WO1991010991A1

WO1991010991A1 - Process and circuit arrangement for adjusting the access arm of a disk operating system

Info

Publication number: WO1991010991A1
Application number: PCT/EP1990/002187
Authority: WO
Inventors: Volker Meyer
Original assignee: Siemens Nixdorf Informationssysteme Aktiengesellschaft
Priority date: 1990-01-11
Filing date: 1990-12-14
Publication date: 1991-07-25
Also published as: EP0510031A1; KR920704271A; JPH05501935A; DE4000663C1

Abstract

The invention relates to a process for adjusting the access arm of a disk operating system, the access arm of which carries at least one transducer which generates a position signal (43) from continuously read positional data. In order to position the transducer on the corresponding track, a servo drive (36) is actuated. The position signal (43) is scanned with a constant scanning frequency (fa) and converted to digital values which are fed to a controller that operates digitally and at a constant operating frequency (fr). Voltage pulses of constant pulse frequency (fp), whose polarity and/or duration (tp) are adjusted in function of the output signal of the controller (48), are fed to a drive coil. The respective ratios of the pulse frequency (fp), the scanning frequency (8a) and the operating frequency (fr) are whole numbers. The process can be used to reduce electric losses in the servo drive (36) and to prevent fluctuations in the control system.

Description

Method and circuit arrangement for adjusting an access arm of a disk storage system

The invention relates to a method for adjusting an access arm of a disk storage system with at least one disk on which. data is stored along a multiplicity of tracks, the access arm carrying at least one transducer head which generates a position signal from continuously read position data and which is positioned on the respective track by the access arm being actuated by a servo drive which ent¬ a drive coil holds which electrical current is supplied as a function of the output signal of a controller, which determines the deviation of the position signal from a target value indicating the target position and regulates it. The invention further relates to a circuit arrangement for performing the method.

Such a method is e.g. used in the positioning control of a magnetic disk memory. The access arms of several plates of a plate stack are mechanically connected to one another to form an access comb. Each access arm carries at least one transducer head, which can be designed as a read / write head or only as a write head or read head. The transducer heads of an access comb are arranged in such a way that they point to traces of the same radial position on the various plate surfaces at every position of the access comb.

A storage surface of a disk is exclusively described with a pattern of position information, which define locations of concentric tracks on the disk surfaces. A specially designed read head, which is also referred to as a servo head, continuously scans this position information, and an evaluation unit uses this information to generate a position signal which can be used to determine whether the servo head is pointing to a predetermined track and whether there is a deviation from the position Target track exists.

A servo drive is used to adjust the position of the servo head, which moves the access arm or the access comb in the radial direction transverse to the traces of the plate surfaces. As an actuator, the servo drive contains a drive coil on which a static magnetic field acts. If a current flows through the drive coil, it is deflected in the magnetic field in a direction dependent on the current direction as a result of the force acting on a conductor through which current flows, and thereby adjusts the relevant access arm or the entire access chamber ^" In the known method, a continuous current is applied to the drive coil, the amount and sign of which depend on the output signal of the controller, which detects the deviation of the position signal from a setpoint and regulates it to zero. The power amplifier of the controlled system required for the continuous supply of current to the drive coil is inevitably also operated in a state in which it generates a high electrical power loss. This results from the fact that with a medium control of the amplifier - ie the drive coil is fed with a medium-strength current - the power loss generated in the power amplifier is at a maximum, since both the voltage drop across the power amplifier and the current are large, and consequently that Product is maximal from these sizes. The power loss is converted into heat and must be dissipated to the environment by complex cooling measures. The result of this heat is that the electronic components of the servo drive are exposed to a high thermal load. Furthermore, the device dimensions must be large in order to create space for adequate cooling.

Another problem with the known method is the application of current control in the power amplifier. Corresponding circuits, if they have a high control speed, show a high tendency to instability, the behavior being sensitive to the change in electrical properties of components which depend on the operating temperature or on construction tolerances. Such instabilities can additionally be fueled by the position control. With critical interferences, eg with a temperature increase, tends; in such controllers therefore too Control vibrations that can only be controlled by reducing the control speed.

It is an object of the invention to provide a simple method for adjusting an access arm of a disk storage system, in which the electrical losses in the servo drive are reduced compared to the prior art and control vibrations are suppressed.

This object is achieved for a method of the type mentioned above in that the position signal is sampled at a constant sampling frequency and converted into digital values which are fed to the controller working digitally and at a constant operating frequency by supplying voltage pulses of constant pulse frequency to the drive coil. the polarity and / or duration of which is set as a function of the output signal of the controller, and that the pulse frequency, the sampling frequency and the working frequency are in an integer relationship to one another.

The invention is based on the knowledge that the discontinuous or digital processing of variables in a control loop compared to continuous or analog processing leads to greater accuracy and better reproducibility of the control result. Accordingly, in the case of the invention, the controller is not realized by resistors, capacitors and amplifiers, but rather by digitally working assemblies which can be controlled by a microprocessor. The result of a controller operating in this way is independent of critical component parameters such as temperature and component tolerance, so that the. Sturgeon p iπdlichkeit against external interference verrin¬ siege and Schwiπgneigung of Regelk ^r ice is reduced. The digital control concept is consequently also applied to the actuator in that the drive coil is supplied with voltage pulses. These voltage pulses cause a current flow in the drive coil, so that a force acts on the drive coil according to the electrodynamic principle, which moves the access arm or the entire access comb in a radial direction, which depends on the sign of the current and thus on the sign of the voltage pulses .

A method of modulation of the voltage pulses consists of the fact to alternately switch ^"positive and negative voltage to the drive coil. The duty cycle of the voltage pulses is adjusted according to the output of the digital controller, which in turn according to the detected deviation of the digital values of the position signal by a digital setpoint The duty cycle of the voltage pulses can be varied between the values 0 and 100%, the value 100% corresponding to a DC voltage without pulse change and the value 0% having the reverse DC voltage. This type of modulation the voltage pulse is referred to as pulse width modulation.

Another possibility of modulating the voltage pulses is to switch only pulses of one polarity to the drive coil, the polarity depending on the polarity of the output signal of the digital controller, the duration, however, depending on the amount. The control range here is ± 100 %, whereby 100% corresponds to a DC voltage again, with 0% the voltage _{impulse is} missing _^ This type of modulation can be called modified pulse width modulation.

Due to the use of voltage pulses, the power stage supplying the drive coil with electrical power is only operated in two operating states, in the switching mode and in the blocking mode. In both states, the power loss arising in the final stage is lower than in continuous operation using the known method. This is due to the fact that the switching stage is supplied with a large current in the switching mode, but with a small voltage drop. In contrast, in the blocking operation of the output stage its voltage drop is large, but the current flow is small. The respective product from the two electrical quantities, i.e. the electrical power loss is minimal in both cases. In this view, the dynamic power losses generated when switching between the operating states of blocking and switching are neglected. This is permissible since these power losses do not count in relation to the current-related static power loss of the output stage.

As a result of this type of energy supply to the drive coil, the electrical power taken from a voltage supply is almost completely given off to the drive coil, which converts it into kinetic energy. It is thereby achieved that the servo drive operates with a high degree of efficiency and the electrical power loss converted into heat is reduced to a minimum value. As a result, complex cooling measures for the circuit electronics can be dispensed with and their housing dimensions can be made small. The supply of the drive coil with voltage pulses has the result that an interference variable with an interference frequency acts on the control system, which is equal to the pulse frequency of the pulses. This can lead to the control loop becoming unstable in the worst case if no countermeasures are taken. Therefore, according to the invention, the pulse frequency, the sampling frequency and the working frequency are chosen so that they are in an integer relationship to each other. If this is the case, there are also fixed phase relationships between the signals of the various modules, for example when scanning the position signal, when regulating and when driving the drive coil. When the position signal is sampled at the sampling frequency, those spectral components in the position signal whose frequency is an integer multiple of the sampling frequency are hidden. The interference amplitudes with the pulse frequency or their harmonics therefore do not reach the digital controller, so that no positive feedback effect can arise which excites the control circuit to oscillate. This measure ensures that the control loop operates very stably and effectively suppresses control vibrations.

In a preferred embodiment of the invention, the pulse frequency of the voltage pulses is an integral multiple of the sampling frequency with which the position signal is sampled. This ensures that the drive coil is supplied with high-frequency voltage pulses even at a low sampling frequency, so that a current smoothed by the inductance flows in it, causing a smooth, jerk-free movement of the access arm. To carry out the method according to the invention, it is advantageous to use a formwork arrangement which is characterized in that four controllable switching elements arranged in a bridge circuit are provided, the one bridge diagonal being connected to the supply voltage and the other bridge diagonal being the Contains drive coil, and that the switching elements can be controlled so that, depending on the direction of the deviation of the transducer head from the target track, two diagonally opposite switching elements are not conductive and the other two switching elements can be switched on during the pulse duration.

This circuit arrangement makes it possible to supply the drive coil with voltage pulses with an optionally adjustable sign, which are derived from only one operating voltage. These voltage pulses of different signs cause currents of opposite polarity to flow in the coil. These generate forces which cause the access arm to accelerate or brake in opposite directions.

In a further development of this embodiment, field effect transistors are provided as switching elements. These semiconductor components are particularly well suited for the rapid switching of voltages without high power losses occurring in them. As a result, the electrical losses in the servo drive are further reduced.

An embodiment of the invention is explained below with reference to the drawing 1 shows a disk storage system with an access comb actuated by a servo drive,

Fig. 2 is a schematic representation of the

Control circuit for digital control of the position of a servo head in a block diagram,

3 shows a circuit arrangement for controlling a drive coil with voltage pulses, and

Fig. 4 curves of the voltage and current of the drive coil and the position signal over time.

FIG. 1 schematically shows a disk storage system with a plurality of disks 10, 12, 14 which rotate about a spindle axis 16. A plurality of tracks in which information can be magnetically recorded are provided on an annular section of the respective disk 10, 12, 14. This information is written or read by read / write heads 30, 32 arranged on access arms. The access arms are mechanically fixed to an access comb 18. This ensures that all read / write heads 30, 32 can be positioned on the same track of the respective disk 10, 12, 14 in a single positioning operation.

The access comb 18 is rotatably mounted on bearings 20, 22 and can carry out pivoting movements about the pivot axis 24. With the help of these swiveling movements, the read / write heads 3.0,: -32 are moved radially to the plates - surfaces 10, 12, 14 and across the tracks. The swiveling movement is brought about by a servo drive 36, which contains, as the drive coil, a plunger coil 40 through which an electric current I flows with an alternating sign. The plunger coil 40 is located in a magnetic field of a permanent magnet 38. Depending on the strength of the current I and on its direction, the coil 40 and thus the access comb 18 are pivoted by the force effect on a conductor through which current flows.

Position information in the form of magnetic flux changes is stored on the underside of the magnetic plate 14 and is read by a special read head or servo head 32. With the help of this position information, individual tracks on the disk 14 can be identified. The servo drive 36 is now adjusted so that the servo head 32 is positioned on a desired target track so that data can be written and read on the tracks of the disks 10 and 12 with the same radial positions.

The control required for positioning the servo head 32 is shown schematically in a block diagram in FIG. 2. The position information on the magnetic disk 14 which is continuously read by the servo head 32 is fed to a signal generating module 42, which generates a position signal from this information, which indicates the current position or actual position of the servo head 32.

The track with a specific track number, on which the servo head 32 is to be positioned, is counted down by that of the servo head in the radial direction

Saving averaged starting with a starting lane. Another way of reaching one Determining the desired track consists in storing an absolute address characterizing it in the respective track as position information and evaluating this address for the track positioning.

After positioning on the desired track, which is also referred to as rough positioning, the read / write heads of the storage disks must be positioned in the center of the respective data track. For this purpose, magnetic position marks are stored on the servo surface in accordance with a known pattern using the known 4-servo signal method. The position of the position marks relative to one another define center lines of data tracks. When these position marks are read by the servo head 32, voltage pulses are generated which are evaluated in the signal generation module 42. A deviation from this center line manifests itself in the position signal 43 generated by the signal generation module 42.

The position signal 43 is sampled by an analog / digital converter 44 with a constant sampling frequency fa and converted into digital values. These are fed to a digital controller 48, which generates a digital output signal that controls a pulse width modulator 52. The digital controller 48 has PID behavior and is designed for optimum control speed and minimal control deviation. It works with a fixed operating frequency fr, ie the system deviation is only determined at certain times and its output signal can also only be changed at such times. A signal 50 is supplied to the digital controller 48, in which the number of the target track is digitally encoded as the setpoint. Another signal 51 contains, as information, the setpoint for fine positioning on the center line of a data track. The signals 50 and 51 are provided by a higher-level controller (not shown).

The pulse width modulator 52 generates pulse-shaped control signals S1, S2 with a constant pulse frequency fp, the respective pulse duty factor of which can be varied between 0% and 100%. The duty cycle here is the ratio of the pulse duration tp to the time T between two pulses. The control signals S1, S2 are fed to a power output stage 54, which generates voltage pulses U of a defined magnitude with different signs. As will be explained in more detail, the voltage pulses dependent on the control signal S 1 serve to accelerate a moving coil in a predetermined direction, while the voltage pulses dependent on the control signal S2 accelerate this moving coil in the opposite direction.

The voltage pulses U are fed to a servo drive 36 which contains a moving coil which is deflected by the current in the coil caused by the voltage pulses in the radial direction x and thereby adjusts the servo head 32. ^{^} The direction of adjustment depends on the current direction and thus on the sign of the voltage pulses U with which the coil is applied.

A clock generator 46 generates synchronous clock pulses with the frequency fe and twice the frequency fp. These clock pulses are made by frequency division Mother impulses derived with a higher frequency. The pulses with the frequency fa are fed to the digital controller 48 and the analog / digital converter 44 and determine the operating cycle of these modules. The pulse width modulator 52 is controlled with clock pulses of the higher frequency fp.

The mode of operation of the control loop according to FIG. 2 is explained below. The target position to be approached by the servo head 32 is communicated to the digital controller 48 via the signals 50, 51. If the actual position of the servo head 32 shown in the position signal 43 deviates from this target value, the digital controller 48 generates an output signal which controls the pulse width modulator 52 and consequently the power output stage 54 in such a way that the voltage pulses U output to the servo drive 36 drive the servo head 32 in Move towards the target track. When the target track is reached, which is preceded by a braking process with voltage pulses of opposite polarity, the digital controller 48 switches over to the fine positioning mode and regulates the position of the servo head 32 exactly on the center line of the data track.

In terms of control technology, the voltage pulses U output by the power output stage 54 represent a disturbance variable with an interference frequency equal to the pulse frequency fp with which the control loop is acted upon. If this disturbance variable is fed back in the control loop so that it fulfills the oscillation conditions for control loops in terms of phase and amplitude, this can become unstable.

In the present exemplary embodiment, the sampling frequency fa and the control frequency fr ha> are as large as the pulse frequency "fp, ie the frequencies are in each case in an integer relationship to one another discontinuous sampling of the position signal 43 and in the discontinuous processing of the signals in the control loop, each with a constant frequency, it is known to suppress those spectral components in the signals which oscillate with the fundamental frequency or with an integral multiple of this frequency. This means that the amplitude values of such vibrations are not amplified in the control loop and thus cannot trigger a positive feedback effect which could set the control loop in vibration. The disturbance variable caused by the voltage pulses U therefore has no negative influence on the stability of the entire control loop.

3 shows an exemplary embodiment of the power output stage 54, which outputs the voltage pulses U to a plunger coil 62 with different signs. Four field effect transistors 64, 66, 68, 60 are connected in a bridge circuit and are supplied with the positive supply voltage V from a voltage source. The voice coil 62 is arranged in the bridge diagonal. If the diagonally opposed field effect transistors 66 and 60 are switched into the blocking state and the transistors 64 and 68 are driven at their gate electrodes with the pulse-shaped control voltage S1, the plunger coil 62 is made up of supply voltage via the current path V, field effect transistor 64, plunger coil 62, field effect transistor 68, measuring resistor 63 and ground, negative voltage pulses supplied. These pulses let a current I flow in the negative coil 62 in the negative direction, so that the coil 62 is accelerated in a direction predetermined by the current direction. If this direction is now to be reversed, for example, in order to brake the moving coil, .- - C or in the opposite direction move, the transistors 64 and 68 are to be switched into the blocking state and the field effect transistors 60 and 66 are to be driven at their gate electrodes with the pulse-shaped control voltage S2. This type of circuit arrangement makes it possible to generate voltage pulses U with different signs from a voltage source with only one voltage V.

A voltage U3 proportional to the current I through the plunger 62 drops across the measuring resistor 63. This voltage U3 can be used to improve the timing behavior of the control loop, in that it is also scanned by the digital controller and a current control loop is subordinate to the position control in a manner known per se. This makes it possible to detect interference on the control circuit even earlier and to increase the positioning speed for the servo head 32.

4 shows curves of the signals S1, S2, and of the current I, expressed by the voltage U3, and of the position signal 43 over time t for a positioning process on desired tracks n. The sizes S1, S2, U3 correspond to the sizes given in FIG. 3. According to the image part d, there is a control deviation 76 between a target position 65 and the actual position, represented by the current position signal 43, at time t = 0. The control loop is now endeavoring to correct this control deviation 76 to the value 0. For this purpose, the power output stage 54 is driven in such a way that it outputs positive voltage pulses U with the control signal S 1 to the duty cycle 100% to the moving coil 62. This is deflected in the radial direction, so that the control deviation 76 'after one Time tl is significantly reduced. The control loop then reacts by reducing the duty cycle tp / T.

In order to allow the servo head 32 to settle on the target track 65, control pulses S1 are emitted to the power end stage 54 and thus voltage pulses U with a negative sign are transmitted to the plunger coil 62 starting at the time t2. As a result of the negative acceleration, the servo head 32 is braked and reaches the target track 65 at time t3. The rough positioning process is thus completed and the digital controller switches over to the fine positioning mode. In this operating mode, only small positional deviations have to be corrected, which are shown in the figure part d) of FIG. 4 greatly exaggerated. Accordingly, the control pulses 70, 72, 74 required for this have correspondingly small duty cycles tp / T.

Claims

Patent claims

1. A method for adjusting an access arm of a disk storage system with at least one storage disk on which data is stored along a plurality of tracks, the access arm carrying at least one converter which generates a position signal from continuously read position data and which generates a position signal the respective track is positioned by being actuated by a servo drive which contains a drive coil, which electric current is supplied as a function of the output signal of a controller which detects and corrects the deviation of the position signal from a target value indicating the target position, characterized in that the position signal (43) is sampled at a constant sampling frequency (fa) and converted into digital values which are fed to the controller (48) operating digitally and with a constant operating frequency (fr) that the drive coil (40, 62) voltage pulses (U) of constant pulse frequency (fp) are supplied, the P polarity and / or duration (tp) is set depending on the output signal of the controller (48), and that the pulse frequency (fp), the sampling frequency (fa) and the operating frequency (fr) are in each case in an integer ratio to one another.

2. Ver ahren ^ according to claim 1, characterized in that the pulse frequency (fp) is an integral multiple of the sampling frequency (fa) "

3. The method according to claim 1 or 2, characterized gekenn¬ characterized in that the operating frequency (fr) of the Digital¬ controller (48) is an integer multiple of the Ab¬ sampling frequency (fa).

4. The method according to claim 1, 2 or 3, characterized ge indicates that the working frequency (fr) is equal to the Abtas frequency (fa).

5. The method according to any one of the preceding claims, characterized in that the sampling frequency (fa), the working frequency (fr) and the pulse frequency (fp) are derived from common mother pulses of higher frequency.

6. Circuit arrangement for performing the method according to one of the preceding claims, characterized in that four controllable switching elements (60, 64, 66, 68) arranged in a bridge circuit are provided, the one bridge diagonal being connected to the supply voltage (V) and the other diagonal of the bridge contains the drive coil (62), and that the switching elements (60, 64, 66, 68) can be controlled such that, depending on the direction of the deviation of the transducer head (32) from the target track (65), two diagonally opposite one another ¬ horizontal switching elements (60, 66 or 64, 68) are not conductive and the other two switching elements (64, 68 or 60, 66) can be switched conductive during the pulse duration (tp).

7. Circuit arrangement according to claim 6, characterized in that field effect transistors are provided as switching elements (60, 64, 66, 68).

8. Circuit arrangement for performing the method according to one of claims 1 to 5, with a clock pulse generating clock generator, a digital controller, which receives digital control signals from an analog / digital converter and controls a pulse width modulator by its output signal , characterized in that the analog / digital converter (44) clock pulses with a constant sampling frequency (fa), the digital controller (48) clock pulses with a constant operating frequency (fr) and the pulse width modulator (52) clock pulses with a constant pulse frequency ( fp) are supplied, and that the pulse frequency (fp), the sampling frequency (fa) and the working frequency (fr) are in each case in an integer relationship.

9. Circuit arrangement according to claim 8, characterized ge indicates that the pulse frequency (fp) is double the te of the sampling frequency (fa).

10. Circuit arrangement according to claim 8 or 9, characterized in that the clock pulses of the Taktgenera¬ gate (48) are derived by frequency division from mother pulses with a higher frequency.