator are implemented using an optical pattern of alternating light and dark radial lines or segments which are adhered to and rotate in unison with the rotor of a motor and are sensed by an optical sensor. The output of the optical sensor includes error conditions including sensor error, pattern error component, the wow component of the motor and the flutter component of the motor. A first pulse generating means eliminates sensor and pattern error components and the flutter component leaving only the wow component of the sensor output. A second pulse generating means that eliminates the higher frequency error components leaving pattern error components, wow components and flutter components, where the system is designed to generate flutter components with a frequency at least ten times the frequency of the wow component. The first pulses serve as disk sector boundary reference pulses.

Abstract: Clocking of Class IV partial response coded binary data is provided by way of circuit means that includes two analog threshold detectors. One detector is responsive to the analog read signal's positive-going amplitude corssing a a preset positive threshold value. The other detector is responsive to the read signal's negative-going amplitude crossing a preset negative value. The time of occurrence of both detector crossing transitions is phase-compared to a clock signal, and a phase error signal is generated for each detector, if a phase error exists. The two phase error signals are integrated by the use of a loop filter. The integrated phase error signal is then used to adjust the phase of a clock signal generator. The output of the clock signal generator is used to accurately clock, and to enable accurate recover of, the binary data that was originally encoded and written in accordance with the Class IV partial responsive coding convention.

Abstract: A magnetic disk drive spindle motor start routine is provided that supersedes the spindle motor primary motor control when a stuck condition, wherein one of more transducer heads are adhered to the disk surface, is found to exist at start. The motor is provided current pulses at the resonant frequency of the worst case stuck system with the sequence limited to a number of pulses that will not cause the torque amplification to exceed a value that would damage the head/suspension assemblies. In another embodiment, the stuck system is provided with a motion sensor to sense positive and negative motion during oscillation of the stuck system and provide current to the motor when motion is positive to effect torque amplification.

Abstract: A synchronizing circuit for synchronizing an asynchronous serial two level input signal to a first clock signal employs a majority decoder which receives the asynchronous input signal at one of its inputs. A first latch is enabled by a second phase shifted clock signal to latch the asynchronous input signal to provide a second input to the decoder. A second latch is enabled by the first clock signal to latch the majority decoder output as the synchronized output signal. An inverted version of the output signal is fed back to the third input of the majority decoder after being delayed by an interval of less than the period of said clock signals.

Abstract: A magnetic recording apparatus and method for the constant frequency recording of concentric data tracks on a magnetic recording disk in a manner to provide continuously reducing track width and continuously reducing track-to-track spacing from the disk's inner track to the disk's outer track. A magnetic recording head moves on a path from the disk's inner track to the disk's outer track. The head's gap-to-radius skew increases as the head moves from the inner track to the outer track. Thus, the gap's width-projection onto a disk radius at the location of each track continuously decreases in length, from the inner track to the outer track. As the track width progressively decreases, the track-to-track spacing also decreases, to thereby provide higher track density at the disk's outer tracks. A flying head having a parallelogram shaped slider is described, allowing use of the invention in non-contact recording.

Abstract: A fastening structure is provided to secure a shaft between parallel walls of a box shaped casting which will secure ans locate the shaft without inducing a distortion of the casting walls. One end of the shaft is secured to the adjacent wall and the other end of the shaft is secured through a wall opening using an annular slug and a deformable washer which is compressed by a bolt attached axially to the shaft end to cause the deformed washer to engage the surfaces defining the wall opening. The uniform axisymmetric expansion of the washer does not affect the position of the shaft and the radial fore secures the shaft end without influencing the structure of the casting.

Abstract: The head suspension assembly of a magnetic disk file is mounted to the disk file`s rigid actuator arm by means of an intermediate, arm-encircling, low profile, mounting band. A number of different mounting bands are disclosed for use in securing the head suspension assembly to the actuator arm. A head suspension assembly is permanently attached to the mounting band during a subassembly manufacturing procedure. The unitary mounting-band/head-suspension subassembly is then removably mounted on the end of the actuator arm by use of the mounting band.

Abstract: A data processing system is shown packaged in an enclosure wherein the system frame or chassis and a cable carrier are separately fabricated and thereafter merged to provide an enclosure with open ends and a central cable carrier that mounts connections for the system components. The system components are slideably received by the enclosure and are self docking as the electrical connections with the system power and signal lines are completed as the components arrive at the fully inserted, latched position. Keying means are also provided to require the user to install the components at both the correct location and with the correct orientation.

Abstract: In a servo controlled device such as a disk drive actuator or a robot that is driven by a DC motor and must seek from one position to another position and settle at the target position in a minimum time, a deadbeat control response can effect settling in a minimum number of servo sample times with no steady state error and no ripple between sampling instants. The device design, initial conditions entering the deadbeat controlled mode and the frequency of servo samples can be used to enable deadbeat control by limiting the control current to a value that will not saturate the motor coil or exceed the linear operating range of the controlled device.

Abstract: A method for the selective decarburization of an iron based substrate, such as a silicon iron substrate, wherein the carbon present in the substrate is extracted by the selective deposition of a titanium nitride layer or film on the substrate.

Abstract: The invention uses a small sized, low cost, high accuracy, semiconductor laser interferometer to position the transducer in a disc storage system. Light is reflected back over the path to be measured and directly back into the semiconductor laser. This causes the power present in the laser to be varied. Unexpectedly, the output of the photosensor associated with the laser has an asymmetric, sawtooth waveform.

Abstract: A disk file 10 employs at least two filters 15,16 for removal of airborne particles, one filter 16 having a relatively coarse mesh and a relatively high rate of airflow and the other 15 having a relatively fine mesh and a relatively low rate of airflow. The fine filter 15 may be combined with a third, breather filter 30 which removes particles from air entering the disk enclosure. The main purpose of the coarse filter 16 is to remove the larger, more damaging particles very rapidly, thereby reducing the risk of damage to the surfaces of the disks 12.

Abstract: The magnetic data storage system includes a magnetic transducer head-suspension assembly with an inline suspension slider with a reduced height leading portion that provides for gimballed connection to the suspension load beam and a full height rear portion that enables the use of the increased height portion for the transducer coil windings. The suspension includes a single piece combined gram form and load beam with the entire head-suspension assembly being contained within the height of the slider full height portion to enable minimization of the head-suspension overall height required in magnetic disk drives with closely spaced disks and back to back transducer heads or minimum height head-suspension assemblies that extend between disk surfaces.

Abstract: A structure is shown having three stages for guiding a system component into an operatively mounted position within an enclosure while simultaneously making a blind attachment of the electrical connectors to complete the power and signal circuits between the component and system contained within the enclosure. A coarse positioning structure includes a guide rail means for directing the component toward the operative, mounted position; an intermediate positioning structure functions within the tolerance of the coarse positioning means to substantially align the component including the electrical connectors and a fine positioning means aligns the electrical connectors for the final positioning as the component is latched into the operating position and electrical connections are completed. The three positioning functions are intitated serially.

Abstract: An equalizer circuit is shown which conditions a read back signal from a data storage device by performing amplitude correction (pulse slimming) and phase correction. The circuit provides a simplified single stage that effects both amplitude and phase correction and does so using components that can be selected or adjusted independently of one another. Further, the principal embodiment shows a single ended circuit that provides the function with a minimum number of components to conserve space on the circuit card as well as reducing cost.