Method for secondary-actuator failure detection and recovery in a dual-stage actuator disk drive

[0001] This application is related to concurrently-filed co-pending application Ser. No. 11/051,392 and titled “DUAL-STAGE ACTUATOR DISK DRIVE WITH SECONDARY ACTUATOR FAILURE DETECTION AND RECOVERY USING RELATIVE-POSITION SIGNAL”.

[0002] This application is also related to previously-filed pending application Ser. No. 10/997,153 filed Nov. 24, 2004 and titled “DISK DRIVE WITH A DUAL-STAGE ACTUATOR AND FAILURE DETECTION AND RECOVERY SYSTEM FOR THE SECONDARY ACTUATOR”.

[0003] 1. Field of the Invention

[0004] This invention relates generally to magnetic recording hard disk drives, and more particularly to a disk drive with a dual-stage actuator for positioning the read/write heads.

[0005] 2. Description of the Related Art

[0006] Magnetic recording hard disk drives with dual-stage actuators for positioning the read/write heads on the disks have been proposed. A rotary voice-coil-motor (VCM) is typically the primary actuator, with the secondary actuator attached to the VCM and the read/write heads attached to the secondary actuator. A servo control system receives servo positioning information read by the read/write heads from the data tracks and generates control signals to the primary and secondary actuators to maintain the heads on track and move them to the desired tracks for reading and writing of data. As in conventional single-stage actuator disk drives, each read/write head is attached to the end of a head carrier or air-bearing slider that rides on a cushion or bearing of air above the rotating disk. The slider is attached to a relatively flexible suspension that permits the slider to “pitch” and “roll” on the air bearing, with the suspension being attached to the end of the VCM actuator arm. The secondary actuator is typically a piezoelectric or electrostatic milliactuator or microactuator located on the VCM actuator arm for driving the suspension, or on the suspension between the suspension and the slider for driving the slider, or on the slider for driving just the read/write head.

[0007] The conventional servo control system for a typical dual-stage actuator disk drive uses a controller designed to assure stability of the VCM with adequate stability margins as if it were to operate without the secondary actuator. Then the controller for the secondary actuator is designed to achieve the desired combined dual-stage bandwidth. The secondary actuator control loop and the combined dual-stage control loop are also designed to ensure adequate stability separately and jointly with the other control loops. This type of servo control system is satisfactory for limited increases in the bandwidth above what is achievable with only the VCM.

[0008] In application Ser. No. 10/802,601 filed Mar. 16, 2004, which is now U.S. Pat. No. 6,977,793 B2, titled “MAGNETIC RECORDING DISK DRIVE WITH DUAL-STAGE ACTUATOR AND CONTROL SYSTEM WITH MULTIPLE CONTROLLERS”, and assigned to the same assignee as this application, a dual-stage actuator disk drive is described that operates with an improved servo control system that has two controllers. One controller is a dual-stage controller that simultaneously generates a primary actuator control signal and a secondary actuator control signal, and uses a degraded-stability primary actuator controller design with relatively high low-frequency open-loop gain and a secondary actuator controller design that provides stability and high mid-frequency to high-frequency open-loop gain resulting in increased bandwidth. The other controller is a single-stage controller that generates only a primary actuator control signal and uses a stable VCM-only controller design. If a potential failure of the secondary actuator is detected, the servo control system selects the single-stage controller.

[0009] In dual-stage actuator disk drives with either the conventional servo control system or the control system of the co-pending application, a failure of the secondary actuator will result in reduced performance and may lead to loss of data and/or failure of the disk drive. In the related co-pending application Ser. No. 10/997,153 filed Nov. 24, 2004, titled “DISK DRIVE WITH A DUAL-STAGE ACTUATOR AND FAILURE DETECTION AND RECOVERY SYSTEM FOR THE SECONDARY ACTUATOR”, and assigned to the same assignee as this application, a secondary-actuator failure detection test is performed by generating a test signal to the secondary actuator and measuring a calibration signal from the read head as the read head detects test blocks in special calibration tracks located on the disk. If the calibration signal indicates only reduced performance of the secondary actuator from which failure is recoverable, the controller parameters are adjusted. The writing of the special calibration tracks containing the test blocks increases the time and cost of the servowriting process.

[0010] What is needed is a dual-stage actuator disk drive that can detect actual failure of the secondary actuator and modify the controller parameters if the failure is recoverable, but without the need for special calibration tracks on the disk.

[0011] This invention is a method for secondary-actuator failure-detection and recovery in a dual-stage actuator disk drive where a calibration test run is by the servo control processor and the position of the secondary actuator is measured relative to its neutral position in response to the calibration test. The secondary-actuator failure detection and calibration test can be performed on a regular schedule or at selected times, such as at disk drive start-up. With the primary actuator at a specific test location, such as an actuator crash stop or the load/unload ramp, the servo control processor generates a test signal to the secondary actuator and receives a relative-position signal (RPS) from the relative-position sensor in response to the test signal.

[0012] The test comprises two measurements: a measurement of the secondary actuator static characteristics, and a measurement of the secondary actuator dynamic characteristics. In both measurements, the RPS is used to measure the secondary actuator movement relative to its neutral position.

[0013] The static characteristics measurement is a calculation of the secondary actuator “stroke”, i.e., the amount of secondary actuator movement as a function of voltage input to the secondary actuator, and a comparison of the calculated stroke to a predetermined range of acceptable stroke values. With the primary actuator biased against the crash stop or load/unload ramp, a first test signal is applied to the secondary actuator and the RPS is used to calculate the stroke. The calculated stroke values are averaged over several disk rotations. If the calculated stroke is outside the acceptable range, then this is an indicator that the secondary actuator has likely failed.

[0014] If the calculated stroke is within the acceptable range, then the dynamic characteristics measurement is made. This measurement is essentially a measurement of the plant frequency response of the secondary actuator. With the primary actuator biased against the crash stop or load/unload ramp, a second test signal is applied to the secondary actuator and the RPS is recorded. The second test signal for the dynamic characteristics measurement is a series of test signals, each a sinusoidal signal at a constant frequency. The RPS is detected during the application of the constant frequency test signal, and the resulting gain and phase of the response are recorded along with the corresponding frequency. This is repeated for each frequency in the series of test signals. This enables the plant frequency response of the secondary actuator to be measured. If the response is significantly different from the expected response, it is virtually assured that the secondary actuator has failed. If the measured frequency response shows minor changes, such as a minor increase or decrease in the gain, or a minor shift in the frequency at which the maximum gain occurs, the controller parameters are adjusted or re-optimized. This re-optimization changes the values of the controller parameters in the memory accessible by the servo control processor. The parameters that can be changed include parameters that affect bandwidth or stability margins, notching of particular frequencies such as the secondary actuator resonant frequency, active damping of the secondary actuator resonance, or other performance, robustness, or stability metrics.

[0015] The invention is applicable to both the dual-stage actuator disk drive with a servo control system having a conventional dual-stage controller, and the dual-stage actuator disk drive according to the previously-cited co-pending application Ser. No. 10/802,601 that has both a dual-stage controller and a selectable single-stage controller.

[0016] In the dual-stage actuator disk drive according to the co-pending application Ser. No. 10/802,601, a potential failure of the secondary actuator is detected either by providing a model of the dynamic response of the primary and secondary actuators and comparing the modeled head-position with the measured head-position, or by measuring the relative position of the secondary actuator with the relative-position sensor and comparing the relative position to a modeled position of the secondary actuator. Upon detection of a potential failure of the secondary actuator, the single-stage controller is selected, the primary actuator is moved to the test location, and the secondary-actuator failure detection and calibration test is run. If the secondary actuator passes both the static characteristics measurement test and the dynamic characteristics measurement test, then the dual-stage controller is re-selected. If the measured frequency response shows minor changes from the optimized frequency response, the controller parameters are adjusted or re-optimized prior to re-selection of the dual-stage controller.

[0017] For a fuller understanding of the nature and advantages of the present invention, reference should be made to the following detailed description taken together with the accompanying figures.

Prior Art

[0033] FIG. 1 is a block diagram of a conventional single-stage actuator disk drive that uses servo positioning information located in angularly-spaced servo sectors for positioning the read/write heads. The disk drive, designated generally as 100, includes data recording disk 104, a voice coil motor (VCM) 110 as the primary and only actuator, an inner-diameter (ID) crash stop 102 and an outer-diameter (OD) crash stop 103 for the VCM 110, an actuator arm 106, a suspension 107, a head carrier or air-bearing slider 108, a data recording transducer 109 (also called a head, recording head or read/write head), read/write electronics 113, servo electronics 112, and servo control processor 115. If the disk drive is a “load/unload” type of disk drive wherein the slider 108 is lifted off or “unloaded” from the disk when the disk drive is not operating, then a load/unload ramp 105 is provided for supporting the slider 108. The VCM 110 moves near or beyond the perimeter of the disk and the suspension rides up the ramp 105 to unload the slider 108 from the disk 104.

[0034] The recording head 109 may be an inductive read/write head or a combination of an inductive write head with a magnetoresistive read head and is located on the trailing end of slider 108. Slider 108 is supported on the actuator arm 106 by a suspension 107 that enables the slider to “pitch” and “roll” on an air-bearing generated by the rotating disk 104. Typically, there are multiple disks stacked on a hub that is rotated by a disk motor, with a separate slider and recording head associated with each surface of each disk.

[0035] Data recording disk 104 has a center of rotation 111 and is rotated in direction 130. The disk 104 has a magnetic recording layer with radially-spaced concentric data tracks, one of which is shown as track 118. The disk drive in FIG. 1 is illustrated as a zone-bit-recording (ZBR) disk drive because the data tracks are grouped radially into a number of annular data zones or bands, three of which are shown as bands 151, 152 and 153, but the invention is fully applicable to a disk drive that does not use ZBR, in which case the disk drive would have only a single data band. The range of rotation of the VCM is between the ID crash stop 102 and the OD crash stop 103, which define the ID and OD of the annular region where data can be written. Each data track has a reference index 121 indicating the start of track. Within each band, the tracks are also circumferentially divided into a number of data sectors 154 where user data is stored. If the disk drive has multiple heads, then the set of tracks which are at the same radius on all disk recording layers is referred to as a “cylinder”.

[0036] Each data track also includes a plurality of circumferentially or angularly-spaced servo sectors. The servo sectors in each track are aligned circumferentially with the servo sectors in the other tracks so that they extend across the tracks in a generally radial direction, as represented by radially-directed servo sections 120. The servo positioning information in each servo sector includes a servo timing mark (STM), a track identification (TID) code, and a portion of a pattern of magnetized blocks or high-frequency bursts that are decoded to provide a head position error signal (PES).

[0037] The servo positioning information in the servo sectors is read by the read/write electronics 113 and signals are input to the servo electronics 112. The servo electronics 112 provides digital signals to servo control processor 115. The servo control processor 115 provides an output 191 to VCM driver 192 that controls current to the VCM 110 to position the head 109.

[0038] Within the servo electronics 112, the STM decoder 160 receives a clocked data stream from the read/write electronics 113. Once an STM has been detected, an STM found signal is generated. The STM found signal is used to adjust timing circuit 170, which controls the operating sequence for the remainder of the servo sector.

[0039] After detection of an STM, the track identification (TID) decoder 180 receives timing information from timing circuit 170, reads the clocked data stream, which is typically Gray-code encoded, and then passes the decoded TID information to servo control processor 115. Subsequently, the PES decoder 190 (also called the servo demodulator) captures the position information from read/write electronics 113 and passes a position error signal (PES) to servo control processor 115.

[0040] The servo control processor 115 includes a microprocessor 117 that uses the PES as input to a control algorithm to generate the control signal 191 to VCM driver 192. The control algorithm recalls from memory a “controller” 116, which is a set of parameters based on the static and dynamic characteristics of the “plant” being controlled, i.e., the VCM 110. The controller 116 is a “single-stage” controller because the disk drive being described has only a primary actuator, i.e., VCM 110, and servo control processor 115 provides only a single output, i.e., signal 191 to VCM driver 192. The control algorithm is essentially a matrix multiplication algorithm, and the controller parameters are coefficients used in the multiplication and stored in memory accessible by the microprocessor 117.

[0041] The method of designing the controller 116 is well-known in the digital servo control and disk drive servo control literature. The controller can be designed in the frequency domain to achieve the desired open-loop input-output frequency response of the VCM 110. The input-output behavior of a dynamic system at any frequency can generally be expressed by two parameters, the gain (G) and the phase (φ) representing the amount of attenuation/magnification and phase-shift, respectively. The gain and phase of a dynamic system represent the frequency response of the system and can be generated by experiment. In disk drive single-stage servo control systems the controller 116 must be a stable design. FIG. 2 is an example of an open-loop frequency response 210 for a disk drive with only a single-stage actuator, i.e., VCM 110. The single-stage controller 116 for this system assures stability. For example, at the gain zero-crossover, the phase margin is about 30 degrees and at −180 degree phase the gain margin is about 5 dB. Also, at a natural resonance near 4 kHz, the phase is near zero such that the system is phase stable.

[0042] FIG. 3 shows one example of a secondary actuator usable with the above-described disk drive wherein the secondary actuator is a rotary microactuator 200 located between the slider 108 and the suspension 107. Rotation of the microactuator 200 as represented by arrow 202 causes rotation of the slider 108 and thus movement of head 109 in the cross-track direction. The microactuator 200 maintains the head 109 on-track, while the VCM 110 moves the slider 108 (and thus head 109) across the tracks. Other types of secondary actuators are also well-known, such as piezoelectric-based actuators. Also, the secondary actuator may be located on the suspension or actuator arm to move a suspension or arm-section relative to the fixed actuator arm, as in U.S. Pat. No. 5,936,805, or between the slider and a slider-segment to move the head relative to the slider, as in U.S. Pat. No. 6,611,399.

[0043] FIG. 3A is an exploded view of microactuator 200, the disk drive suspension 107, and the air-bearing slider 108. The slider 108 has an air-bearing surface 32 that faces the disk of the disk drive and a trailing surface 34 that supports the read/write head 109. The microactuator 200 depicted in FIG. 3A is a rotary electrostatic microactuator as described in detail in U.S. Pat. No. 5,959,808 and L. S. Fan et al., “Electrostatic Microactuator and Design Considerations for HDD Applications”, IEEE Transactions on Magnetics, Vol. 35, No. 2, Mar. 1999, pp. 1000–1005. The rotary electrostatic microactuator 200 comprises fixed and movable portions on an insulating substrate 51, which is typically a silicon base with an insulating surface layer, such as a layer of polyimide, silicon oxide or silicon nitride. The substrate 51 is mounted to the end of the suspension 107. The fixed portion of the microactuator 200 includes a central post 52 fixed to substrate 51 and two sets of stationary electrodes 53, 53′ also fixed to substrate 51. The movable rotary portion of the microactuator 200 includes a support frame 54, a plurality of flexible members or springs 55 connecting the frame 54 to the fixed post 52, and a plurality of movable electrodes depicted as two sets 56, 56′ attached to frame 54. The sets of movable electrodes 56, 56′ are interleaved with the sets of fixed electrodes 53, 53′, respectively. The frame 54 may optionally support a platform to which the slider 108 is bonded. All of the elements attached to the substrate 51 are formed by lithographically patterning the substrate 51 and depositing electrically conductive material, such as nickel or copper.

[0044] A voltage applied to stationary electrodes 53 will apply an electrostatic attractive force between stationary electrodes 53 and movable electrodes 56, tending to cause the frame 54 to rotate counterclockwise. A voltage applied to stationary electrodes 53′ will apply an electrostatic attractive force between stationary electrodes 53′ and movable electrodes 56′, tending to cause the frame 54 to rotate clockwise. Thus a first set of stationary electrodes 53 and movable electrodes 56 are associated with counterclockwise rotation and a second set of stationary electrodes 53′ and movable electrodes 56′ are associated with clockwise rotation. The frame 54 rotates (as depicted by arrows 202) due to the flexibility of the springs 55 and thus moves the read/write head 109 to maintain its position on a data track on the disk.

[0045] The conventional control system for a disk drive with a dual-stage actuator is similar to that described above except that there is a second output from the servo control processor that is sent to the driver for the secondary actuator, and the controller is a dual-stage controller. In the dual-stage control system, the VCM or primary actuator is typically responsible for large-amplitude, low-frequency motion and the microactuator or secondary actuator is typically responsible for small-amplitude, high-frequency motion. The design of a dual-stage controller for a hard disk drive dual-stage servo control systems is well-known, as described for example, in Y. Lou et al., “Dual-Stage Servo With On-Slider PZT Microactuator for Hard Disk Drives”, IEEE Transactions on Magnetics, Vol. 38, No. 5, September 2002, pp. 2183–2185; and T. Semba et al., “Dual-stage servo controller for HDD using MEMS microactuator”, IEEE Transactions on Magnetics, Vol. 35, September 1999, pp. 2271–2273. Generally, the design of a dual-stage controller starts with the VCM controller, typically with a design that is very similar to a single-stage VCM controller, such as the design for the VCM controller whose open-loop frequency response 210 is shown in FIG. 2. In particular, the stability of the VCM is assured with adequate robustness or stability margins as if it were to operate without the secondary actuator. Then the controller for the secondary actuator is designed to achieve the desired combined dual-stage bandwidth. The secondary actuator control loop and the combined dual-stage control loop are also designed to ensure adequate stability separately and jointly with the other control loops. This process is satisfactory for limited increases in the bandwidth above what is achievable with only the VCM.

[0046] An example of an open-loop frequency response for a dual-stage hard disk drive with a conventional stable VCM controller design is shown as frequency response 220 in FIG. 4 and compared with the frequency response 210, which is also shown FIG. 2. The response 220 is similar to response 210 at low frequencies, but has higher gain in the mid-frequency range, and a higher bandwidth. For disk drives in general and for the frequency responses described herein, the low frequency range is generally meant to be below approximately 300 Hz, the mid frequency range is generally meant to be approximately 300 Hz to 2 kHz, and the high frequency range is generally meant to be above approximately 2 kHz. However, the boundaries between what are considered low, mid, and high frequency ranges are more accurately linked to where the microactuator begins to dominate the overall frequency response and the open loop bandwidth (0 dB crossover) achievable with a single-stage actuator. This is strongly related to the physical size of the disk drive.

The Invention

[0047] In a dual-stage actuator disk drive as described above, failure of the secondary actuator will result in reduced performance and may lead to loss of data and/or failure of the disk drive. Thus it is important to be able to detect actual failure of the secondary actuator and re-optimize the controller parameters if the failure is recoverable.

[0048] This invention is a dual-stage actuator disk drive that uses a secondary-actuator failure-detection and calibration test run by the servo control processor and a relative-position sensor for measuring the position of the secondary actuator relative to its neutral position in response to the calibration test. The secondary-actuator failure detection and calibration test can be performed on a regular schedule or at selected times, such as at disk drive start-up. With the primary actuator biased against a crash stop or load/unload ramp, the servo control processor generates a test signal to the secondary actuator and receives a relative-position signal (RPS) from the relative-position sensor in response to the test signal.

[0049] The test comprises two measurements: a measurement of the secondary actuator static characteristics, and a measurement of the secondary actuator dynamic characteristics. In both measurements, the RPS is used to measure the secondary actuator movement relative to its neutral position.

[0050] The static characteristics measurement is a calculation of the secondary actuator “stroke”, i.e., the amount of secondary actuator movement as a function of voltage input to the secondary actuator, and a comparison of the calculated stroke to a predetermined range of acceptable stroke values. With the primary actuator biased against the crash stop or load/unload ramp, a first test signal is applied to the secondary actuator and the RPS is used to calculate the stroke. The calculated stroke values are averaged over several disk rotations. If the calculated stroke is outside the acceptable range, then this is an indicator that the secondary actuator has likely failed.

[0051] If the calculated stroke is within the acceptable range, then the dynamic characteristics measurement is made. This measurement is essentially a measurement of the plant frequency response of the secondary actuator. With the primary actuator biased against the crash stop or load/unload ramp, a second test signal is applied to the secondary actuator and the RPS is recorded. The second test signal for the dynamic characteristics measurement is a series of test signals, each a sinusoidal signal at a constant frequency. The RPS is detected during the application of the constant frequency test signal, and the resulting gain and phase of the response are recorded along with the corresponding frequency. This is repeated for each frequency in the series of test signals. This enables the plant frequency response of the secondary actuator to be measured. If the response is significantly different from the expected response, it is virtually assured that the secondary actuator has failed. If the measured frequency response shows minor changes, such as a minor increase or decrease in the gain, or a minor shift in the frequency at which the maximum gain occurs, the controller parameters are adjusted or re-optimized. This re-optimization changes the values of the controller parameters in the memory accessible by the servo control processor. The parameters that can be changed include parameters that affect bandwidth or stability margins, notching of particular frequencies such as the secondary actuator resonant frequency, active damping of the secondary actuator resonance, or other performance, robustness, or stability metrics.

[0052] While the invention is fully applicable to a dual-stage actuator disk drive with a servo control system having a conventional dual-stage controller as described above, the invention will be described in detail below as implemented in the dual-stage actuator disk drive with the improved servo control system of the previously-cited co-pending application Ser. No. 10/802,601.

[0053] FIG. 5 is a block diagram of the control system of the present invention for a dual-stage hard disk drive. The head 109 reads the servo pattern from the disk, the read/write electronics 113 processes the signal from the head 109, and the servo electronics 112 generates the PES from the signals from read/write electronics 113, all as described in the prior art.

[0054] The servo control processor 400 receives the PES from servo electronics 112, and provides a primary control signal 191 to VCM driver 192 and a secondary control signal 229 to microactuator driver 230. The servo control processor includes a microprocessor 117 and uses a dual-stage controller 410 to generate control signals 191, 229. The dual-stage controller 410 incorporates a degraded-stability VCM controller with relatively high low-frequency open-loop gain, and a secondary actuator controller providing stability to the dual-stage controller and high mid-to-high-frequency open-loop gain, resulting in increased bandwidth. However, if the microactuator 200 fails while the disk drive is under the control of dual-stage controller 410, then VCM 110 will become unstable. If the microactuator 200 fails then the servo control processor 400 switches to use of a single-stage stable controller 420 and generates only a primary control signal 191 to VCM driver 192. The single-stage controller 420 can be a VCM controller based on the frequency response 210 (FIG. 2) or any VCM controller that is stable without the microactuator 200. This stable VCM controller 420 will most likely have decreased performance, but will prevent catastrophic failure of the disk drive that would result in loss of data.

[0055] In the preferred embodiment, the detection of potential failure of microactuator 200 is by a microactuator relative-position sensor 240. The sensor 240 measures the displacement of microactuator 200 relative to its neutral position (and thus to VCM 110) and provides a relative-position signal (RPS) to servo control processor 400. If the secondary actuator is an electrostatic microactuator, then sensor 240 can be a capacitance sensing circuit, as described in M. T. White and T. Hirano, “Use of the Relative Position Signal for Microactuators in Hard Disk Drives”, Proceedings of the American Control Conference, Denver, Colo., Jun. 4–6, 2003, pp. 2535–2540.

[0056] FIG. 6 is the open-loop frequency response 412 for the dual-stage controller 410 with a degraded-stability VCM controller design compared with the frequency response 210 for the single-stage stable controller 420. As frequency response 412 shows, the low frequency gain may be increased by relaxing the stability requirements of the VCM, but at the expense of robustness. The resulting open-loop frequency response of the dual-stage system 412 has increased gain over a wider frequency compared to the open-loop frequency response of the dual-stage system 220 shown in FIG. 4. This will result in better disturbance rejection and performance. The frequency response of the dual-stage system 412 has gain and phase margins that are comparable to the frequency response of the single-stage system 210 in FIG. 2. The phase margin near 2.8 kHz is about 30 degrees and the gain margin near 3.8 kHz is about 5 dB.

[0057] FIG. 7 is the gain portion of three frequency responses. Solid line 210 represents the gain of the VCM open-loop frequency response for a stable single-stage VCM controller. Dotted line 220 represents the improvement to response 210 with the addition of the secondary actuator (microactuator 200) and is the typical shape for a conventional dual-stage controller. It has increased bandwidth and increased gain in the mid-frequency range. This will result in improved disturbance rejection and faster response at these frequencies. However, because the low-frequency gain is still determined by the single-stage VCM controller, there is no improvement at low frequency. Cross-hatched line 412 represents the further improvement with the use of an increased-gain degraded-stability VCM controller and the secondary actuator. This response also has increased low-frequency gain, and is comparable in shape to the response for the VCM-only design, but shifted higher in frequency. However, increasing the low-frequency gain will also decrease the phase margin for the VCM controller, potentially to the point of instability of the VCM. The secondary actuator controller is then designed to make the combined system stable, as well as increasing the mid-frequency to high-frequency gain. Using the secondary actuator to ensure the stability of the combined system typically takes significantly less stroke than using the secondary actuator to increase the low-frequency gain, and is therefore a more efficient use of the limited secondary actuator stroke to achieve high bandwidth with adequate disturbance rejection.

[0058] With the dual-stage controller having the characteristics represented by line 412 in FIG. 7, failure of the secondary actuator results in an unstable system. This could lead to inoperability of the hard disk drive, or even failure with loss of data. This is avoided by detecting a potential failure of the secondary actuator and switching to a stable VCM-only controller. The potential failure of the secondary actuator is detected either by providing a model of the dynamic response of the primary and secondary actuators and comparing the modeled head-position with the measured head-position, or by measuring the position of the secondary actuator relative to the primary actuator with the relative-position sensor 240 and comparing the relative position to a modeled position of the secondary actuator. In the present invention, after a potential failure of the secondary actuator has been detected and switching to the VCM-only controller has occurred, the secondary-actuator failure-detection and calibration test is run by the servo control processor.

[0059] A schematic structure of the control system of the present invention is shown in FIG. 8. C_MACT and P_MACT represent the microactuator controller and plant, respectively, and C_VCM and P_VCM represent the VCM controller and plant, respectively. The controllers C_MACT and C_VCM together represent the dual-stage controller 410. The microprocessor 117 in servo control processor 400 (FIG. 5) runs the control algorithm using the parameters of controllers C_MACT and C_VCM and generates control signals u_MACT and u_VCM (229 and 191, respectively, in FIG. 5). The control system includes a model 430 of the microactuator plant and a model 440 of the VCM plant. These models may be determined from frequency response measurements of the microactuator and VCM, finite element models (FEM), or other well-known system identification techniques.

[0060] FIG. 8 shows two methods for determining potential failure of the microactuator 200. In the preferred method the calculated microactuator control signal u_MACT is input to the microactuator model 430 and the estimated microactuator position y_MACT(EST) from the model is compared to the RPS from sensor 240 at junction 450. In an alternative “PES-based” method the calculated microactuator control signal u_MACT is input to the microactuator model 430 and the calculated VCM control signal u_VCM is input to the VCM model 440. The modeled expected or estimated output y_EST is then compared with the measured output y_MEAS at junction 460.

[0061] FIG. 9 is a flow chart for the operation of the disk drive of the present invention. The flow chart portion 500 describes the method for detecting potential failure of the secondary actuator and switching to the VCM-only stable controller if potential failure is detected, and flow chart portion 600 describes the method for detecting actual failure of the secondary actuator and possible re-optimization of the controller parameters to enable switching back to the dual-stage controller having a degraded-stability VCM controller.

[0062] Referring first to portion 500, the control system starts (block 501) and continues to operate using the dual-stage controller 410 with the degraded-stability VCM controller (block 505). In block 510, the position of the head is measured (y_MEAS) if the PES-based method is used, or the relative position of microactuator 200 is measured (RPS) if the relative position sensing method is used. In block 515 the expected or estimated head position y_EST is calculated from the models 430, 440 if the PES-based method is used, and the expected or estimated relative position (y_MACT(EST)) is calculated from microactuator model 430 if the relative position sensing method is used. The difference (DIFF) is then tested to see if it is within pre-determined bounds (block 520). If yes, the control continues (block 505).

[0063] If DIFF is outside the bounds, this indicates potential failure of the microactuator 200. Once a potential failure of the secondary actuator has been detected the servo control processor 400 recalls the stable VCM controller 420 (FIG. 5) from memory (block 525). The processor sends a location signal to the VCM 110 to move the head to the test location, where further movement of the VCM is substantially limited. In the preferred embodiment, the test location is a crash stop, so the VCM 110 is then biased to maintain its position at the inner diameter with the ID crash stop 102 compressed (block 530). In this way, the position of the VCM is essentially fixed. Because the VCM is biased against the crash stop, the effects of other sources of error are virtually eliminated. This is a safe and accurate way to determine the motion of the microactuator in a condition that is virtually isolated from the VCM. The operation then moves to the steps described in flow chart portion 600. In an alternative embodiment, if the disk drive is a load/unload disk drive, the processor sends a signal to VCM 110 to move to the ramp 105 and the VCM 110 is then biased so that it maintains its position at the ramp 105. The following description of the invention is made with a crash stop being the test location, but the invention is fully applicable if the load/unload ramp is the test location.

[0064] Referring to portion 600, the secondary-actuator static characteristics measurement is performed (block 605). With the position of the VCM 110 biased against the ID crash stop, a test signal is applied to the microactuator 200 to calculate its stroke. At the check point (block 610) if the stroke is two low or too high, as compared to predetermined acceptable stroke values, then the secondary actuator has failed. The secondary actuator is disabled by selecting the VCM-only controller and an error is posted to the disk drive system (block 615). If the calculated stroke meets the acceptable criteria, then the secondary-actuator dynamic characteristics measurement is started (block 620).

[0065] The secondary actuator dynamic characteristics measurement measures the secondary actuator plant frequency response. At the check point (block 625) if the measured response is significantly different from the expected result, then the secondary actuator is disabled by selecting the VCM-only controller and an error is posted to the disk drive system (block 630). If the measured response shows only minor changes, such as a minor gain increase or decrease, or a minor shift in the peak frequency, then the secondary actuator can be recovered from potential failure (block 635). The controller parameters are then adjusted or re-optimized (block 640). The dual-stage controller with the degraded-stability VCM controller and the new secondary actuator controller parameters is re-selected (block 505) and operation continues. As an additional feature, a counter counts the number of times the controller parameters are re-optimized, which indicates the number of times a failure of the secondary actuator has been detected but both the static and dynamic characteristics measurements have been successful. If this count exceeds a certain threshold within a certain time period, indicating that failures are becoming too frequent, then subsequent modification of the controller parameters is terminated, an error is posted and the secondary actuator is disabled (block 630).

[0066] FIG. 10 is a schematic showing the rotary electrostatic microactuator 200 (FIG. 3A), its associated driving circuitry, and the relative-position sensor 240 that provides the RPS used in the secondary-actuator failure detection and calibration test of the present invention. Because the output force of an electrostatic microactuator is proportional to the square of the voltage difference between the stationary and movable electrodes (53 and 56; 53′ and 56′), it is desirable to linearize the voltage-vs.-force relationship. FIG. 10 shows the common differential driving method. In this method, an analog control voltage x is the input to the system. This voltage is processed in two ways, and applied to the two input terminals of the microactuator. In one path, a fixed bias voltage of A is added, followed by fixed gain amplification G by a high voltage amplifier, resulting in a voltage G*(A+x), which is connected to the microactuator's V1 terminal for the stationary electrodes 53. In another path, the analog input signal x is inverted, and the same fixed bias voltage A is added, and amplified by the same fixed gain G, resulting in a voltage G*(A−x), which is connected to the microactuator's V2 terminal for the stationary electrodes 53′. Since the voltage V1 generates counter-clockwise torque and the voltage V2 generates clockwise torque, the net torque will be the difference of these two torques, which is proportional to
(G*(A+x))²−(G*(A−x))²=4G²Ax.
The result is that the torque is linear to the input control voltage x.

[0067] The relative-position of the microactuator 200, i.e. the position of rotatable frame 54 relative to its neutral position, can be determined by modeling the electrostatic microactuator as two variable capacitors. The capacitance between electrodes 53 and 56 is equal to the capacitance between electrodes 53′ and 56′ when the frame 54 is at its neutral position. When the frame 54 moves from its neutral position, one of the capacitances increases and the other capacitance decreases. A sensing signal V_s*sin(w_s*t) is added to one side of the drive signal G*(A+x) and subtracted from the other side of the drive signal G*(A−x), as represented schematically by an oscillator 242 connected to the two drive signal input lines to respective terminals V1, V2. The oscillator 242 generates this fixed, small amplitude, high frequency (e.g., +/−1V, 2 MHz) sensing signal, which is applied to the two capacitors of the microactuator. When there is any imbalance between the two capacitors, meaning that the microactuator has moved away from the neutral position, a current with a frequency the same as the sensing signal frequency will appear at the middle point of the two capacitances to ground (shown schematically as AC current meter 243). The amplitude of the signal at this particular frequency is proportional to the capacitance imbalance, which is again proportional to the position of the microactuator. The amplitude of the current signal at the frequency of the added sensing signal frequency is detected by amplitude demodulator 244 and the output is the relative-position signal (RPS). The amplitude is zero when the two capacitances are equal. The use of a capacitance sensing circuit as a relative-position sensor for an electrostatic microactuator is described in detail by M. T. White and T. Hirano, “Use of the Relative Position Signal for Microactuators in Hard Disk Drives”, Proceedings of the American Control Conference, Denver, Colo., Jun. 4–6, 2003, pp. 2535–2540.

[0068] The above-described method for relative-position sensing of an electrostatic secondary actuator uses the force-generating elements for position sensing, i.e., the sets of stationary and fixed electrodes. If the secondary actuator is a piezoelectric actuator, the force-generating piezoelectric material, such as lead zirconium titanate (PZT), can be used for position sensing by monitoring the charge. In addition, relative position of a secondary can be sensed by integration of standard displacement using known sensors, such as potentiometers, strain gauges, encoders, capacitance probes, and piezoelectric material.

[0069] In the present invention the secondary-actuator failure detection and calibration test comprises two measurements: a measurement of the secondary actuator static characteristics and a measurement of the secondary actuator dynamic characteristics. The static characteristics measurement checks the secondary actuator movement amount per some constant first test signal, and the dynamic characteristics measurement checks the plant frequency response of the secondary actuator movement against a second test signal at one or more frequencies. The first and second test signals generated in the event of failure of the secondary actuator are added to the drive signals, as described above for FIG. 10, and the RPS generated by relative-position sensor 240 in response to these test signals is used in the measurements of the secondary actuator static and dynamic characteristics.

1) Secondary Actuator Static Characteristics Measurement

[0070] FIG. 11 is a flow chart showing the details of the secondary actuator static characteristics measurement (block 605 in FIG. 9). The secondary actuator input is initially set as a neutral bias. A constant input voltage is then applied as a positive bias to the secondary actuator (block 605A) and the relative position is measured from the RPS (block 605B) as described above. Next a constant input voltage is applied as a negative bias to the secondary actuator (block 605C) and the relative position is measured from the RPS (block 605D) as described above. The measurements (blocks 605B and 605D) are done multiple times and the average is calculated to get an accurate measurement. The secondary actuator stroke is then calculated (block 605E) as the measured movement of secondary actuator for the known positive and negative voltage inputs. The calculated stroke is then compared with acceptable stroke values at check point 610 (FIG. 9).

2) Secondary Actuator Dynamic Characteristics Measurement

[0071] FIG. 12 is a flow chart showing the details of the secondary actuator dynamic characteristics measurement (block 620 in FIG. 9). This measurement is performed after the static characteristics measurement. The secondary actuator input is initially set as a neutral bias. Next the first frequency of the secondary actuator test signal is selected and a sinusoidal input is made to the secondary actuator at the first frequency (block 620A). Generally the amplitude of the sinusoidal input signal is held constant during the entire measurement. Then the RPS is detected for several samples and the secondary actuator relative position calculated (block 620B). FIG. 13 shows a typical test signal at a fixed frequency and a typical RPS response. (No units are listed on the vertical axis because FIG. 13 is intended to merely show the representative shapes for a test signal and the RPS response.) The magnitude of the RPS response is shown as the peak values, and the phase can be roughly calculated as the difference between the zero crossings of the test signal and the RPS response. However, for a more accurate calculation of the gain and phase at the test signal frequency, a discrete Fourier transform (DFT) is applied to the RPS values at the input test signal frequency to calculate the gain and phase (block 620C). Then the gain and phase for that input test frequency are stored in a table (block 620D). The next input frequency is selected for the test signal (block 620E) and the process is returned to block 620A. The steps from blocks 620A to 620E are repeated until all desired frequencies are tested. For the convenience of the DFT calculation, multiples of the disk rotational frequency may be selected. After all desired frequencies have been tested, the table is searched for the maximum gain value and its corresponding frequency (block 620F).

[0072] After the maximum gain value has been obtained, the check point (block 625 in FIG. 9) is performed according to the details shown in FIG. 12. The maximum gain value is checked to determine if it is within an acceptable range of predetermined values (block 625A). For example, if the maximum gain is too low, e.g., not at least 20 db over the static gain, this is an indication of actual failure of the microactuator (block 630 in FIG. 9). There are potential microactuator failure modes in which the motion is excessive, in which case the maximum gain value may be too high. If the maximum gain is not outside the range of acceptable gain values, then the measured frequency at which the maximum gain occurs is compared to a predetermined acceptable range (block 625B). For example, if the maximum gain occurs at more than 1 kHz from the known resonant frequency of the microactuator, this is an indication of actual failure of the microactuator (block 630 in FIG. 9). However, if the frequency at which the maximum gain occurs is within acceptable limits, then the controller parameters can be re-optimized (block 640 in FIG. 9). This re-optimization may take the form of a table lookup of pre-calculated parameters or may be performed in real-time through the use of an adaptive scheme. The controller may be adjusted to achieve such features as desired bandwidth or stability margins, notching of particular frequencies such as the microactuator resonance, designing for active damping of the microactuator resonance, or other performance, robustness, or stability metrics. However, if there is only a minimal or no substantial difference from the optimum values, then control can be returned to block 505 (FIG. 9) without re-optimizing the controller parameters.

[0073] FIG. 14 shows one example of the plant frequency response for a typical micro-electro-mechanical system (MEMS) type microactuator located between the suspension and the slider. The motion of this type of microactuator is about 1 micrometer for a 30 V input at low frequency. So for the static characteristics test, a 0.5 micrometer movement would be expected for a 15 V input. If the measured output was less than about 0.2 micrometer or more than about 1 micrometer during the static characteristics measurement, then this would likely indicate failure of the microactuator. In a server class disk drive, the spindle motor frequency is between 167 Hz and 250 Hz, and the servo sector sampling frequency is currently 50 kHz or above. So for the dynamic characteristics measurement, an acceptable appropriate frequency range for the series of input test signals would be from about 200 Hz to 10 kHz. Thus as one example, the first sinusoidal test signal of FIG. 13 could be at 167 Hz with subsequent test signals increased in 167 Hz increments until the last test signal is reached at around 10 kHz. This type of microactuator also has a resonant frequency around 2 kHz, so if the maximum gain from the dynamic characteristics measurement is less than about 1 kHz or greater than about 3 kHz, then this would likely indicated failure of the microactuator.

[0074] While the present invention has been particularly shown and described with reference to the preferred embodiments, it will be understood by those skilled in the art that various changes in form and detail may be made without departing from the spirit and scope of the invention. Accordingly, the disclosed invention is to be considered merely as illustrative and limited in scope only as specified in the appended claims.