HARD DISK DRIVE OPERATION PROCEDURE, CONTROL CIRCUIT, HARD DISK DRIVE AND CORRESPONDING COMPUTER DEVICE

Description

DESCRIPTION of the industrial invention entitled: “Operation procedure of hard disk drives, control circuit, hard disk drive and device corresponding processor” by: STMicroelectronics International NV, nationality Dutch agent through head office and branch Switzerland located at Chemin du Champ-des-Filles 39, 1228 Plan-les-Ouates Geneva, Switzerland. Designated Inventors: Michele BOSCOLO BERTO, Maurizio RICCI, Marco FERRARI, Diego TOGNOLI, Giuseppe MAIOCCHI. Filed on: May 24, 2024 **** DESCRIPTION TEXT Technical field The description is for hard disk drives (HDD, “Hard Disk Drive”). Solutions as described here can be applied, for example, in hard disk drives used in processing devices such as computers or computers, servers, data centers and the like. Description of the related technique Hard disk drives (HDD) include so-called moving coil motors (VCM motors or, for short, VCM (“Voice Coil Motor”)) which are used to position the coil heads reading and writing. When a HDD is in operating mode, a VCM engine is used to position the read and write heads above the surface of the hard drive where they are stored data. When the HDD is not in operating mode, the read and write heads are positioned (park) in a safe and secure area using a retraction (or parking) procedure. The speed of the heads (and thus of the VCM) during the parking phases (retraction) can be controlled in a discontinuous driving mode with on times (of duration TON) where energy is supplied to the alternating motor with off times (of TOFF duration) in which no power is supplied energy to the engine. During an on phase, a current is made to flow flows in the motor winding(s) and generates an engine torque; during an off phase, the current is interrupted and, if the speed is not zero, at motor winding heads (windings) generates a counter electromotive force (BEMF) voltage “Back Electro-Motive Force”). The BEMF voltage is (directly) proportional to the engine speed VCM and therefore it can be used as a feedback signal in a speed control ring. In modern HDDs, a parking area is provided of the heads located on the outermost part of the disk (currently referred to as OD or outer diameter (“Outer Diameter”)) where a ramp facilitates parking in head safety. In a retraction phase, the heads should hopefully move from the surface of the disk to the parking position at the end of the ramp to a controlled (low) speed. The path of movement to the position of parking includes critical points where control of speed may not have a bandwidth (BW, “BandWidth”) wide enough to facilitate the compliance with speed specifications in terms of minimum / maximum speed. For example, in traditional retraction techniques discontinuous which involve a constant off time of TOFF duration, bandwidth cannot be extended to high frequencies as a minimum duration is desired of the off time (and an on time associated with it for get a duty cycle value as set) long enough to facilitate measurement accurate BEMF. The voltage measured across the VCM motor when the current is interrupted during an off time in a discontinuous driving mode is affected by a residual voltage generated by the sudden interruption of the current. This residual voltage negatively affects on the measurement of the BEMF (measurement of velocity) and therefore on HDD performance during retraction and is function of the current amplitude during the on phase which precedes the instant of time at which the velocity of the VCM is measured by BEMF. You can then select a value for the duration of the off time TOFF (to be kept as a fixed value) during the entire retraction procedure) considering a worst case situation, i.e. the current condition maximum, which usually coincides with the end point movement (EOT, “End-Of-Travel”). The fact of selecting a value for the TOFF duration which is too short hinders proper control action of the VCM, for example in detecting the EOT condition. This can lead to various unwanted consequences: for example, acoustic noise can be generated in response to the fact that the VCM reaches the EOT point due to an erroneous detection of the EOT in relation to the fact that a discontinuous piloting is maintained for a period of time excessive and / or the VCM speed can be estimated in incorrectly in those areas where the current has a relatively high amplitude. Purpose and summary One purpose of the solutions described here is to help address the issues discussed in precedence. According to one or more forms of implementation, this purpose is achieved with a procedure as set out in claims that follow. Solutions as described here are related to a corresponding controller circuit for disk drives rigid. Solutions as described here are for a unit to the corresponding hard drive. Solutions as described here are related to a corresponding processing device equipped with one or more hard disk drives. A computer, server, or data centers are examples of such a device. The claims are an integral part of the technical teaching provided here in relation to the embodiments as provided herein. In solutions such as those described here, a discontinuous mode retraction approach with a variable operating frequency which can be modulated as a function of the driving current. This facilitates an increase in bandwidth of the speed control ring: in fact, yes note that, during a phase of retraction of the heads when reading and writing an HDD, a point occurs critical in the initial portion of the ramp provided for park the heads, where a current of relatively low piloting. The speed control in that portion can be improved by increasing the operating frequency and therefore the obtained bandwidth (BW) of the ring of speed control. It is noted that another critical aspect of a phase of retraction consists in properly detecting the point of end of journey (EOT). In the EOT area, the currents are relatively high and an accurate measurement of the BEMF would involve a relatively long TOFF duration of off times. In such area, a limited BW bandwidth of the ring speed control is not particularly critical and a reduction in the frequency of can be tolerated operation during discontinuous driving (and of resulting in a reduction in the BW of the ring speed). Solutions such as those described here facilitate a reduction of the speed variations during an ascent of the ramp and correct detection of the end point displacement (EOT) by modulating the duration of the off time TOFF (and thus the operating frequency) during discontinuous piloting) which can be a function of the amplitude of the driving current. Brief description of the drawings Solutions will now be described here, purely for information purposes. for example, with reference to the attached figures, in which: Figure 1 is a block diagram of a disk drive hard disk drive (HDD) equipped with a reconstruction circuit of the BEMF (“BEMF monitor”) used for the reconstruction of the BEMF during VCM piloting; Figure 2A illustrates one possible behavior (a possible waveform) in time of a current in a VCM engine during a discontinuous pilot; Figure 2B illustrates one possible behavior (a possible waveform) in the corresponding time of a differential voltage across a VCM motor during a discontinuous piloting; Figure 3A is an example of a possible trend of VCM peak current values ​​used during a test to measure the residual voltage; Figure 3B is an example of a possible trend of the VCM differential voltage used during a test to measure the residual voltage; Figure 4 is a schematic representation of the structure of an HDD showing a movement path of the read and write heads during a procedure of retraction (parking); Figure 5A illustrates one possible behavior (a possible waveform) in time of a current in a VCM engine during a discontinuous piloting in a phase of retraction, i.e. during a displacement of the VCM from one position of internal diameter (ID, “Inner diameter”) to a outside diameter (OD) position; Figure 5B is a more detailed (zoomed) view of a portion of Figure 5A where the VCM speed, with the zoomed area showing the current and the velocity of the VCM across the edge of the ramp; and Figure 6 is an illustrative flowchart of a possible sequence of steps in a procedure like here described. Unless otherwise indicated, numbers and symbols corresponding in the different figures refer to generally to corresponding parts. The figures are drawn to clearly illustrate the relevant aspects of the forms of implementation and are not necessarily drawn to scale. The edges of the features drawn in the figures do not necessarily indicate the boundaries or extent of the characteristics. Furthermore, for simplicity and ease of explanation, a same designation can be applied to this description to indicate a line or a node of a circuit as well as a signal that occurs on that line or at that node. The quantitative values ​​indicated in any of the figures attached here (or obtainable from them) are purely indicative These are examples and are not limiting. Detailed description In the following description one or more are illustrated specific details, in order to provide an understanding in-depth analysis of examples of implementation forms of this description. The embodiments can be obtained without one or more of the specific details, or with other processes, components, materials, etc. In other cases, Known operations, materials or structures are not illustrated or described in detail so that certain aspects the forms of implementation will not be made unclear. A reference to “a form of implementation” in the framework of this description intends to indicate that a particular configuration, structure, or feature described with reference to the form of implementation is included in at least one embodiment. Therefore, the phrases like “in an embodiment” or similar that may be present in one or more points of this descriptions do not necessarily refer to the actual to the same form of implementation. Furthermore, particular configurations, structures or features can be combined in any suitable way in one or more forms of implementation. The references used here are provided simply for convenience and therefore do not define the scope of protection or the scope of the forms of implementation. In all figures attached hereto, unless the context indicate otherwise, similar parts or elements are indicated with similar references / numbers, without repeating for brevity a corresponding description. Again, for simplicity and ease of explanation, the same designation can be applied throughout this description to indicate: a certain node or a certain line as well as a signal occurring at such node or on such line, and / or a certain component (such as a capacitor, a resistor or an inductor of a coil) as well as its electrical parameters. As discussed in the introductory part of this description, HDD is an abbreviated designation for hard disk drive hard disk (“Hard Disk Drive”). A hard disk drive is a component of various types of computing devices such as personal computers, servers, data center or similar and is the physical location where they are store the information. Hard disk drives (HDDs) use a motor of the shaft or spindle to rotate one or more disks as well as one or more moving coil motors or VCMs to make move the heads relative to the disk(s). That is, say, the spindle motor is used to turn the hard disk(s) on which the data is stored data and the VCM engine or engines is / are used for place the read / write heads on the surface of the hard disk(s). Hard disk drives thus comprise at least two motors (spindle and VCM) which are driven by devices power semiconductor (integrated circuits) indicated by usually as a power combo. In current HDDs, a ramp is provided on the side outside of the disk (outer diameter, OD); this ramp is used to position the read and write heads when not in use (i.e., when the heads are in the parking position). The procedure implemented to move the heads from a parking position to a parking position operation on the surface of the disk is indicated as “VCM loading”; the reverse procedure, i.e., move the heads from the disk surface to a parking position is indicated as “VCM unloading”, “VCM retraction” or “VCM parking”. Carrying out loading and unloading procedures at a controlled speed makes it easier to avoid a damage to the heads or the disc surface (scratching of the disk). As shown in Figure 1, in applications HDD, MCU servo controller can be provided for adjust the VCM drive current as needed you want, for example, to keep the heads on the track (on track) and / or to facilitate an execution adequate “VCM loading” and “unloading” procedures of the VCM” / “VCM retraction”. This can be done by a (speed) control word provided by the MCU controller. To control the speed of the VCM you can use a feedback signal. Voltage (Counter Electromotive Force or BEMF) generated by the VCM engine during movement can be used for this purpose, to the extent that the BEMF is (directly) proportional to the VCM speed. This general principle underlies the functioning of a VCM in a HDD is conventional in technique and for For the sake of brevity, a more detailed description is not provided here. detail. Furthermore, it will be appreciated that this description It mainly concerns the fact of piloting in a suitable a VCM engine in a discontinuous mode while also facilitating accurate measurement of the BEMF. As noted, in disk drive applications hard disk (HDD) there are at least two motors (spindle and VCM) that are driven by devices (integrated circuits) of power: these devices are currently indicated as a power combo. The spindle motor is used to rotate the hard drives on which data is stored, and the VCM motor is used to position the read heads and writing on the surface of hard drives. On the outside of the disc (outer diameter, OD) of today's HDDs there is a ramp that is used for place the heads, when not in use, in a position of “parking”. The procedure performed to move the heads from the parking position to the operating position on the disc surface is currently referred to as “load of the VCM” while the complementary procedure, i.e. the movement of the heads from the disk surface to the parking position, is currently indicated as “VCM unloading”, “VCM retraction” or “VCM parking” VCM”. In order to avoid unwanted damage to the heads or to the disc surface (disc scratch) the loading and unloading procedures are carried out at a controlled (low) speed. To control the speed of the VCM based on a feedback signal you can use a control loop of speed. Advantageously, a signal of voltage generated by the VCM motor during movement (the counter-electromotive force (BEMF) since this is directly proportional to the speed of the VCM. Figure 1 is a block diagram of a 100 circuit which can be used for a reconstruction of the Force Against Electromotive, BEMF while driving a Moving Coil Motor, VCM, indicated with the reference 120 and represented as a (voltage) source of the Counter Electromotive Force, BEMF with associated an inductor L and a resistor R (in series). MM A BEMF 100 reconstruction circuit as illustrated in Figure 1 can be used when the loading / unloading procedures are carried out in mode continues, where there is no power outage that allows you to read the BEMF directly through the VCM terminals. A BEMF 100 reconstruction circuit as illustrated in Figure 1 includes input nodes VCMSNSP and VCMN coupled through a VCM 120, plus at least an additional input node VCMSNSN with a resistor R S coupled between the VCMSNSN and VCMSNSP nodes. The voltage through the resistor R (the voltage drop between the S (VCMSNSN and VCMSNSP nodes) is a measure of an IM current that flows towards the VCM 120 starting from a node of VCMP power supply. A first sum node SUMM1 (signed) produces a signal based on the difference between the voltages at the nodes VCMSNSN and VCMSNSP, which is thus a measure of the current I M and is amplified by a first 110 gain stage to produce an A1 signal. A second (signed) sum node SUMM2 produces a signal based on the difference between the voltages at the nodes VCMSNSP and VCMN, which is thus a measure of the fall of voltage across the VCM 120. A third (signed) sum node SUMM3 produces a signal based on the difference between the voltage at the output of the second sum node SUMM2 and the voltage A1. The The difference signal thus obtained is amplified by a second gain stage 130 to produce a signal A2 which can be converted to digital by means of a analog to digital converter 140. When carrying out loading / unloading procedures in discontinuous mode, some circuit 100 blocks not come into play because the current is zero. During a discontinuous drive off time, the A1 signal is zero (to the extent that R *I =0) and the signal A2 is therefore SM equal to the BEMF (possibly scaled for a “pure” A2 gain factor introduced by block 130). In a discontinuous mode, the output of circuit 100 is so the “pure” BEMF. In fact, one advantage of this mode of discontinuous piloting consists in the absence of calibrations to reconstruct the BEMF (the gain of the block 110 is an indifferent (“no-care”)); on the other hand, one disadvantage is a bandwidth (BW) lower speed control compared to the Continuous piloting procedures. The control BW the speed is limited by the driving frequency discontinuous. The circuitry represented and indicated in the complex from reference number 100 in Figure 1 has associated with it - according to solutions that will be described later - two further blocks 160 and 180 (a logic block of retraction and a digital-to-analog converter, DAC (“Digital-to-Analog Converter”)) that provide a signal to be applied to a voltage (power) amplifier 200 used to provide voltage between VCMP and VCMN nodes to control the operation of the VCM in a mode discontinuous during a retraction phase. For the rest it is noted (this also applies to the description provided with reference to additional figures) that the VCM 120 may not be part of the circuitry for VCM control / reconstruction of the BEMF as supplied. This circuitry and the VCM can be connected only by the end user. When measuring the BEMF of a VCM, two approaches can be adopted: approaches: a first approach involves reconstructing (typically in analog form) the BEMF voltage in based on the characteristics of the VCM engine: see, for example, the resistor R and the detection circuit of M current, i.e., the resistance R as represented S in Figure 1; a second approach involves direct measurement of the BEMF through the engine terminals made after the flow of current is interrupted for a short time. Circuit 100 can be used in both approaches provided that: when using continuous mode, one operation adequate almost invariably requires calibration accurate, when using discontinuous mode, such accurate calibration is no longer required. The first approach can be used while the VCM is continuously driven: the current in the VCM motor is not never interrupted but is supplied continuously. The second approach can be used with the VCM driven discontinuously: the current in the motor is interrupted for a certain off time (duration TOFF) regular intervals to facilitate reading of the BEMF voltage (which is proportional to speed) directly to the VCM engine terminals. Usually it is preferable to control the speed of the VCM in a continuous mode versus a control discontinuous. However, continuous monitoring may not be always feasible as it involves calibration accurate and relatively complex circuit design BEMF reconstruction that can be performed adequately only if certain conditions are met conditions. The circuitry indicated as a whole by a number reference 100 in Figure 1 (again, leaving out the blocks 160 and 180 discussed below) is an example of a “BEMF monitor” which can be used for a BEMF reconstruction during continuous piloting but which requires calibration as the results from it supplied are influenced by the motor resistance R M (for example, it is observed that the gain A1 of the first gain stage 110 is determined by the ratio R / R ) and MS the resistance R is not constant but varies as the M temperature. Because of this temperature dependence, a circuitry indicated in the complex by the number of reference 100 in Figure 1 can be used adequately as a “BEMF monitor” if calibrated in a moment immediately preceding its use. This may be feasible for the loading and unloading phases. discharge under normal operating conditions of an HDD, but it may be difficult to do in those cases in which the heads are expected to be parked in response to a critical emergency event (for example: a power supply failure). In the latter case, a head parking can correspond to an emergency retraction phase that hopefully it should be implemented immediately and without any latency as possibly introduced by a calibration phase. Furthermore, in emergency conditions, the or MCU of they are usually not active and therefore not able to support a calibration and / or emergency retraction using of firmware (FW) routines. In emergency situations, there expects the entire emergency retraction to be made using integrated hardware (HW) procedures (“embedded”) in the “power combo” device. In the event of such an emergency retraction, one can use a discontinuous mode pilot since this facilitates a measurement of the BEMF of the VCM engine directly to the VCM motor terminals, without any interference no calibration is involved. Figures 2A and 2B represent, with respect to a scale of time t in common abscissa, possible behaviors (waveforms) over time: a current I through a VCM motor having a M peak PC value (Figure 2A); and a VCM voltage of the VCM across the same motor V VCM under zero VCM velocity conditions (Figure 2B). Figures 2A and 2B are both related to a discontinuous piloting in which the off times and the on times they have TOFF and TON durations of 450 us and 1 ms respectively. As noted, these quantitative values ​​are purely by way of example and are not limiting. The actual VCM voltage of the VCM is actually out V scale (see OOS arrow pointing up) in the Figure 2B, which is intended primarily to illustrate (in (a dotted line) a possible behavior in the time of the residual voltage RV, also with a large out of scale extension (see OOS arrow pointing downwards) with reference to a voltage level zero 0V. For the rest, it is noted that the real BEMF is zero in I am finding that the VCM is stopped during a test. As already discussed, during a discontinuous piloting unwanted side effects may occur which are related to the presence of a residual RV voltage through the load presented by the VCM engine in response to the fact that the current through it is interrupted abruptly as is the case during a discontinuous piloting (This phenomenon is highlighted in Figure 2B). As visible in Figure 2B, the residual voltage RV it wears off naturally over time. Its unwanted effects can thus be mitigated (as is traditional in technique) by measuring the BEMF of the engine VCM at the end of the off time, with duration TOFF, using relatively long values ​​for the TOFF duration (typically hundreds of microseconds). The fact of using relatively long values ​​for the TOFF duration negatively affects VCM control as these long values ​​can give rise, for for example, at a higher ripple speed and at a limited control loop bandwidth (BW) of speed. Furthermore, the fact of using relatively long values for the duration TOFF produces an intrinsic long latency between the change in speed of the VCM (for example due to (a collision with the ramp) and the reaction of the control speed. This is visible in Figure 5B (discussed further below) where reference VI indicates a speed sampling position at the end of an off time of duration TOFF and the reference VII indicates the latency between two consecutive speed measurements introduced by the low frequency of discontinuous driving. Figure 5B (primarily references VI and VII) does understand that a more discontinuous frequency is desirable high in order to counteract an unwanted drop in VCM speed during a “collision” with the ramp. In turn, the fact of reducing the TOFF duration of the off times to limit these undesirable effects almost inevitably leads to other effects as a result unwanted, such as an incorrect estimate of speed actual VCM engine power, distorted by the presence of the residual voltage RV (whose amplitude - as visible in the Figure 2B - may be far from negligible). Italian patent application No. 102023000012570 (Michele BOSCOLO BERTO, inventor), not yet available to the public at the time of filing this application, describes a possible way to address this problem compensating for residual tension. This approach facilitates the use of off times having a (very) short TOFF duration with some implementation costs. In fact, for the same TOFF duration of time off, the residual voltage (RV in Figure 2B, for example) It is proportional to the amplitude of the current in the VCM motor during the conduction phases (TON duration on times) of a discontinuous piloting. The function that relates the amplitude of the current during the on times with the amplitude of the residual voltage during off times is nonlinear: the solution described in Italian patent application no. 102023000012570 (already mentioned) involves compensating the residual stress based on a sort of “linearization” of this relationship. Figures 3A and 3B represent, with respect to a scale in common abscissa (in terms of the number N of tests in a sequence) a possible “trend” of the peak current of the VCM (Figure 3A) and the differential voltage of the VCM (Figure 3B) measured 100 us after the onset of TOFF. It is noted again that the real BEMF is zero in response to the fact that the VCM is stopped. The BEMF generated by the VCM engine during the procedure of retraction can be, for example, from 200 mV to 300 mV (again, these quantitative values, as well as the values ​​indicated in / obtainable from Figures 3A and 3B, are purely by way of example and not as a limitation). The BEMF generated by the VCM engine during the procedure of retraction can thus easily go beyond these values ​​at some instants during a retraction phase, in especially when the current amplitude is relatively high. This may be the case if you select a TOFF value that is too short for the duration of the off times, noting also that the TOFF duration of the times of off cannot be easily reduced without penalizing the accuracy of the VCM speed measurement. In the solution described here, the TOFF duration of the off times (and thus the frequency of the pilot discontinuous) is made selectively adaptive, which makes it easier to address critical issues related to to a discontinuous retraction at a lower cost in comparison with alternative solutions. In addition to a “BEMF monitor” 100 as described previously, the circuit shown in Figure 1 also includes a retraction logic block 160 to which the available (reconstructed) BEMF signal is provided at the monitor output of BEMF 100 (ADC 140 converter) which is provided with a BEMF reference (BEMF reference) speed) in the form of a digital value BEMF_REF used to calculate the velocity error and to calculate the adequate voltage to control the speed of the VCM. Figure 1 also includes a converter digital / analog (DAC) 180 which can be considered provides a VCM driving voltage suitable for being applied between VCMP and VCMN nodes through a stage of power 200 voltage controlled. Power stage 200 can also be enabled and disabled through an enable signal (“enable”) En to implement a piloting mode discontinuous. The enable signal En is controlled from Retraction Logic block 160. As further discussed below, the solutions described here effectively address the related problems to a minimum desired value for the TOFF duration of the times off and facilitate a BEMF measurement by contrasting at the same time also the resulting limitation of the maximum operating frequency of the control loop speed, which in turn is connected to the residual voltage generated through a VCM engine driven in a mode discontinuous. Instead of compensating for residual tension, an approach the solutions described here are therefore based on the following: to exploit the (very) limited amplitude of the voltage residual at low driving currents. Solutions such as those described here therefore facilitate the use of short-duration TOFF values ​​for off times in I find that the driving current is relatively low (which facilitates the use of frequencies high for the speed control ring) using at the same time long off times (low frequencies) when the driving current is relatively high. It is observed that the current profile during a phase of retraction proves very suitable for solutions like here described. In fact, during the ramp up phase, where the read and write heads meet and move over the ramp, a width of is desirable high bandwidth (BW) in order to reduce variations in speeds due to an interaction (a collision) of the heads with the ramp. Under these conditions, the current in the engine can be relatively low, which facilitates a use of high frequencies (thus making it possible to have a large BW bandwidth for control loop of speed and detect speed changes with a reduced latency). On the other hand, during the final advance towards the final parking position, with a fine movement (EOT) finally detected, the current in the motor can be relatively high: this would in itself be in contrast with the use of high frequencies, but in these conditions do not require a large bandwidth and the bandwidth of the control loop speed can be limited accordingly. In situations where the VCM retraction is made in a discontinuous manner as a phase of emergency retraction, the fact of having a TOFF duration of the shortest power outage time possible is advantageous as it makes it easier to have high operating frequencies designed to limit the typical side effects of a check-up discontinuous: ripple speed and bandwidth (BW) limited speed control ring. A limited bandwidth of the ring speed control can be a problem in presence of sudden changes in mechanical torques applied outside the VCM that may occur when the read and write heads are done advance from the surface of the disc towards the position of EOT. For example, these external pairs are almost zero when the heads advance over the surface of the disk and can undergo a sudden change when the heads they meet (hit) the edge of the ramp. In response to these sudden changes, the VCM speed can undergo marked variations. As noted, these variations can be counteracted increasing the bandwidth of the control loop speed and reducing the latency between measurements consecutive speeds. However, this way of operating can be limited by the refresh rate of the speed measurement (which practically corresponds at the frequency of a discontinuous piloting). Figure 4 is a schematic side view of some elements in an HDD hard disk drive, including: D disks (only one is illustrated for simplicity) in which data is stored, a SM spindle motor to rotate (turn) the disc (discs), H read and write heads for reading data from and write data to the disk surface (of the discs) D, and a VCM moving coil motor to move the heads H on the surface of the disk(s) D and towards a ramp R (positioned in the outermost part of the disk (of disks) where the H heads are parked in a parking position. This HDD structure is also conventional in technique and for the sake of brevity, no details are provided this evening. more detailed description. The right side of Figure 4 represents a theoretical linear development (reproduced not to scale for simplicity and ease of understanding) of a possible profile of a parking ramp R. In Figure 4, the path of moving the H heads driven by the VCM motor towards their final parking position which includes four zones identified with the Roman numerals I, II, III and IV. Of these: Zone I represents the advancement of the H heads above the surface of the disc(s) D towards the ramp R; Zone II represents the H heads that meet (which hit) the edge RE of the ramp R adjacent to the disc surface; Zone III represents the advancement of the H heads above a flat area RFZ of the ramp R; and Zone IV represents the H heads that reach an end-of-movement (EOT) condition where they are parked (and possibly held by a ML magnetic latch for easy positioning stable H-heads in the parking position). Of course, Figure 4 provides a deliberately simplified representation of the ML magnetic latch for the sole purpose of showing its effect on the VCM: in fact, the ML magnetic latch can be placed in a different physical location (near the VCM motor pin, for example). Figure 5A represents a possible behavior (a possible waveform) in time of a current through a VCM during a retraction procedure made in a discontinuous mode in which they are identified four zones I, II, III and IV as discussed previously, thus making it possible to identify: an I-zone, where the H-heads move on the disc surface, a zone II, where the H heads collide and “rise” on the edge of the RE ramp, a zone III, in which the H heads advance on the part RFZ ramp plan; and a zone IV, where the H heads reach the point end of journey (EOT). Figure 5A also shows an area additional (indicated by reference V) in which the system detects EOT condition (EOT validation). This can occur in response to the detection of a speed of VCM which is zero (or, in general, is below a certain threshold) for a certain number of discontinuous cycles consecutive. The upper curve in Figure 5B is a reproduction zoomed (with enlarged scales in the abscissas and in the ordinates) of the curve in Figure 5A which describes in detail the behavior of the signal around a value 0 in zones I, II and III with particular attention to time when the H heads meet (hit) the R ramp. The lower curve in Figure 5B represents a possible behavior in the corresponding time (which shares the same time scale on the x-axis) VCM speed of the VCM. S As discussed, the upper curve in Figure 5B also represents possible sampling positions (of the velocity) of the BEMF located at the end of a time of off (indicated by arrow VI) and the latency between two consecutive BEMF sampling events (referred to as VII). In a worst-case scenario, a collision with the ramp (indicated by the arrow VIII in the lower curve in Figure 5B) occurs immediately after a speed sampling (arrow VI). In this case, the system detects speed drop with maximum delay. In this worst case scenario, the speed drop is maximized and can be limited (only) using a high discontinuous driving frequency aimed at reducing the latency between two consecutive speed measurements. As noted, the lower curve in Figure 5B is a (approximate) representation of instantaneous velocity VCM of the VCM that shares an abscissa time scale S enlarged common with the upper curve in the same figure. The lower curve in Figure 5B thus represents in further detail the speed behavior of the H heads (and the VCM) as the H heads “climb” on the ramp R. Again, the quantitative values ​​given in (or otherwise obtainable from) these figures are purely title for example and not for limitation. As can be understood from Figure 5A, once the EOT point is reached (zone IV), the VCM speed is reset to zero and the speed ring can react increasing the driving current up to a value maximum. As known to those skilled in the art, a procedure for detecting the end-of-displacement condition (EOT detected) may put an end to discontinuous piloting after a certain number of discontinuous cycles with the speed measured below a certain threshold (zone V). Figures 5A and 5B confirm again that, during the retraction procedure: you want to have a high drive frequency and a high control loop bandwidth of the speed at the point where the heads meet (hit) the ramp R, and the frequency and bandwidth requirements are less stringent in other areas where changes in external couples are less abrupt and are free from discontinuity. Figures 5A and 5B show that the current in the area (Zone I) before going up the ramp (Zone II) is quite small. Under these conditions, the residual voltage RV is correspondingly (rather) low. Here you can use (very) short off times with a short TOFF duration (and thus a high frequency and a low latency), which facilitates a quick reaction of the speed ring when the H heads meet (hit) the R ramp. While climbing the ramp (zone II), the current increases in order to counteract the loss of speed but the current value remains relatively low, which facilitates the use of off times having a duration TOFF “intermediate”, which can however be shorter than the (fixed) duration used in those systems where the TOFF duration of the off times is selected for the purpose of a accurate BEMF detection under all conditions possible, thus taking into account the maximum current set at the EOT point. In summary, Figure 4 is an example of a operating procedure of a moving coil motor, VCM 120 in a HDD hard disk drive during a recovery phase retraction in which the read and write heads H of the hard disk drive are retracted from the VCM 120 by a disc surface D towards a parking condition of end of EOT shift with H heads rising above a ramp R starting from one edge of the adjacent ramp RE to the surface of the disk D. During such a retraction phase, the VCM 120 is operated in a discontinuous mode comprising an alternation of on times of duration TON and off times off of duration TOFF in which the supply of a current of Piloting I at VCM 120 is made easier and more difficult, M respectively. This process of operating a diesel engine moving coil, VCM 120 in a HDD hard disk drive during a retraction phase it is compatible with those devices in which the intensity of the driving current I through the VCM 120 is controlled (in a known manner of M for themselves to experts in the technique: by means of a MCU microcontroller in “good” operating condition or using a dedicated HW integrated into the combo device power in emergency conditions) as a function of a Counter Electromotive Force, BEMF of the VCM. Figures 5A and 5B indicate the possibility of using short off times (and a bandwidth / frequency high speed control ring) in low current conditions (Zone I). More specifically, it is observed that the times of moving the heads through Zone I are substantially identical regardless of whether whether long or short TOFF off times are used so that no error is introduced by a small value for TOFF: a high bandwidth / frequency used in Zone I does not introduce a significant error due to to the low current conditions involved; and In Zone II, travel times can be different because the current is high there (and the voltage residual is correspondingly high): the time of shift with a long off time (this can be taken as a reference) is shorter than the case of a short off time and the residual voltage introduces a controlled speed error. Consequently, a small value for the TOFF duration is generally unsuitable to be used as a fixed value: However, it can be used when the current is small. For the rest, it is noted that the edge of the ramp RE is right at the end of a low current zone (displacement of the heads on the surface of the disk), so that it can a high frequency can be used to advantage (a large bandwidth of the control loop speed) when the VCM engine is moving in the Zone I and is approaching the R ramp. Solutions such as those described here facilitate a reduction of the speed variation at the most critical point (essentially Zone II in Figures 5A and 5B) in compared to those systems that for the duration of the TOFF time of off use a single (fixed) value that is selected almost inevitably as a long duration in order not to compromise a BEMF measurement, which has in its time as a result a relatively constant frequency low for intermittent operation. Figure 6 is an example of phases of a flowchart. flow in a procedure that implements a solution as described here. For example, such a procedure can be implemented in an MCU servo controller as shown in Figure 1 during normal operating conditions (when the MCU is active (“alive”) and functioning properly) or can be implemented in a dedicated integrated hardware in the power combo for those cases where the MCU is not operational as, for example, during conditions of emergency. In any case, the procedure can be configured (in manner known in itself to those skilled in the art) for adjust the VCM drive current as desired in order to carry out “VCM loading” procedures and “VCM unloading” / “VCM retraction”. This can be done by using a control word DAC provided by the MCU controller or via a word of DAC control provided by the logic block retraction 160 based on a BEMF voltage generated by the VCM engine during movement, to the extent that the BEMF is (directly) proportional to the VCM speed. In solutions as described here, a circuit of reconstruction of BEMF 100 as illustrated in Figure 1 is advantageously completed with the logic block of retraction 160 to which the BEMF signal is supplied (rebuilt) available at the output of the 100 circuitry (ADC 140 converter) plus a converter digital / analog (DAC) 180 that can be considered to provide a VCM voltage suitable to be applied between the nodes VCMP and VCMN through the 200 controlled power stage in tension. The flowchart in Figure 6 is an example of a procedure applicable in driving a VCM engine in a hard disk drive (HDD) in a mode discontinuous during a retraction phase in which the engine VCM is supplied with energy with a flowing current through it during on times (of TON duration) that they alternate with off times (of TOFF duration) in which VCM motor is not supplied with energy by current flow through it interrupted. In solutions as described here, the duration of such times are not fixed: for example, the length of the waiting times Power interruption (TOFF) can be modulated in in such a way as to be a function of the amplitude (of the intensity) of the current used to drive the motor. Possible options (generally not mutually exclusive) to achieve this result can understand, for example: The TOFF duration can be selected between two or more values, The TOFF duration can vary between a minimum value and a maximum saturation value, the TON duration of the energization times (excitation) (TON) can be determined as a function of the duration of the power outages, i.e. TOFF; for example, TON can be selected equal to K*TOFF where K can be, for example, 1.5 or 2, TOFF duration can vary between a minimum value (lower limit) and a maximum saturation value (upper limit). The Digital to Analog Converter (DAC) 180 can be configured to have a DAC gain of K[DAC] (normalized) (selectable between 0 and 1, i.e. a gain normalized from 0% to 100%) with a TOFF duration determined as: TOFF = MinTOFF*(1+K*[DAC]) where MinTOFF is a minimum value (lower bound) for TOFF and K*[DAC] is an adjustable boost factor TOFF duration steps of off times. For example, the DAC 180 can be made programmable to make K*[DAC] variable as a function of a word of Ktoff [1:0] input of two bits according to the following table of the truth. Ktoff[1:0] K*[DAC] TOFF 00 0 MinTOFF 01 1 TOFF = MinTOFF*(1+K*[DAC]) 10 2 11 4 A TON duration can be determined in a way corresponding as TON=2*TOFF (i.e., a 66% duty cycle). The ranges or absolute values ​​of the values ​​for TOFF and TON may depend on various factors / related parameters to the intended application (a desired target speed for retraction can be an example of such a factor / parameter). By way of non-limiting example, possible TOFF variation ranges can be from 200 us at 600 us (max TOFF at saturation), while possible variation ranges for TON can be from 400 us to 1200 us (max TON at saturation). TOFF values ​​near or at the upper limit can be selected in view of the intended application as to facilitate proper measurement of the BEMF. The blocks in the flowchart in Figure 6 are indicative of the following phases or actions: block 1000 – retraction started (the speed of the VCM is initialized) block 1002 – set a target speed value T and the velocity error is calculated block 1004 – DAC 180 triggered based on error speed block 1006 – TOFF duration calculation (taking into account of a possible saturation value) block 1008 – calculation of TON duration (taking into account of a possible saturation value) block 1010 – disk current pulse force forced block 1012 – VCM speed sampling block 1014 – possible validation of end of movement (EOT) block 1016 – check whether EOT has been reached; in response to a negative result (N) of block 1016, the system returns to block 1002 with a (new) value of VCM T speed block 1018 – in response to a positive result (Y) of the block 1016 the retraction procedure is finished. In summary: every time you read the speed, it is calculated a new DAC value based on the speed error measured; the DAC value is converted into a voltage of piloting (and so in a VCM current); every time you write a new DAC value (you (force a new current) updated values ​​are calculated for the durations TON and TOFF of the on times and off times off. The procedure outlined above is suitable for be implemented in a controller unit (MCU in Figure 1) of the HDD, provided that such controller is “active”, which may not be the case for a phase of emergency retraction. Advantageously, the procedure can be implemented using hardware circuitry integrated into the combo HDD power. A procedure as illustrated in the flowchart of Figure 6 facilitates the use of smaller values ​​for TOFF (short off times) and thus a BW bandwidth of higher speed ring during the critical point of retraction, where the VCM begins the climb on the ramp (Zone II in Figures 5A and 5B). A higher frequency can thus be used when the driving current is small, condition when the VCM engine is approaching the edge of the RE ramp. A higher frequency means bigger BW of the speed control ring and a latency of reduced speed sampling, i.e. a drop in reduced speed when climbing the ramp. A procedure as illustrated in the flowchart of Figure 6 facilitates a discontinuous retraction of the VCM with a variable frequency, with the frequency is modulated in function of the driving current. You can use a high frequency – with off times short – when the current is low, and vice versa, A lower frequency can be used – with times of longer off – when the current is higher. When a discontinuous retraction is working to high frequency, control bandwidth BW speed can be increased, which makes it easier to better speed control when heads they meet (hit) the edge of the ramp, starting to go up the ramp. Solutions as described here therefore provide an effective alternative to other solutions such as a discontinuous mode operation in which a voltage residual is compensated or an operation in mode continues using a BEMF monitor circuit that requires a calibration just before being used. Without prejudice to the basic principles, details and forms of implementation may vary, even in a appreciable, compared to what has been described purely by way of example, without leaving the scope of protection. The scope of protection is defined by the claims annexes.

Claims

1. A method, comprising: operating a moving coil motor, VCM (120) in a hard disk drive (HDD) during a retraction phase wherein read and write heads (H) of the hard disk drive are retracted by the VCM (120) from a surface of the disk (D) to an end-of-travel (EOT) park condition with the heads (H) ascending a ramp (R) starting from an edge of the ramp (RE) adjacent to the surface of the disk (D), wherein, during the retraction phase, the VCM (120) is operated in a discontinuous mode comprising alternating on times of duration TON and off times of duration TOFF wherein the supply of a drive current (IM) to the VCM (120) is, respectively, facilitated and resisted,the process involving varying the TOFF duration of the off times with the TOFF duration of the off times increasing from a lower limit value at the edge of the ramp (RE) to a higher limit value as the heads (H) rise above the ramp (R)., 2. A method according to claim 1, comprising varying the TOFF duration of the off times according to the relationship: TOFF = MinTOFF*(1+K*[DAC]) where MinTOFF is the lower limit value at the edge of the ramp (RE) and K*[DAC] is an adjustable increment factor of the TOFF duration of the off times.

3. A method according to claim 1 or claim 2, wherein the lower limit value and the upper limit value of the TOFF duration of the off times are about 200 us and about 600 us, respectively.

4. A method according to any preceding claim, wherein the method comprises varying the duration TON of the on times with the duration TON of the on times increasing from a lower limit value at the edge of the ramp (RE) towards an upper limit value as the heads (H) rise above the ramp (R).

5. A method according to claim 4, comprising varying the duration TON of the on times as a function of the duration TOFF of the off times.

6. A method according to claim 5, comprising varying the duration TON of the on times according to the relationship: TON = K*TOFF where: TOFF is the duration of the off times; and K is a proportionality factor equal to 1.5 or 2.

7. A method according to any of claims 4 to 6, wherein said respective lower limit value and said respective upper limit value of the duration TON of the on times are, respectively, about 400 us and about 1200 us.

8. A method according to any preceding claim, comprising controlling the intensity of the driving current (Im) through the VCM (120) as a function of a Counter Electromotive Force, BEMF of the VCM (12 0).

9. A control circuit (MCU, 180) configured to operate a voice coil motor, VCM (120) in a hard disk drive (HDD) during a retraction phase wherein read and write heads (H) of the hard disk drive are retracted by the VCM (120) from a surface of the disk (D) to an end-of-travel (EOT) park condition with the heads (H) rising over a ramp (R) starting from an edge of the ramp (RE) adjacent to the surface of the disk (D), wherein the control circuit (MCU, 180) comprises retraction logic circuitry (180) configured to be coupled (VCMSNSN, VCMSNSP, VCMN) to the VCM (120) and to operate the VCM (120) during the retraction phase by the method of any preceding claim.

10. A hard disk drive (HDD), comprising: at least one voice coil motor (VCM), and a control circuit (MCU, 180) according to claim 9, having said retraction logic circuitry (180) coupled (VCMSNSN, VCMSNSP, VCMN) to the VCM (120) and configured to operate the VCM (120) during the retraction step by the method according to any of the preceding claims 1 to 8.

11. A computing device comprising a hard disk drive (HDD) for storing signals therein, the hard disk drive (HDD) being according to claim 10.