JP2011258306A - System and method for module type power management - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a system and a method for managing power consumption of a hard disk drive system.SOLUTION: A module type applicable power management 110 includes a power island register and an oscillation control register. Both registers are capable of writing through firmware and writing through a host interface. A hard disk controller 140, an interface controller 120, a channel reading module 130, and a power management system are implemented over two or more different power islands and use two or more different clocks. Power to the two or more different power islands is at least partially controlled by the power management system through the power island register, and the two or more different clocks are respectively controlled by the power management system through the oscillation control register.

Description

  The present invention relates to a system and method for managing power loss of an electrical system, and more particularly to a system and method for managing power consumption of a hard disk drive system.

  Hard disk drives are complex electromechanical systems and are more complex systems due to both power and heat dissipation problems. These power issues and heat dissipation issues have been addressed in a variety of individual directions (eg, low power application specific semiconductor design, hardware, firmware, or software to reduce clock speeds). Any use of). An example is a hard disk drive that manages power loss in accordance with software instructions from a host. Such host instructions are typically distributed over the ATA bus and are inherently very coarse. This lack of precision prevents the drive power from being effectively managed while keeping the hard disk drive operation at a sufficient level.

  Accordingly, there is a need in the art for advanced power management systems and methods for at least the reasons described above.

  The present invention relates to systems and methods for managing power loss in electrical systems, and more particularly to systems and methods for managing power consumption of disk drives.

  Various embodiments of the present invention provide a modular adaptive power management system. Such modular adaptive power management systems include a hard disk drive controller, a channel read module, a host interface controller, and a power management system. The hard disk drive controller includes a processor that executes firmware and the host interface controller allows host access via the host interface. The host interface or translator interface may be, for example, an ATA interface, a SATA interface, and / or other emerging serial interfaces such as MMC, CE-ATA, SDIO, etc. The power management system includes a power island register and an oscillation control register. The power island register and the oscillation control register are both writable indirectly, at least via firmware and / or host interface, respectively. The hard disk controller, interface controller, channel read module, and power management system are implemented across two or more separate power islands and use two or more separate clocks. Power to the two or more separate power islands is controlled, at least in part, by the power management system via a power island register, and the two or more separate clocks are each via an oscillation control register. Controlled by power management system.

  Other embodiments of the present invention provide reusable modular power management circuits for use in various systems. Such a power management circuit includes a power island control register and an oscillation control register. A power island control register is associated with a number of power island control bits, each of which is operable to disable power delivery to a given power island. The oscillation control register is associated with a number of oscillation control bits, each of which is operable to disable a given clock. The circuit further includes a translator interface. The translator interface is operable to receive standard power management instructions and convert the standard power management instructions into writes to the power island control registers and oscillation control registers. is there. Writing to the power island control register and the oscillation control register implements a power management scheme adapted in connection with standard power management instructions. In some cases, the standard power management instruction is an ATA or SATA instruction, or an instruction associated with one of the emerging serial interfaces, such as MMC, CE-ATA, SDIO, etc.

  In some examples of the foregoing embodiments of the invention, the circuit further includes an embedded processor that executes firmware. Such an embedded processor can at least partially control the operation of the translator interface. In some cases, the firmware can be upgraded. In one implementation of the foregoing embodiment of the present invention, one of the oscillation control bits is operable to disable the oscillator and another of the oscillation control bits is the oscillator. Is operable to gate all clocks that are taken from, and yet another bit of the oscillation control bits gates some but not all of the clocks that are taken from that oscillator It is possible to operate.

  Yet another embodiment of the present invention provides a hard disk drive system that includes a hard disk controller, an interface controller, a channel read module, and a power management system. These hard disk controllers, interface controllers, channel read modules, and power management systems are implemented across two or more separate power islands, and the power supply to those two or more separate power islands is at least In part, controlled by a power management system. In some cases, these hard disk controllers, interface controllers, channel read modules, and power management systems utilize two or more separate clocks, each of which is a power management system. Controlled by. In some cases, the interface controller provides a host interface such as, for example, an ATA interface, a SATA interface, and / or other emerging serial interfaces such as MMC, CE-ATA, SDIO, etc. Is operable. In some cases, the hard disk drive system further includes a processor that executes the firmware. In such cases, the firmware is operable to implement an adaptive power management scheme via the power management system. In certain cases, the interface controller is operable to provide a translator interface operable to receive a host instruction that causes the selected power mode to be implemented via the power management system. .

  Yet another embodiment of the present invention provides a modular adaptive power management system that includes a power island controller and an oscillation controller. These oscillation controllers and power island controllers are implemented in a modular power management circuit. In some cases, the power management system further includes a processor having firmware executable to operate the power management system. Further, in some cases, the power management system is deployed in the context of a device that includes two or more power islands. In such a case, the power island controller is one register bit that is writable by firmware and is operable to disable power to one of those power islands. And another register bit that is writable by firmware and operable to disable powering to another of those power islands. In some cases, the firmware is operable to implement an adaptive power management scheme.

  In yet another case of the foregoing embodiment, the power management system further includes a host interface. In such a case, the aforementioned register bits can be written via the host interface. In certain cases, a host communicating via the host interface commands the implementation of a coarse power management scheme. In such cases, the firmware may be operable to implement a finer adaptive power management scheme. In some cases, the power management system is register based and can be deployed in the context of various host systems.

  This summary provides only an overview of some embodiments according to the present invention. Many other objects, features, advantages and other embodiments of the present invention will become more fully apparent from the following detailed description, the appended claims and the accompanying drawings.

  A further understanding of the various embodiments of the present invention may be obtained by reference to the drawings described in the remaining portions of the specification. In the drawings, like reference numerals have been used throughout the drawings for like elements. In some examples, a sublabel consisting of lowercase letters is associated with a reference sign to represent one of a plurality of similar elements. If a reference to a reference is made without specifying an existing sublabel, this is an attempt to reference all such similar elements.

1 illustrates a hard disk drive system incorporating a modular adaptive power management module according to various embodiments of the invention. FIG. 6 is a graph illustrating exemplary adaptive transitions between different power levels in an active state according to one or more embodiments of the present invention. FIG. 4 illustrates different systems including power islands under the control of a power management system according to various embodiments of the invention. FIG. 4 illustrates different systems including power islands under the control of a power management system according to various embodiments of the invention. 1 illustrates a register-based modular power management system according to one or more embodiments of the present invention. FIG. FIG. 2 illustrates a reusable modular power management circuit according to one or more embodiments of the present invention.

  The present invention relates to a system and method for managing power loss of an electrical system, and more particularly to a system and method for managing power consumption of a hard disk drive system.

  Many computer devices use hard disk drives. For example, hard disk drives are commonly incorporated into personal computers, including both desktop and laptop computers. In addition, many consumer electronic devices are beginning to incorporate hard disk drives as a mechanism for increasing the storage capacity of the devices. For example, digital video cameras, digital still cameras, personal digital assistants, mobile phones, and the like are becoming commonplace to incorporate hard disk drives. In many cases, these devices operate at very high clock speeds and require relatively low access times for data residing on the associated hard disk drive. By increasing the speed at which the data retention platter rotates relative to the read / write head assembly, it is possible to implement a high speed hard disk drive. Since these platters are connected to a rotating spindle, the rotation speed of the platter is often referred to as the spindle speed. Increasing the spindle speed not only increases the speed at which platter data can be accessed during sequential access, but also reduces the rotational delay, thereby reducing the average time required to perform random data access. Due to the benefits gained by increasing spindle speed, efforts to further increase spindle speed continue.

  Increasing the spindle speed increases vibration energy and thermal energy, each of which is detrimental to hard disk drive operation. By introducing effective power management, it is possible to suppress the increase in thermal energy and improve the performance of the hard disk drive. Power management is most important in hard disk drives that should be deployed in mobile consumer electronics, but power management is much larger and less mobile (eg, laptop personal computers, or even Even in the case of hard disk drives that should be deployed for corporate use, etc., it is quite important.

  Various embodiments of the present invention work in conjunction with existing techniques for adaptively controlling the active power of hard disk drives to provide elaborate adaptations that can limit heat loss while providing a desired level of performance. Power management is realized. Such an embodiment of the present invention provides an adaptive transition from a low power loss mode to a high power loss mode and / or from a high power loss mode to a low power loss mode, eg via a standard ATA bus. It can be included at a fineness not provided by the control by the distributed host command. In some embodiments of the invention, the adaptive power management system described above is implemented as a dedicated power management controller. Such dedicated power management controllers may be modular so that they can be adapted to and deployed in connection with various hard disk drive designs and / or other system designs. Is possible.

  Various embodiments of the present invention provide a modular adaptive power management system. Such modular adaptive power management systems include a hard disk drive controller, a channel read module, a host interface controller, and a power management system. The hard disk drive controller includes a processor that executes firmware and the host interface controller allows host access via the host interface. The host interface or translator interface may be, for example, an ATA interface, a SATA interface, and / or other emerging serial interfaces such as MMC, CE-ATA, SDIO, etc. The power management system includes a power island register and an oscillation control register. In this specification, the term “register” is used in its broadest sense, meaning any memory element. Thus, in some cases, the register may be a single flip-flop or latch, a group of flip-flops or latches, or some other addressable memory element. Note that a “register” may include a subset or superset of individual register bits that can be written and / or read at a given address or memory address. The power island register and the oscillation control register are both writable indirectly, at least via firmware and / or host interface, respectively. The hard disk controller, interface controller, channel read module, and power management system are implemented across two or more separate power islands and use two or more separate clocks. As used herein, the phrase “power island” is used in its broadest sense to mean any region that is provided by a power source and that is controllable away from other regions of a given system. Power to the two or more separate power islands is controlled, at least in part, by the power management system via a power island register, and the two or more separate clocks are each via an oscillation control register. Controlled by power management system.

  FIG. 1 illustrates a hard disk drive system 100 that includes a modular adaptive power management module 110 according to one or more embodiments of the present invention. In some embodiments of the present invention, the modular adaptive power management module 110 may be deployed on many different devices and / or customized for operation on one or more devices. Implemented with possible modular design. In another embodiment of the present invention, modular adaptive power management module 110 provides one or more of host-initiated power management, firmware-controlled adaptive power management, and / or hardwired adaptive power management. It is possible. In yet another embodiment of the present invention, modular adaptive power management module 110 provides both design and adaptive power management modularity. In yet another embodiment of the present invention, modular adaptive power management module 110 provides modularity of design, adaptive power management, and power management with host instructions. In addition to the modular adaptive power management module 110, the hard disk drive system 100 includes an interface controller 120, a channel read module 130, a hard disk controller 140, a motor controller 150, a preamplifier 160, a spindle motor 170, and a read. / Write head assembly 180 and disk platter 190.

  The interface controller 120 may be any interface adapted to allow instructions to be passed to the host controller and / or instructions to be passed from the host controller. Accordingly, the host interface controller may be, but is not limited to, an ATA interface controller, a SATA interface controller, an MMC interface controller, a CEATA interface controller, or an SDIO interface controller. Those skilled in the art will recognize a variety of interface controllers that can be implemented in connection with one or more embodiments of the present invention based on the disclosure provided herein.

  The data on the disk platter 190 consists of a collection of magnetic signals that can be detected by the assembly 180 when the read / write head assembly 180 is properly positioned on the disk platter 190. In a typical read operation, read / write head assembly 180 is accurately positioned by motor controller 150 over a desired data track on disk platter 190. The motor controller 150 moves the read / write head assembly 180 to the proper data track on the disk platter 190 in accordance with instructions from the hard disk controller 140 to read / write heads to the disk platter 190. Position assembly 180 and drive spindle motor 170. The spindle motor 170 rotates the disk platter 190 at a predetermined rotational speed (RPM).

  When the read / write head assembly 190 is positioned near the proper data track, as the disk platter 190 is rotated by the spindle motor 170, a magnetic signal representing the data on the disk platter 190 is read. Detected by write head assembly 180. The detected magnetic signal is given as a minute continuous analog signal representing magnetic data on the disk platter 190. This small analog signal is transferred from read / write head assembly 180 to channel read module 130 via preamplifier 160. Preamplifier 160 is operable to amplify the minute analog signal read from disk platter 190. Further, the preamplifier 160 is operable to amplify data from the channel read module 130 that is to be written to the disk platter 190. The channel read module 130 then digitizes the received analog signal to recreate the digital data originally written to the disk platter 190. The digitized data is passed through the interface controller 120 to a host computer or a device that controls the operation of the hard disk drive system 100.

  The write operation is almost the reverse of the read operation described above, and digital data is received from the host via the interface controller 120. Digital data is passed to the channel read module 130 and the hard disk controller 140. The hard disk controller 140 provides proper positioning of the read / write head assembly 180 by the motor controller 150 in the vicinity of a particular data track on the data platter 190 and the desired number of revolutions of the disk platter 190. Instruct. When the above instructions are made, the digital data is converted to an analog data stream by channel read module 130 and passed to read / write head assembly 180 via preamplifier 160. Read / write head assembly 180 then writes the data to disk platter 190.

  In some cases, the read operation and the write operation described above are not performed continuously. In such a case, there is an opportunity to perform power management. The modular adaptive power management module 110 performs such power management on the hard disk drive system 100. In some specific embodiments of the present invention, modular adaptive power management module 110 provides programmable power management to increase drive energy efficiency while meeting desired system performance.

  As already proposed, in one or more embodiments of the present invention, the modular adaptive power management module 110 distributes and enables clocks to other major sub-blocks in the hard disk drive system 100. It is a stand-alone circuit that operates on In some embodiments of the present invention, modular adaptive power management module 110 allows firmware running on hard disk controller 140 to control power distribution to various other circuits in hard disk drive system 100. It supports several registers that make it possible. The modular adaptive power management module 110 can control the overall power loss of the hard disk drive system 100 by using several power management techniques in cooperation. In certain embodiments of the invention, the modular adaptive power management module 110 includes the ability to enable and disable the operation of synchronizing the entire system clock of the hard disk drive system 100. In certain cases, such a function can be provided by gating the entire system clock; in other cases, the aforementioned function is responsible for generating the entire system clock. It can be provided by turning off the power supply to the oscillator circuit.

  Further, the modular adaptive power management module 110 can include the ability to enable and disable sub-clocks of the overall system clock. Thus, for example, the modular adaptive power management module 110 can include the ability to disable the clock provided to the channel read module 130 and at the same time enable the clock provided to the interface controller 120. It should be noted that the modular adaptive power management module 110 may be mounted on the same piece of silicon as other elements of the hard disk drive system 100, and one of the other elements in a different piece of silicon and / or different package or It may be mounted away from a plurality of elements.

  Using the clock gating and / or clock off approach described above can dramatically reduce the amount of dynamic power lost in a CMOS circuit. In general, this is sufficient. This is because the loss of dynamic power in a given circuit can be many times the loss of static power (eg, leakage current). However, it may be advantageous to further control the leakage power. Thus, some embodiments of the present invention also provide the ability to limit leakage current. Specifically, in some embodiments of the present invention, modular adaptive power management module 110 includes the ability to shut off power to individual portions of hard disk drive system 100. In such cases, the individual design portions can be placed in power islands or regions where the various circuits share a common power source that is separate from the circuits in other regions. Thus, for example, although channel read module 130 and interface controller 120 receive power from the overall power source, power to each of channel read module 130 and interface controller 120 can be individually enabled and disabled. As is the case, the channel readout module 130 and the interface controller 120 can be implemented. Thus, disabling both power to channel read module 130 and disabling both dynamic power and static power loss by channel read module 130 by disabling power to channel read module 130 and enabling power to interface controller 120. The functionality of the interface controller 120 can be maintained.

  This multiple voltage island topology becomes more important as designs continue to move to different technologies. For example, in some technologies, dynamic power loss may be a relative advantage over any static power loss. On the other hand, in other technologies, static power loss may be more significant than dynamic power loss. Specifically, in some technologies, if there is an overall power loss, the main cause may be a leakage problem. By shutting down power to unused logic and shutting it down completely, this leakage factor is eliminated and overall system power consumption in standby and sleep mode operation is reduced.

  In some specific embodiments of the present invention, the modular adaptive power management module 110 includes an interface to the interface controller 120, which in some embodiments of the present invention provides power from the host controller. Used by hard disk drive system 100 to receive management instructions. Thus, although only one of many examples, the host can command three levels of operation: awake, awake, and sleep. Regardless of which mode is selected, when a media access command such as the above-mentioned read command or write command is received, the hard disk drive system 100 starts to respond to the request for the received command. Must. The time taken to start the operation in response to the received command request may vary depending on which mode the hard disk drive system 100 is currently operating. Therefore, there is a trade-off between response time and power loss. This trade-off can vary depending on how much the device is currently used and how fast response time is required by a given application. Thus, for example, in portable devices where power is very valuable, response time may be sacrificed to minimize power loss. On the other hand, in the case of a fixed device with no power limitation, the response time may be more important, and the implemented power management method may not be very aggressive.

  Specifically, the modular adaptive power management module 110, when instructed to operate in awake mode, provides enough power to respond to read and write requests with minimal delay. Guarantee. In contrast, when instructed to operate in standby mode, the modular adaptive power management module 110 responds with a predetermined delay that is expected to perform full power distribution after receiving a read or write request. Minimize power loss while giving time. Furthermore, when instructed to operate in sleep mode, the modular adaptive power management module 110 avoids significant delays whenever it becomes necessary to respond to a received read or write request. The power distribution is extremely cut off so that it is not possible. In some cases, such a sleep mode may include the hard disk drive system 100 being almost completely turned off. Such a complete power off allows a significant power savings, and in some cases where substantial caching is provided by the host computer or host device, the negative effects of such a power off. Is limited. As an example, MP3 devices can utilize a very low power and high delay approach where a significant amount of audio data is kept in the system cache. In some embodiments of the invention, the modular adaptive power management module 110 is enabled to enable the modular adaptive power management module 110 to be implemented whether either an aggressive power management scheme or a passive power management scheme is implemented. Programmability is provided to the management module 110. Such programmability can be implemented via a number of control registers included in the modular adaptive power management module 110, which are accessed via a processor running firmware. Yes, it is implemented as part of the hard disk controller 140. This allows the modular adaptive power management module 110 to be reused and utilized across multiple designs and / or application spaces. The user can adapt the trade-off between response time and power loss according to individual application requirements.

  There are various host interfaces for communication with the hard disk drive system 100. Such a host interface typically includes a predetermined power management strategy, which may or may not operate optimally for a particular hard disk drive system 100. There are cases where it does not. For example, the ATA interface is well known and provides a stable interface standard that the industry can design in compliance. However, the stability of the standard makes it difficult to meet optimal power management. The power management instructions provided as part of the ATA standard are generic and apply to any type of device that communicates via the ATA bus interface. Such an abstraction allows only coarse power management and is hardly adapted to a particular hard disk drive implementation. Furthermore, in many cases it is not possible for a particular host to implement a very aggressive power management scheme.

  Some embodiments of the present invention provide a self-managed or adaptive power management scheme that operates in addition to host instructions. In some cases, power management by host instructions operates as coarse power management, and self-management operates as fine control within the coarse control described above. In embodiments where adaptive power management is provided in addition to power management by a host command, such adaptive power management can operate within the range commanded by the host. Thus, for example, if the host commands operation in awake mode, adaptive power management is implemented to give a power loss that varies adaptively from the upper limit of the awake mode to the upper limit of the standby mode. Is possible. Similarly, when the host commands operation in standby mode, adaptive power management is implemented to give a power loss that varies adaptively from the upper limit of standby mode to the upper limit of sleep mode. It is possible.

  FIG. 2 is a graph 200 illustrating exemplary adaptive transitions between different power levels in an active state. The graph 200 includes a power axis 210 and a time axis 220. As indicated by block 225, initially, the hard disk drive system 100 is actively processing a read or write instruction. During this time, the hard disk drive system 100 is fully operational and the power lost thereby is maximized in the hard disk drive system 100. When the instruction processing is complete, the hard disk drive system 100 transitions to a high performance idle state represented by block 235. In this state, the hard disk drive system is still nearly operational and there is little power reduction possible. Further, in this state, there is no appreciable delay in receiving and processing subsequent instructions. As an example, the tristate bus can be placed in a tristate condition while other logic remains fully enabled during this state.

  If the inactive state continues for a predetermined time, the hard disk drive system 100 transitions to a fast idle state represented by block 245. In the fast idle state, an additional factor of power loss is eliminated at the expense of longer delay in responding to subsequent incoming commands. Similarly, after a further predetermined time has elapsed, the hard disk drive system 100 transitions to a low power idle state represented by block 255. In the low power idle state, an additional factor of power loss is eliminated at the expense of longer delay in responding to subsequent incoming commands. Eventually, after some further delay, the hard disk drive system 100 transitions to the standby power state. Thus, the host bus is used to command coarse power states (eg, active, standby, and sleep) and adaptive power management is used to use power at a finer level between the coarse levels due to host commands. It is possible to correct how it is done. Alternatively, if host instructions are not available, the standby state can simply be considered as one of the other power states to which adaptive power management transitions.

  FIG. 3 shows a system 300 adapted for use with an MMC interface and another system 301 adapted for use with a SATA interface. These two systems are merely exemplary, and one of ordinary skill in the art can implement power management according to one or more embodiments of the present invention based on the disclosure provided herein. Note that a variety of possible systems will be recognized. Specifically, the system 300 includes a standard controllable storage system 310 that includes a motor controller 320, a storage platter 322, and a read / write head assembly 324. The motor controller 320 includes a switching regulator 318, a switched capacitor 312, a spindle driver 314, and a voice coil motor driver 316. Controllable storage system 310 is well known in the art, and those of ordinary skill in the art will understand that controllable storage system 310 is one or more embodiments of the present invention upon reading this disclosure. Note that it would be possible to replace other known systems that can apply power management according to. The motor controller 320 is electrically coupled to the stabilized power supply via the LC filter 326. One skilled in the art will recognize other filters that can be used in connection with the present application based on the disclosure provided herein.

  The system 300 further includes an MMC interface 330 that allows the MMC power supply 332, the MMC clock 336, and the MMC signal 338 to pass to the MMC host interface circuit 360 via the clock buffer 348 and the tristate buffer 344. Supplied. In addition, system 300 includes a mini-voltage regulator 342 operable to generate a 1.2 volt power supply from MMC power supply 332. System 300 also includes an oscillating circuit 350 associated with crystal oscillator 352. Further, the system 300 includes a processor and / or other control logic 390 and a power management circuit 380 in accordance with an embodiment of the present invention. The system 300 further includes a clock generation circuit 372 and a channel read digital circuit 396.

  The power management circuit 380 controls power to the various power islands 340, 370, 392, and 394 in the system 300 and provides separate clocks to the various parts of the system 300. The role of the power management circuit 380 is to gate one or more of those clocks to eliminate the dynamic power loss associated with each part, and the power islands 340, 370, 392, and 394 Turning off power to one or more of them to eliminate both dynamic and static power losses associated with each island. In some cases, power management circuit 380 and / or interface circuit 360 supports responding to ATA / ATAPI instructions that are communicated without using an active clock. In such an embodiment, the oscillator circuit 350 can be turned off and the system 300 can still be recovered by a host instruction.

  System 300 is configured as separate power islands 340, 370, 392, and 394, each including circuitry shown in adjacent dashed lines. For example, switch 374 controlled by power management circuit 380 can be opened to turn off power to power island 370. Thereby, all power supply to the clock generation circuit 372 is turned off, and not only dynamic power loss but also static power loss is eliminated. Similarly, a switch 398 controlled by the power management circuit 380 can be opened to turn off power to the power island 394. Thereby, all power supply to the channel readout digital circuit 396 is turned off, eliminating not only dynamic power loss but also static power loss. Although not shown, switches may also be deployed with respect to power island 340 and power island 392 because the power supplied to the circuits associated with the individual voltage islands is controlled by power management circuit 380. It can be deployed so that it can be switched on / off below. In some embodiments of the invention, the aforementioned switches are controlled via register bits assigned to one or more of those switches. Thus, for example, switch 398 can be associated with a particular register bit within power management circuit 380. The switch 398 can be opened when a logic “1” is written to the register bit, and the switch 398 can be closed when a logic “0” is written. Such register bits will be described in detail later with reference to FIG. Further, one of ordinary skill in the art can be used to gate the feed to one or more of power islands 340, 370, 392, and 394 based on the disclosure provided herein. Various controls will be recognized.

  System 300 is divided into multiple power islands so that various parts of system 300 that must be continuously powered are separated from other parts of system 300 that need to be intermittently powered. Is possible. Continuous power feed is maintained for those portions of the system 300 that are required to sensitize the system 300 to a wake-up condition and control the system 300 power-up sequence. Such a continuously powered portion can be included in a power island that is “always on” and, in some cases, powered by a linear regulator built into the system 300. As will be appreciated by those skilled in the art upon reading this disclosure, the various power islands can be powered by a common power source, but they are isolated from each other.

  Similarly, power supply to the controllable storage system 310 is controlled by an Awake signal 328. In the illustrated embodiment shown, power generation is integrated with the motor controller 320. If the awake signal 328 is a logic “0”, power is generated. If the awake signal 328 is logic “1”, no power is generated. The awake signal 328 can also shut down power and control for several motors available in individual hard disk drive systems (eg, including spindle motors and voice coil motors). Shutting down these motors and parking the heads will result in significant system level power savings. This is because these electromechanical interfaces traditionally consume most of the system current of hard disk drive systems. In some embodiments of the present invention, the awake signal 328 is controlled by one or more register bits maintained as part of the current management circuit 380. Such register bits will be described in detail later with reference to FIG. Those skilled in the art will recognize various controls that can be used to gate power to the controllable storage system 310 based on the disclosure provided herein.

In some embodiments of the present invention, one or more programmable discrete I / Os are provided by the power management circuit 380 to control the operation of the system 300. For example, an awake signal 328 is provided that controls the power generation circuitry of the controllable storage system 310. In addition, a signal that shuts down all analog circuits, a signal that shuts down the oscillator circuit, and / or a signal that shuts down all non-essential digital circuits can be included. Those skilled in the art will recognize other I / Os that can be included to control various other functionalities based on the disclosure provided herein.

In one particular embodiment of the present invention, the following exemplary power-up sequence is implemented in system 300 when transitioning from a low power state to a high power state.
(1) The MMC clock 336 is monitored to detect activity at the host interface.
(2) When the host interface 360 detects an activity in the MMC interface 330, the power management circuit 380 is notified.
(3) If the host interface 360 determines that firmware control is required to recover the system 300, the power management circuit 380 commands to turn on all circuitry necessary to operate the firmware. Is done. In some cases, this involves resetting the suicide bit (this will be described in detail later in connection with FIG. 4).
(4) The power management system 380 enables the internal system clock to the host interface 360.
(5) The power management system 380 waits for firmware interaction to set some register bits designed to cause the system 300 to perform an appropriate recovery.
(6) The power management circuit 380 executes the instruction received from the firmware via the register bit of the power management circuit 380.
It should also be noted that various embodiments of the present invention can be designed to support multiple modes of operation. The interaction between the host interface 360 and the power management system 380 may vary somewhat from mode to mode. However, the same power management system 380 can be deployed to support various intended modes of operation by simply changing the firmware that interacts with the power management system 380.

  For example, according to one or more embodiments of the present invention, an ATA mode of operation can be implemented. In such a case, the host mainly controls the sequence of operation modes for turning off the power supply via the host interface 360. However, the power management system 380 can include additional, more aggressive power management functions that can be used within the architecture of the system 300 to better manage the power of the system. Specifically, it is possible to gate the clock and switch the power supply by using adaptive power management that is added to power management by a host command. This can be in addition to the standard sleep and standby modes provided by the ATA bus. In standby, the clock is selectively shut down, but power to the core voltage domain of system 300 is not turned off. On the other hand, all electromechanical interfaces can be shut down. When the host instructs the drive to go to sleep, a suicide mode of operation can be initiated in which power to system 300 is dramatically reduced.

  As another example, an MMC / CE-ATA mode of operation targeted for embedded applications can be implemented. Such a mode provides a serialized physical interface with an overriding ATA instruction set. In such a mode, the host provides little (if any) support when powering off. In such embodiments of the invention, host interface 360 is modified to operate with an external clock in some cases using MMC clock 336. When recovery of the system 300 is requested by such an external clock, the internal clock is enabled via the power management circuit 380 and the recovery process proceeds. In contrast, if the host interface 360 determines that recovery is not required, the system 300 is maintained in its low power state. In some cases, system 300 can be modified to support multiple modes of operation simultaneously. For example, both the MMC / CE-ATA mode and the ATA mode can be simultaneously implemented on the same silicon piece.

  FIG. 3b shows a system 301 adapted for the SATA mode of operation. System 301 is substantially the same as previously described system 300, except that controllable storage system 310 is replaced by a controllable circuit 311 comprising a motor controller 313 and an associated switching regulator 315. That is. Further, the host interface 330 of the system 300 has been replaced with a host interface 331 that includes a host power source 333 and a SATA signal 339. Further, the host interface 360 and associated I / O of the system 300 has been replaced with a SATA PHY layer 367 having interface circuits 345, 347, 349, 355, 357, 359, 361, 363, and 365.

  Since system 301 includes a SATA PHY layer 367, special consideration is given to interaction between the SATA PHY layer 367 and the power management circuit 380. A standard SATA interface has three power states of interest: “active” (SATA PHY layer is ready, fully powered, and both transmit and receive are enabled. ), “Partial” (suppressed power supply and can resume operation in less than 10 microseconds), and “slumber” (suppressed power supply and resumed operation in less than 10 milliseconds) Includes power status. Thus, unlike the previous embodiments of the present invention, these defined power states require a predetermined transient response that is accountable for any operation of the power management circuit 380.

  One or more advantageous features provided by embodiments of the present invention as shown in FIGS. 3 (a) and 3 (b) include system power supply sequence, system clocking, and system reset generation. This includes, but is not limited to, adaptive control. In some cases, only the minimum necessary blocks are activated because additional power is required to meet operational requirements. Specifically, the problem of power loss can be effectively managed by combining power island control and internal clocking control. Further, in some cases, a mini-regulator 342 is used to power an “always-on” power island to manage leakage current.

  FIG. 4 shows a power management system 400 (in broken lines) that can be used in place of the modular adaptive power management module 110. The power management system 400 includes a translator module 410, a register set 420, an adaptive controller 430, a power island controller 440, and an oscillation controller 450. Translator 410 receives information from interface 476, which formats instructions from either or both software 472 running on the host device or firmware 474 embedded as part of a hard disk drive system. Is operable. Power island controller 440 and oscillation controller 450 provide control outputs to various system circuits 480. In some embodiments of the present invention, power management system 400 is only operable to receive instructions from firmware 474. In such a case, firmware 474 can convert all instructions from the host into instructions adapted for implementation by power management system 400. Thus, for example, if a host controller that is executing software 472 requests operation in standby mode, firmware 474 causes power management system 400 to interpret the received instruction in the requested standby mode interpretation. Convert to instructions adapted to work. Further, if an adaptive power management scheme is implemented, firmware 474 controls execution of the adaptive power management scheme by providing one or more instructions to translator 410 via interface 476. In the case described above, translator 410 simply provides a mechanism for writing instructions to the appropriate registers in register set 420.

  Alternatively, some embodiments of the present invention do not include firmware control. In such embodiments, translator 410 can be modified to receive host instructions directly from software 472. In such a case, translator 410 can be designed to recognize one or more host instructions (eg, instructions that indicate a level of activity such as awake, standby, sleep, etc.). . When the translator recognizes a particular instruction, it sets one or more registers in the register set 420 to obtain an interpretation of each instruction that is specifically adapted to the system in which the power management system 400 is deployed.

４個の「電力状態」ビットは、読み取り専用ビットであり、電力管理システム４００が配備されているシステムの電力状態を反映するために連続的に更新されている。これらのレジスタ・ビットは、常時オンの電力ドメインにおいて維持され、したがって、ファームウェアを介してでもホストへでも常時アクセス可能である。動作時には、これら４個の「電力状態」ビットは、適応コントローラ４３０によって示される、次に来る電力状態を伝達するために使用されることが可能である。これらの「電力状態」ビットは、ファームウェアによってプロービングされる場合には、次に来る適応電力状態を選択する契機として使用されることが可能である。電力管理システム４００がスタンバイ・モードまたはスリープ・モードで動作している場合に、「カートリッジ・ウェイク」ビットが論理「１」にセットされると、ウェイクアップ命令として動作する定義済み外部ビットがイネーブルになる。同様に、電力管理システム４００がスタンバイ・モードまたはスリープ・モードで動作している場合に、「プロセッサ・ウェイク」ビットが論理「１」にセットされると、ファームウェアを実行しているプロセッサに対するウェイクアップ命令として動作する定義済み外部ビットがイネーブルになる。 Register set 420 includes a number of registers that can be set to values that control the operation of power management system 400. In certain embodiments of the invention, the register set 420 includes a power mode register 223, a clock control register 224, and a low power control register 225. The power mode register 223 includes a number of bits dedicated to defining the power mode in which the power management system 400 is operating. Specifically, one embodiment of the present invention includes the nine register bits shown in Table 1 below.

The four “power status” bits are read-only bits and are continuously updated to reflect the power status of the system in which the power management system 400 is deployed. These register bits are maintained in the always-on power domain and are therefore always accessible through the firmware and to the host. In operation, these four “power state” bits can be used to convey the next power state indicated by the adaptive controller 430. These “power state” bits, when probed by firmware, can be used as an opportunity to select the next adaptive power state. When the power management system 400 is operating in standby mode or sleep mode, if the “cartridge wake” bit is set to logic “1”, a predefined external bit that operates as a wake-up instruction is enabled. Become. Similarly, when the power management system 400 is operating in standby mode or sleep mode and the “processor wake” bit is set to logic “1”, the wake-up to the processor running the firmware Defined external bits that act as instructions are enabled.

  When the “Suicide” bit is set to logic “1”, the power management system 400 disables the clocks to all modules except those required to restore operation of the system in which the power management system 400 is deployed. To be forced to. In addition, in some cases, setting the “Suicide” bit turns off power to all non-essential power islands. Thus, the “Suicide” bit provides a single writable bit that can be used to completely shut down system operation. In some embodiments, multiple “suicide” bits are provided, each of which is adapted to reduce power consumption in one or more portions of the system in which the power management system 400 is deployed. . For example, in addition to a clock suicide bit that can be operated to gate all non-essential clocks, a power island suicide bit is provided that is operable to power off all non-essential power islands. Is possible. Such an approach further increases the power management flexibility in a given system. This increased flexibility can be used, for example, to turn off all non-essential clocks while maintaining power to the various power islands. This ensures a quicker return from a given power state at the expense of leakage current consumed in the circuits associated with those power islands.

  The “sleep” bit is set to logic “1” whenever a sleep mode is commanded by the host. Note that this register bit is adapted to be used when the host is commanding a power mode compliant with the standard ATA bus. In other embodiments of the invention, other bits may be implemented to command a power mode specific to another standard bus. In some cases, this bit is not set when the power management system 400 is operating under exclusive control of the firmware. When this bit is set to logic “1”, the clock is turned off in all parts of the system where the power management system 400 is deployed. As such, putting the system in an awake state is typically present in an external clock that is supplied only to a limited portion of the design that is capable of restoring operation. Alternatively or additionally, operation can be restored if a reset is received. In addition, when this bit is set, power to all non-essential power islands is turned off. Power supply is recovered under the same conditions used for clock recovery, but may be established somewhat earlier than the clock is reestablished.

  The “standby” bit is set to logic “1” whenever a standby mode is commanded by the host. Note that, like the “sleep” bit, this register bit is adapted to be used when the host is commanding a power mode that conforms to the standard ATA bus. In other embodiments of the invention, other bits may be implemented to command a power mode specific to another standard bus. In some cases, this bit is not set when the power management system 400 is operating under exclusive control of the firmware. When this bit is set, the clock is turned off in most parts of the system where the power management system 400 is deployed. For those parts that control power management and / or take a significant amount of time to reboot, a clock is maintained. The system wakes up from this state whenever an instruction is received from the host or firmware control.

  For some examples only, if the low power mode is to be entered, the “Suicide” bit is set under firmware control. When the “Suicide” bit is set, the power management system 400 forces the clocks from all non-essential modules of the system in which the power management system 400 is deployed to be turned off (and in some cases, Turn off the power supply). Typically, shutting down these clocks saves a significant amount of power at the system level (and shutting down the power supply can further reduce any power consumption).

  Immediately after a power-on reset (power-on reset), all discrete I / O is enabled and the register bits described above are initialized to their default states. Thus, the “Suicide” bit is initialized to an inactive (ie, logic “0”) state. In addition, at this point all power islands are powered and all clocks are active. Such a power on state is indicated in the “power state” bit of the power mode register 423.

  If the system in which the power management system 400 is deployed is set to transition to a power mode that can be achieved adaptively or under the control of the firmware using the “Suicide” bit, Set the "Suicide" bit according to the desired power state. Writing to the “Suicide” bit of the power mode register 423 allows the power management system 400 to initiate the following actions, for example, within the hard disk drive system 100. First, the power management system 400 disables clocks to the various modules of the hard disk drive system 100, for example. This includes disabling the clock to the embedded processor on which the firmware is running. This effectively stops the embedded processor and terminates all firmware activity. Second, power management system 400 sets all register bits in register set 420 to reflect and implement the commanded operating state. This includes shutting off the oscillator that is the source of all clocking, simply gating one or more clocks, and / or turning off power to one or more power islands. May include. Third, the power management system 400 can set a bit that allows the determined instruction to restore system operation. This may include, for example, system recovery as soon as the invariant bit is set by a basic host instruction and / or after a predetermined period of time has elapsed.

  Clock control register 424 includes a number of bits dedicated to defining various clock control schemes that can be used in connection with power management system 400. Each bit of the clock control register 424 controls the power management system 400 to gate the individual circuit portions with which it is associated. For example, if the power management system 400 is used in conjunction with the hard disk drive system 100, one clock control bit is associated with the channel read module 130 and another clock control bit is associated with the hard disk controller 140. It is possible. This allows the clock to the channel read module 130 to be gated while the clock to the hard disk controller 140 remains active. Further, the clock control register 424 may include a bit that gates all clocks by gating the entire system clock. Furthermore, the clock control register 424 includes a bit that, when set, turns off the power to the oscillator completely. Note that the clock control register 424 may include a number of words, each individually addressable, readable, and writable.

  The low power control register 425 has a number of bits that can disable a pull-up resistor or group of pull-down resistors associated with one or more buses controlled by the power management system 400. Thus, for example, separately controlling a group of pull-up resistors associated with a bus supplying data to one module and a group of pull-up resistors associated with another bus supplying data to another module. It is possible. This allows the power loss associated with various pull-up and pull-down resistors in a given design to be flexibly controlled. Further, the low power control register 425 includes a number of bits that are designed to separately control the power supply to one or more power islands under the control of the power management system 400. Thus, it is possible to separately control the power supply to each individual power island of a given design. Yet another bit of the low power control register 425 is to disable the group of tristate signal outputs. In this way, it is possible to minimize the power loss associated with various tristate signal outputs. In addition, the low power control register 425 includes a bit operable to bring all associated systems from a state initiated by setting the “Suicide” bit in the power mode register 425 to an awake state. Is possible. In addition, the low power control register 425 includes bits operable to select an internally generated clock or an externally generated clock to operate the host interface. Note that the low power control register 425 may include a number of words, each individually addressable, readable, and writable.

  The adaptive controller 430 operates to provide a hardwired adaptive power management scheme that is separate and separate from any firmware control. Thus, for example, one or more embodiments of the present invention can be implemented without the assistance of a processor executing firmware. Alternatively, the adaptive controller 430 may be used in the context of a system that includes a processor running firmware to implement some of the functionality that would otherwise be implemented in the form of firmware. Is possible. As yet another alternative, the power management system 400 can be implemented without using the hardwired adaptive controller 430. In some cases, the hardwired adaptive controller 430 includes a group of hardware timers that indicate how long the power management system 400 is operating in a particular power state and is lower It is operable to cause entry into the power state. Those skilled in the art will recognize a variety of circuits that can be used in performing adaptive control by the hardwired adaptive controller 430 based on the disclosure provided herein.

  In addition, the power management system 400 includes a power island controller 440 and an oscillation controller 450. Each of the power island controller 440 and the oscillation controller 450 provides a signal to the system circuit 480 that implements various desired power states. Specifically, the power island controller 440 provides a signal to the system circuit 480 that enables and disables power supply to one or more power islands in the system where the power management system 400 is deployed. It is possible to operate. In some cases, power island controller 440 is just a pass-through (buffered or unbuffered) of register bits included in register set 420. In other embodiments, the power island controller 440 uses the register bits in the register set 420 in addition to the input from the hardwired adaptive controller 430 to control the power supply to the one or more power islands. It is possible to operate. Those skilled in the art will recognize a variety of circuits that can be deployed to implement the aforementioned functionality of power island controller 440 based on the disclosure provided herein. Let's go. Oscillation controller 450 is operable to provide a clock signal or clock enable signal to system circuit 480. This provides the power management system 400 with the ability to turn off or disable one or more clocks used to synchronize various circuits in the system where the power management system 400 is deployed. In some cases, oscillation controller 450 is just a pass-through (buffered or unbuffered) of register bits included in register set 420. In other embodiments, the oscillation controller 450 enables and / or disables one or more clocks using register bits in the register set 420 in addition to inputs from the hardwired adaptive controller 430. Is operable. Those skilled in the art will recognize a variety of circuits that can be deployed to implement the aforementioned functionality of the oscillation controller 450 based on the disclosure provided herein.

  FIG. 5 illustrates a reusable modular power management circuit 500 in accordance with one or more embodiments of the present invention. Modular power management circuit 500 includes a translator interface 510 capable of receiving instructions via a host or host interface 580. Thus, for example, if the host is an ATA hard disk drive system, various ATA instructions can be provided to modular power management circuit 500 via translator interface 510. The translator interface 510 then adapts the standardized instructions from the host interface 580 to implement a power management scheme that is highly adapted to the system in which the modular power management circuit 500 is deployed. Convert to instruction. In some cases, translator interface 510 is implemented through or communicatively coupled to embedded processor 520 that is capable of performing some or all of the conversion functions. Embedded processor 520 executes firmware 530 that can be updated via firmware update interface 530. As such, the power management implemented via the modular power management circuit 500 is modified after deployment to provide different power management schemes as needed by individual end users. Is possible.

  In operation, the translator interface 510 receives instructions from the host 580 and converts them to register writes. Specifically, the translator interface 510 writes to the power island register 540 and the oscillation control register 550. Each bit (545) of power island register 540 is associated with a respective bit of power island I / O buffer 560, and each bit (555) of oscillation control register is a respective bit of oscillation control I / O buffer 570. Associated with. The power island I / O buffer 560 drives a number of power island control signals 590 including power island controls 1-N (591-594). The designer uses the power island control signals 591-594 to drive the respective switches that gate the power supply to the various power islands formed in the system in which the modular power management circuit 500 is deployed. Each can be used. Similarly, the oscillation control I / O buffer 570 drives several oscillation control signals 595 including oscillation control signals 1 to N (596 to 599). The designer uses each of the oscillation control signals 596-599 to turn off the oscillator used in the system where the modular power management circuit 500 is deployed or drive the respective circuit that gates the clock. Is possible. Thus, as one specific example, one of the oscillation control signals 596-599 can shut down a particular oscillator, and another signal of the oscillation control signals 596-599 can be All clocks derived from the oscillator can be gated, and yet another signal of the oscillation control signals 596-599 is not all of the clocks derived from that particular oscillator. It is possible to gate. These three separate oscillation controls help to implement different levels of power management and / or moderate power management aggressiveness at any level of power management. One of ordinary skill in the art will be aware of the myriad power management controls that can be implemented using modular power management circuitry according to embodiments of the present invention based on the disclosure provided herein. Will.

  Although only one specific example of the operation of modular power management circuit 500, translator interface 510 can receive ATA sleep mode commands from host 580. Translator interface 510 then translates the received instruction into a register write that causes one or more power islands to be shut down and one or more clocks to be turned off or gated. It is possible. As such, the modular power management circuit 500 provides a modular platform that allows designers to easily implement custom power management schemes that interact with standard power management instructions. Thus, power management can be made more sophisticated by using reusable circuitry.

  Finally, the present invention provides a novel system, apparatus, method, and configuration for power management. Up to now, one or more embodiments of the present invention have been described in detail, but those skilled in the art will recognize various alternatives, modifications, and equivalents that do not depart from the spirit of the present invention. Let's go. Accordingly, the above description should not be taken as limiting the scope of the invention, which is defined by the appended claims.

Claims (5)

A hard disk controller,
An interface controller;
A channel readout module;
A hard disk drive system including a power management system,
The hard disk controller, the interface controller, the channel read module, and the power management system are implemented across two or more separate power islands, and the power supply to the two or more separate power islands is: A hard disk drive system controlled at least in part by the power management system.
The hard disk drive system according to claim 1, wherein
The hard disk controller, the interface controller, the channel read module, and the power management system utilize two or more separate clocks, each of the two or more separate clocks being powered by the power management system. Hard disk drive system controlled.
The hard disk drive system according to claim 1, wherein
The hard disk drive system, wherein the interface controller is operable to provide a translator interface selected from the group consisting of an ATA interface, a SATA interface, an MMC interface, a CE-ATA interface, and an SDIO interface.
The hard disk drive system according to claim 1, wherein
A hard disk drive system further comprising a processor executing firmware, wherein the firmware is operable to implement an adaptive power management scheme via the power management system.
The hard disk drive system according to claim 4,
The interface controller is operable to provide a translator interface that receives host instructions that cause a selected power mode to be implemented via the power management system. Hard disk drive system that is operable to

Images