Reallocation of regulator phases within a phase-redundant voltage regulator apparatus

Abstract

A phase-redundant voltage regulator apparatus includes groups of regulator phases, each having a multi-phase controller (MPC) connected to each regulator phase. The MPC transfers, to control logic, phase fault signals and a shared current (ISHARE) phase control signal received from each dedicated regulator phase of a phase group. Spare regulator phases include output ORing devices to limit current flow into spare regulator phase outputs. Output switching devices are configured to electrically couple spare regulator phase outputs to a common regulator output. Control logic is connected to the phase groups MPC and asserts phase enable signals to, transfers ISHARE phase control signals to, and receives phase fault signals from the spare regulator phases. The control logic electrically interconnects a spare regulator phase to a phase group including a failed regulator phase in response to receiving a phase fault signal from an MPC.

Description

BACKGROUND

The present disclosure generally relates to voltage regulator circuits. In particular, this disclosure relates to reallocation of shared, redundant regulator phases within a phase-redundant voltage regulator circuit.

A voltage regulator is an electronic device or system designed to receive an input voltage and automatically maintain a constant voltage level on one or more output terminals. Depending on the design, a voltage regulator can be used to regulate one or more alternating current (AC) or direct current (DC) voltages. Voltage regulators can be included in electronic devices such as computer power supplies where they can be used to stabilize DC voltages used to supply power to electronic components such as processor(s), memory devices, and other types of integrated circuits (ICs). A voltage regulator circuit can receive a feedback voltage received from a sense point located adjacent to electronic component(s) that the voltage regulator supplies power to. This feedback voltage can be used to modulate an output voltage of the voltage regulator. This modulated output voltage can result in the voltage received by the supplied electronic component(s) being maintained at a stable value regardless of the current draw of the components or voltage drop across conductors interconnecting the voltage regulator to the components.

A field-effect transistor (FET) is a transistor that uses an electric field to control the shape and hence the conductivity of a channel of one type of charge carrier in a semiconductor material. FETs can be unipolar transistors as they can involve single-carrier-type operation. FETs can be majority-charge-carrier devices, in which the current flow is carried predominantly by majority carriers, or minority-charge-carrier devices, in which the current is mainly due to a flow of minority carriers. A FET device can consist of an active channel through which charge carriers, electrons or holes, flow from a source to a drain. Source and drain terminal conductors can be connected to the semiconductor through ohmic contacts. The conductivity of the channel can be a function of the electric potential applied across the gate and source terminals.

SUMMARY

Embodiments can be directed towards a phase-redundant voltage regulator apparatus. The phase-redundant voltage regulator apparatus can include a plurality of regulator phases, each including a regulator electrically coupled to receive, at a regulator input, an input voltage and provide, at a regulator output, a respective output voltage. The phase-redundant voltage regulator apparatus can include a set of phase groups of the plurality of regulator phases including a first and a second phase group. Each phase group can include a common regulator input electrically interconnected to regulator inputs of regulators of the phase group, a common regulator output electrically interconnected to regulator outputs of regulators of the phase group and at least one redundant regulator phase. Each phase group can also include at least one dedicated regulator phase of the plurality of regulator phases and at least one spare regulator phase of a set of spare regulator phases. Each phase group can also include a multi-phase controller (MPC) electrically coupled to each dedicated regulator phase of the phase group, the MPC configured to transfer, to control logic, phase fault signals and a shared current (I_SHARE) phase control signal received from each dedicated regulator phase of a respective phase group. The phase-redundant voltage regulator apparatus can also include the set of spare regulator phases of the plurality of regulator phases including a first and a second spare regulator phase. Each spare regulator phase can include a secondary output ORing device electrically coupled and configured to limit current flow into a secondary output of the spare regulator phase and a first output switching device configured to, in response to a first phase enable signal, electrically couple the regulator output of the spare regulator phase to a first common regulator output. Each spare regulator phase can include a second output switching device configured to, in response to a second phase enable signal, electrically couple the regulator output of the spare regulator phase to a second common regulator output and control logic. The control logic is electrically connected to MPCs of each phase group of the set of phase groups, the control logic configured to receive I_SHARE phase control signals from, and exchange phase fault signals with the MPCs. The control logic is also electrically connected to spare regulator phases of the set of spare regulator phases. The control logic is also configured to assert phase enable signals to, transfer I_SHARE phase control signals to, and receive phase fault signals from the spare regulator phases. The control logic is also configured to, in response to receiving a phase fault signal from an MPC, electrically interconnect a spare regulator phase to a phase group that includes a failed regulator phase.

Embodiments can also be directed towards a method for reallocating a set of spare voltage regulator phases between phase groups of voltage regulator phases. The method includes using control logic that is responsive to a system control function and responsive to monitored phase fault signals received from the phase groups. The method includes using the control logic to store, into a non-volatile memory within the control logic, an association between a first portion of the set of spare voltage regulator phases and an “allocated” status. The method also includes using the control logic to store, with the control logic, into the non-volatile memory within the control logic, an association between a second portion of the set of spare voltage regulator phases and an “unallocated” status. The method also includes using the control logic to detect, with the control logic, a phase fault signal from a first compromised phase group of the phase groups. The method also includes using the control logic to transfer, in response to detecting the phase fault signal, at least one spare voltage regulator phase of the second portion of the set of spare voltage regulator phases to the first compromised phase group.

Embodiments can also be directed towards a method for reallocating a set of spare voltage regulator phases between phase groups of voltage regulator phases, the method comprising using control logic that is responsive to a system control function and responsive to monitored phase fault signals received from the phase groups. The method includes using the control logic to store, into a non-volatile memory within the control logic, an association between a first portion of the set of spare voltage regulator phases and an “allocated” status. The method includes using the control logic to store, with the control logic, into the non-volatile memory within the control logic, an association between a second portion of the set of spare voltage regulator phases and an “unallocated” status. The method includes using the control logic to detect, with the control logic, a phase fault signal from a first compromised phase group of the phase groups. The method includes using the control logic to transfer, in response to detecting the phase fault signal, at least one spare voltage regulator phase of the second portion of the set of spare voltage regulator phases to the first compromised phase group.

The above summary is not intended to describe each illustrated embodiment or every implementation of the present disclosure.

BRIEF DESCRIPTION OF THE DRAWINGS

The drawings included in the present application are incorporated into, and form part of the specification. They illustrate embodiments of the present disclosure and, along with the description, serve to explain the principles of the disclosure. The drawings are only illustrative of certain embodiments and do not limit the disclosure.

FIG. 1 is a block diagram depicting a phase-redundant voltage regulator apparatus, according to embodiments of the present disclosure.

FIG. 2 includes two views depicting a voltage regulator phase and a spare voltage regulator phase, according to embodiments consistent with the figures.

FIG. 3 is a block diagram depicting a phase-redundant voltage regulator apparatus with shared, assignable redundant spares, according to embodiments consistent with the figures.

FIG. 4 is a flow diagram depicting a method for transferring an assignable spare regulator phase between phase groups of voltage regulators, according to embodiments consistent with the figures.

FIG. 5 is a flow diagram depicting a method for reallocating regulator phases between phase groups of regulators, according to embodiments consistent with the figures.

FIG. 6 is a flow diagram depicting a method for reallocating regulator phases between phase groups of regulators, according to embodiments consistent with the figures.

While the invention is amenable to various modifications and alternative forms, specifics thereof have been shown by way of example in the drawings and will be described in detail. It should be understood, however, that the intention is not to limit the invention to the particular embodiments described. On the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the invention.

In the drawings and the Detailed Description, like numbers generally refer to like components, parts, steps, and processes.

DETAILED DESCRIPTION

Certain embodiments of the present disclosure can be appreciated in the context of reallocating shared redundant regulator phases within a phase-redundant voltage regulator circuit. Reallocating such shared regulator phases can be useful in providing, at reduced cost and design complexity, phase redundancy and robust power delivery. The reallocating of shared redundant regulator phases within a voltage regulator circuit can provide increased reliability power delivery to electronic systems.

Embodiments can provide reallocating of shared redundant regulator phases to electronic equipment such as servers, which can be used to provide data to clients attached to a server through a network. Such servers can include, but are not limited to web servers, application servers, mail servers, and virtual servers. While not necessarily limited thereto, embodiments discussed in this context can facilitate an understanding of various aspects of the disclosure. Certain embodiments can also be directed towards other equipment and associated applications, such as providing reallocating of shared redundant regulator phases to electronic equipment such as computing systems, which can be used in a wide variety of computational and data processing applications. Such computing systems can include, but are not limited to, supercomputers, high-performance computing (HPC) systems, and other types of special-purpose computers. Embodiments can also be directed towards reallocating of shared redundant regulator phases for equipment used in such as telecommunications, airborne, and automotive applications.

The acronym “FET” is used herein in reference to a field-effect transistor which can be useful, within a voltage regulator design, to interconnect two circuit nodes by providing a relatively low impedance electrical connection between the nodes. It can be understood that in embodiments, various types of FETs can be selected by a voltage regulator designer to meet electrical performance criteria of certain voltage regulator designs. Such FET types can include, but are not limited to enhancement-mode, depletion-mode, N-channel field-effect transistor (NFET) and P-channel field-effect transistor (PFET) devices. Such FETs can also be referred to “power FETs,” and “power metal-oxide-semiconductor field-effect transistors” (MOSFETs), without loss of meaning. It can also be understood that, in some embodiments, various other types of electronic devices can be used in place of FETs. Such electronic devices can include, but are not limited to, bipolar devices such as NPN and PNP transistors as well as transistors fabricated in other semiconductor technologies.

For ease of discussion and illustration, the terms “ORing device” and “ORing FET” can be used interchangeably, without loss of meaning, to refer to a semiconductor device configured to prevent the flow of current into a voltage regulator phase output. Although a FET symbol, e.g., 244, FIG. 2, has been used to represent such a device in the Figures, this is not to be construed as limiting; other types of devices, as described above, can also be used for a similar purpose.

The descriptor “ORing,” used herein in reference, for example, to “ORing devices,” “ORing FETs” and “ORing regulators,” can be understood as referring to the logical OR function and output protection function of a device, e.g., FET, or voltage regulator. Such devices and voltage regulators are protected from reverse current flow into a device output by a FET configured to perform in a similar manner as an output protection diode.

The terms “phase” and “regulator phase” are used interchangeably herein in reference to a redundant voltage regulator used within a voltage regulator apparatus. The voltage regulator phase is generally activated by a multi-phase controller (MPC) at a time that is different than activation times of other voltage regulators. This is commonly referred to as being activated “out of phase” with the other voltage regulators.

A wide range of electronic systems employ voltage regulators to provide power delivery, at a stable voltage, to electronic components within the systems. Such electronic systems can include, but are not limited to, computers, computing equipment, servers and equipment used in telecommunications, airborne, and automotive applications. Components within the systems can include, but are not limited to, central processing units (CPUs), graphics processing units (GPUs), other integrated circuit (IC) types, hard disk drives, solid-state drives (SSDs), memory and other types of electronic components and devices.

A voltage regulator can be electrically connected between a power supply and components specified to receive a voltage that is lower than a voltage provided by the power supply. For example, a power supply can provide a variety of voltages, for example 12 V, 24 V or 48 V DC, to a voltage regulator. Output voltages of the voltage regulator can include, for example, 3.3 V, 2.5 V, 1.8 V, 1.0 V, 0.7 V or any other voltage(s) specified as suitable for the powered electronic components and systems.

Voltage regulators including parallel-connected phase-redundant regulators can provide numerous benefits to various electronic systems that they supply power to. For example, such voltage regulators can provide increased system reliability and robustness resulting from the ability of a voltage regulator apparatus to dynamically and automatically replace a failed or faulted regulator phase with a redundant spare or “backup” phase. A regulator phase can fail due to the failure, e.g., short-circuiting, of one or more components such as capacitors, FETs, amplifiers or driver circuits within the regulator phase.

Electronic system reliability can also be increased as a result of reduction of the number of decoupling capacitors within a phase-redundant voltage regulator design. This reduction in capacitor count can result in a lowering of the effective failure rate of the voltage regulator. The addition of one or more redundant or spare phases to a voltage regulator design can also increase the reliability of the design by lowering the overall current demands placed on each of the multiple redundant phases included within the voltage regulator.

The use of phase-redundant voltage regulators having phases activated at staggered times can also result in decreased output voltage ripple and enhanced transient response to variations in current load placed on the phases. Multiple redundant regulator phases can be particularly useful in managing overall voltage regulator system cost. According to embodiments, the cost of a voltage regulator that uses multiple smaller, redundant regulator phases can be significantly less than the cost of a regulator that uses fewer, larger regulator phases each having greater current output.

Certain types of electronic systems can use phase-redundant voltage regulators to supply power to certain subsystems and electronic components. In order to assure reliable operation of the system in the event of failure of one or more of the voltage regulators, additional redundant voltage regulator phases can be added to each individual voltage domain or output level in order to restore regulator current supply capacity resulting from the failure of one or more voltage regulator phases. In addition to voltage regulator phase failures, the current supply needs of particular voltage domains may change over time, as a result of varying system configurations and/or throttled load circuits, e.g., CPU, GPU, or memory circuits

However, a methodology of adding additional redundant voltage regulators to each voltage level may not be particularly efficient or cost-effective. Excessive and unnecessary cost, consumed power, system space/area, and design complexity can result from the use of this methodology. A more efficient voltage regulator design methodology is desirable; one which reduces the number of components used, cost, consumed power, system area, and design complexity.

According to embodiments, redundant, spare voltage regulators can be added to a voltage regulator design, e.g., either within or outside of phase groups, and can be allocated or shared between various voltage domains on an “as needed” basis. Control logic can monitor the voltage supply circuits for voltage regulator failures, and can, in response to regulator failures, system load and system configuration changes, supervise the reallocation of spare voltage regulator phases as needed, in order to provide robust, uninterrupted power delivery on all voltage levels/domains.

The figures herein depict certain example circuits, functions, electrical interconnections and component interactions used to implement embodiments of the present disclosure. These are provided by way of example, and are not to be construed as limiting. Embodiments can also include circuits, functions, electrical interconnections and component interactions not described or depicted herein, within the spirit and scope of the present disclosure.

For ease of illustration and discussion, a limited number of regulator phases, regulator phase groups and spare regulator phases are depicted and discussed herein. For example, two regulator phase groups each including two regulator phases and one spare regulator phase, in conjunction with two spare regulator phases independent of phase groups can be used to illustrate embodiments of the present disclosure. However, this is not to be construed as limiting; any other number of regulator phases, regulator phase groups and spare regulator phases can be used within embodiments. According to embodiments and design practices, the number “N” of regulator phases used to satisfy the current requirements of a voltage domain within an electronic system can be a relatively small number such as 1 or 2, or it can be much larger, such as 20 or more. Any number of regulator phase groups can be used within an electronic system to satisfy the needs for particular, unique voltage domains. A number of spare regulator phases can be selected by a power system designer based upon an expected average failure rate of a regulator phase, as well as other design criteria.

Embodiments of the present disclosure can be useful in reducing voltage regulator packaging size and cost, relative to voltage regulators not using shared, assignable redundant regulator phases to provide regulated voltage levels. An electronic system configured according to embodiments of the present disclosure can have increased reliability and reduced cost, complexity, parts count, and packing area for included voltage regulators.

A phase-redundant voltage regulator apparatus with shared, assignable redundant regulator phases, designed according to certain embodiments, can be compatible with existing and proven electronic systems, and can be a useful and cost-effective way to add shared, assignable redundant regulator phases to voltage supplies powering an electronic system. A phase-redundant voltage regulator apparatus constructed according to embodiments of the present disclosure can be installed within an existing electronic system.

Embodiments of the present disclosure can be useful for implementing phase-redundant voltage regulators with shared, assignable redundant regulator phases for use within electronic systems by using existing and proven IC and printed circuit board (PCB) fabrication technologies and material sets, electronic design methodologies, design tools, and manufacturing processes.

FIG. 1 is a block diagram depicting a phase-redundant voltage regulator apparatus 100 that includes multiple redundant voltage regulators phases 126A, 126B and 126C. Multiple redundant regulator phases 126A, 126B and 126C are electrically connected in parallel between common regulator input V_IN and common regulator output V_OUT, each of the regulator phases receiving an input voltage at V_IN input 136, and providing an output voltage on V_OUT output 148.

A quantity of “N+1” or “N+2” voltage regulator phases can be electrically connected in parallel, where “N” represents a minimum number of phases needed to supply a specified current, and the additional one or two phases can be useful in replacing one or two failed voltage regulator phases. In the case of failure or “faulting” of one or more redundant phases, faulty redundant phase(s) can be disabled in order to share a current load between the remaining active phases, and thus to ensure uninterrupted power delivery. A redundant phase can also be used to implement “masked redundancy,” meaning that when a phase fails, the fault is not reported to a system control function 108. The redundant phase can then be used to create a high-reliability regulator instead of being used as a redundant phase. This swapping of regulator phases can be controlled by MPC 122.

Each of the regulator phases 126A, 126B and 126C includes a buck regulator 116, an input protection device 114, an ORing device 118 and a phase-redundant controller 106. The phase-redundant controller 106 can be useful, in conjunction with input protection device 114, in isolating the input of a failed buck regulator 116 from the other phases in the apparatus 100. Phase-redundant controller 106 can monitor input and output current and output voltage of a regulator phase, e.g., 126A, and can control input protection device 114 in response to, for example, abnormal currents or voltages within the phase. Such abnormal currents or voltages can result from the failure, e.g., short-circuiting, of components such as capacitors or FETs within buck regulator 116.

The input protection device 114 can be used for providing input overcurrent protection and output overvoltage protection for a respective regulator phase, e.g., regulator phase 1126A. Input protection device 114 can protect the regulator phase 1126A by electrically isolating, i.e., disconnecting, the common regulator input V_IN from buck regulator 116, in response to a signal generated by phase-redundant controller 106. Each of the regulator phases 126A, 126B and 126C also includes an ORing device 118 that can be used to limit or prevent reverse current flow into the V_OUT output 148 of the regulator phase. Such reverse current flow could possibly result from the short circuiting or failure of a FET or a capacitor within a regulator phase.

An MPC 122 is electrically coupled to each of the regulator phases 126A, 126B and 126C, through detected current outputs 102 and control signals 124. Master controller 112 of the MPC 122 generates control signals 124 to periodically and sequentially activate each of the regulator phases 126A, 126B and 126C for predetermined periods of time. In some applications this activation can be used, for example, to generate controlled current sharing between phases. The MPC 122 can be used to maintain current sharing between active regulator phases following a failure or fault of one or more of the regulator phases with one or more phases provided for enabling redundancy. Current sharing can also be implemented between multiple active regulator phases when none of the phases have previously failed.

Redundant faults reporting circuit 104 of the MPC 122 can collect and report phase faults/failures based on the multiple detected current signals presented on detected current outputs 102A, 102B and 102C from the respective regulator phases. The detected current signals can be used and interpreted in order to indicate the failure of one or more particular regulator phase(s). In some applications, the MPC 122 can provide reduced ripple and transient response time from the redundant regulator phases by staggering the respective V_OUT outputs 148 to be out-of-phase with each other. The MPC 122 is also configured to adjust the control signals 124 in response to a feedback output voltage, received at feedback input 101, from the common regulator output V_OUT.

The serial control bus 120 can be used to interconnect the regulator serial interface 110 of the MPC 122 to a system control function 108. The regulator serial interface 110 can send and receive control and monitoring signals to the system control function 108, which can exert control over one or more phase-redundant regulator apparatuses 100, within an electronic system. In applications, system control function 108 can represent a hardware and/or software unit that can be used in an electronic system such as a computer or server, to monitor and control various aspects of the systems hardware functions. A system control function 108 may be used, within an electronic system to monitor and control functions such as power supply and voltage regulator functions, system clock frequency, cooling, and the like.

In some applications, the serial control bus 120 can be, for example, a Serial Peripheral Interface (SPI) bus, a Power Management Bus (PMBus), or an Inter-Integrated Circuit (I²C) interface. The serial control bus 120 can be used to send monitoring data, for example, indicating which regulator phases have failed or faulted, to the system control function 108, and can also receive commands and controls from the system control function 108.

FIG. 2 includes two block diagram views including a view depicting a voltage regulator phase 126 and a view depicting a spare voltage regulator phase 227, according to embodiments consistent with the figures. The various circuits, functions and functional blocks depicted in FIG. 2 are generally consistent with those depicted in and described with reference to FIG. 1. FIG. 2 can be useful in providing an expanded, more detailed depiction of these functions, as well as depicting the addition of switch devices 217 and 219 and ORing devices 118, 118A to a regulator phase, e.g., 126A, FIG. 1 in order to form spare voltage regulator phase 227. It can be understood that view 126 corresponds to a single regulator phase, e.g., regulator phase 1126A, FIG. 1. The addition of multiple switch devices 217 and 219 and ORing devices 118, 118A can be particularly useful in allowing the regulator output 242 of buck regulator 116 to be selectively electrically connected to more than one output, e.g., primary output 204 and secondary output 206 of the spare voltage regulator phase 227.

Voltage regulator phase 126 includes buck regulator 116 which can receive an input voltage at regulator input 240 and drive an output voltage onto regulator output 242. Driver M1 is configured to enable and disable FETs within buck regulator 116, in response to a control signal, e.g., 124, FIG. 1, received at control input 250. According to embodiments, control signal 124 can be a digital signal having logical “0” and logical “1” levels of 0 V and 3.3 V, respectively. In some embodiments, other voltage levels may be used.

Voltage regulator phase 126 also includes an input protection device 114, an ORing device 118 and a phase-redundant controller 106. Phase-redundant controller 106 includes current detect circuit 234, output overvoltage protect circuit 232 and input overcurrent protect circuit 230. Input overcurrent protect circuit 230 is electrically connected and configured to monitor current received at the V_IN input 136, while output overvoltage protect circuit 232 can monitor output voltage at regulator output 242. Latch 228 is part of input overcurrent protect circuit 230, which is electrically connected to output overvoltage protect circuit 232. Current detect circuit 234 can be used to monitor output current at regulator output 242.

Phase-redundant controller 106 can be useful in monitoring input and output current and output voltage of the voltage regulator phase 126 and can, in response, control input protection device 114, which can be used for providing input overcurrent protection and output overvoltage protection for voltage regulator phase 126.

Phase-redundant controller 106 can control input protection device 114 through asserting the Q output of latch 228 of input overcurrent protect circuit 230. The Q output of latch 228 is used to control a gate input G input of input protection FET 238. Input protection FET 238 has a drain input D coupled to the V_IN input 136 a source input S coupled to the regulator input 240. A control signal applied to a gate input G of input protection FET 238 can activate or deactivate input protection FET 238, which either electrically connects or disconnects, respectively, V_IN input 136 and regulator input 240 of buck regulator 116. The control of input protection device 114 by phase-redundant controller 106 can be useful in providing input overcurrent protection and output overvoltage protection to the voltage regulator phase 126 by electrically isolating the regulator input 240, when buck regulator 116 faults or fails, from the other phases within a voltage regulator apparatus.

In some applications, current detect circuit 234 can provide, on detected current output 202, a signal having an analog voltage level between 0 V and 3.3 V that proportionally represents a level of detected current flowing out of V_OUT output 148 the voltage regulator phase 126. In some applications, other analog voltage levels/ranges may be used. In embodiments, the detected current signal can be useful in indicating the failure of a particular regulator phase.

ORing device 118 can be used to limit or prevent reverse current flow into the V_OUT output 148 of the voltage regulator phase 126, and thus also regulate reverse current flow into regulator output 242. Such reverse current flow can result from the short-circuiting or failure of a FET or a capacitor within buck regulator 116. Comparator 246 has inputs electrically connected to a source terminal S and to a drain terminal D of the output ORing FET 244. The output of comparator 246 is electrically connected to gate terminal G of the output ORing FET 244. Comparator 246 configured and connected to, in conjunction with the output ORing FET 244, limit current flow into regulator output 242, in response to a voltage at V_OUT output 148 that is greater than a voltage at regulator output 242. In the case where a voltage at V_OUT output 148 is greater than a voltage at regulator output 242, indicating a reverse current flow, the comparator 246 outputs to the gate G of output ORing FET 244 a low voltage, thus disabling output ORing FET 244 and preventing further reverse current flow.

The phase-redundant controller 106, input protection device 114 and buck regulator 116 of spare voltage regulator phase 227 are generally consistent with voltage regulator phase 126 in regards to electrical interconnection and function. Spare voltage regulator phase 227 also includes output switching devices 217 and 219, which for ease of illustration and discussion, are depicted as and referred to as “FET 217” and “FET 219” herein, without loss of meaning. In embodiments, switch devices 217 and 219 can each also be an NFET, a PFET, an NPN transistor, a PNP transistor, or other suitable type of transistor or semiconductor device.

FET 217 and FET 219 each have a gate input G electrically connected to enable input 216 and enable input 218, respectively. These interconnections allow FET 217 and FET 219 to be activated, i.e., enable an electrically conductive path between the drain D and source S terminals, in response to phase enable signals received on respective enable inputs 216 and 218 of phase-redundant controller 106. This activation can be used to electrically connect regulator output 242 of buck regulator 116, through ORing device 118 and secondary ORing device 118A, respectively, to either primary output 204 or secondary output 206. In embodiments, primary output 204 and secondary output 206 can be electrically connected to common regulator outputs of a phase group (See FIG. 3).

The electrical configuration and function of ORing device 118 and secondary ORing device 118A of spare voltage regulator phase 227 are consistent with the configuration and function of ORing device 118 of voltage regulator phase 126. Similar to ORing device 118 of voltage regulator phase 126, ORing device 118 and secondary ORing device 118A of spare voltage regulator phase 227 can be used to limit or prevent reverse current flow into the primary output 204 and secondary output 206, respectively, of the spare voltage regulator phase 227, and thus also regulate reverse current flow into regulator output 242.

FIG. 3 is a block diagram depicting a phase-redundant voltage regulator apparatus with shared redundant spares 300, according to embodiments generally consistent with the figures. FIG. 3 is particularly useful in depicting a phase-redundant voltage regulator apparatus, consistent with FIG. 1 and FIG. 2 that includes shared spare voltage regulator phases 227A, 227B, 227C and 227D that are electrically assignable to phase groups 374 and 376.

Numerous aspects of the embodiments depicted in FIG. 3 are particularly consistent with those depicted in FIGS. 1 and 2, and described in the associated text, and will not be further described herein. These aspects include circuits, logical and control functions, interactions between functions, electrical interconnections, signal usage and signal voltage ranges.

The elements of phase-redundant voltage regulator apparatus 300 including control logic 366, MPCs 122, voltage regulator phases, e.g., 126A, and spare voltage regulator phases, e.g., 227A, are electrically interconnected and thereby communicate through the use of various types of signals. These types include phase fault signals, phase enable signals, and I_SHARE phase control signals. These signal types can be useful in enabling both monitoring of regulator phase function and faults, as well as control and reallocation of regulator phases by control logic 366.

Phase fault signals are used in embodiments to indicate a fault or failure of a particular regulator phase within the voltage regulator apparatus 300. Phase fault signals can be generated by a redundant faults reporting circuit 104, in response to detected current outputs 102 received from individual phase-redundant controllers (PRCs) 106 of voltage regulator phases, e.g., 126A. For example, a detected current output 102 received by redundant faults reporting circuit 104 that is significantly higher or lower than other detected current outputs 102 can resulting in redundant faults reporting circuit 104 generating a respective phase fault signal for the corresponding phase. Phase fault signals can be also be generated directly by PRCs 106 of spare regulator phases, e.g., 227A, 227B, 227C and 227D.

Phase fault signals include V₁ fault in signal 302A, V₂ fault in signal 302B, V₁ N−1 fault signal 308A, V₁ N−2 fault signal 310A, V₂ N−1 fault signal 308B, V₂ N−2 fault signal 310B, S1 fault signal 312A, S2 fault signal 312B, S3 fault signal 312C and S4 fault signal 312D.

The naming convention of phase fault signals, as listed above, includes an associated common regulator output voltage, e.g., “V₁” or “V₂,” a spare regulator indicator, e.g., “S1,” “S2,” “S3” or “S4,” corresponding to spare regulator phases 227A, 227B, 227C and 227D, respectively, and a fault “level” indicator, e.g., “N−1” or “N−2.” A fault level indicator of “N−1” indicates that the fault signal indicates a “single-fault” type, that is, the particular phase group from which the fault signal originates has only one reported fault among the voltage regulators in the phase group. Similarly, a fault level indicator of “N−2” indicates that the fault signal indicates a “double-fault” type, i.e., the particular phase group from which the fault signal originates has two reported faults among the voltage regulators in the phase group. A “spare fault” signal, e.g., S1 fault signal 312A, can be used to indicate the failure of a spare regulator phase, e.g., 227A. The various phase fault signals can be useful in providing indications of faults that can be used by control logic 366 in determining how to re-allocate spare voltage regulator phases to provide continued, robust voltage regulator performance in the event of one or more phase failures.

Phase enable signals are used in embodiments to enable the outputs of spare regulator phases, e.g., 227A, in order to selectively interconnect them to common regulator outputs, e.g., V₁. In embodiments, phase enable signals are generated by control logic 366. Phase enable signals include S1 enable V₁ signal 316A, S1 enable V₂ signal 318A, S2 enable V₁ signal 316B, S2 enable V₂ signal 318B, S3 enable V₁ signal 316C, S3 enable V₂ signal 318C, S4 enable V₁ signal 316D and S4 enable V₂ signal 318D. The naming convention of phase fault signals, as listed above, includes an associated common regulator output voltage, e.g., “V₁” or “V₂,” and an associated spare regulator indicator, e.g., “S1,” “S2,” “S3” or “S4.”

In embodiments, I_SHARE phase control signals are signals that represent, through an analog voltage, an amount of current to be provided by each regulator phase within a phase group. In embodiments, a current “I_SHARE”=I_TOTAL/N, where =I_TOTAL is the total current output of a phase group and “N” is the number of phases in the phase group. In some embodiments, I_SHARE signals can be analog voltages in a range between 0 V and 3.3 V, respectively. By way of example, an I_SHARE signal voltage of 0 V can represent 0 A of regulator phase current output, while an I_SHARE signal voltage of 3.3 V can represent the full rated, e.g., maximum, regulator phase output current. I_SHARE voltages between 0 V and 3.3 V can accordingly represent regulator phase output currents linearly proportional, between OA and the full rated regulator phase output current, to the respective I_SHARE voltage. In some embodiments, other scaling/control schemes and/or voltage ranges may be used.

An I_SHARE phase control signal can be generated, for example, by an MPC 122 of a phase group, e.g., 374, as an indicator of the current output required from each voltage regulator within the phase group. The I_SHARE phase control signal may vary over time, in order to modulate the voltage regulator's output to supply particular dynamic current loads. An I_SHARE phase control signal can also be received and used to modulate a spare voltage regulator phase in order to cause it to perform similarly to active voltage regulator phases within a phase group. Thus, an I_SHARE phase control signal, in conjunction with phase enable signals, can be particularly useful in replacing a failed voltage regulator phase with an active spare voltage regulator phase. I_SHARE phase control signals include V₁ I_SHARE S1 signal 304A, V₁ I_SHARE S2 signal 306A, V₂ I_SHARE S1 signal 304B, V₂ I_SHARE S2 signal 306B, S1 I_SHARE signal 314A and S2 I_SHARE signal 314B, S3 I_SHARE signal 314C and S4 I_SHARE signal 314D.

According to embodiments, phase fault signals and phase enable signals can be digital signals having logical “0” and logical “1” levels of 0 V and 3.3 V, respectively. In some embodiments, other voltage levels can be used. FIG. 3 depicts the above-described signals as lines with arrowheads; it can be understood that the end of the signal line having the arrowhead indicates the destination of the particular signal, while the opposite end indicates the source of the signal.

Phase-redundant voltage regulator apparatus 300 includes phase groups 374 and 376, each including multiple voltage regulator phases and an interconnected MPC 122. Phase group 374 includes voltage regulator phases 126A and 126B and spare regulator phase 227C. Phase group 376 includes voltage regulator phases 126A and 126B and spare regulator phase 227D. Voltage regulator apparatus 300 also includes spare regulator phases 227A and 227B. Each of the spare regulator phase 227A, 227B, 227C and 227D includes output switching devices 217 and 219, and ORing devices 118 and 118A. Control logic 366, including non-volatile memory 368 and regulator serial interface 110, is electrically connected to each of the voltage regulator phases, spare voltage regulator phases, and MPCs.

Consistent with FIG. 1, each phase group 374 and 376 includes multiple redundant voltage regulator phases 126A, 126B and spare regulator phases 227C and 227D, respectively, which can be useful for providing phase-redundancy in the delivery of power to common regulator outputs V₁ and V₂, respectively. The V_OUT output 148 of each voltage regulator phase within a phase group is connected to a common regulator output, e.g., V₁. It can be understood that the V_OUT outputs 148 of regulator phases 126A, 126B of each phase group are each wired to only one common regulator output; either V₁ or V₂. Regulator phases 126A, 126B are thus referred to herein as “dedicated” voltage regulator phases, as the V_OUT output 148 cannot be reconfigured to be connected to another common regulator output. In contrast, spare regulator phases, e.g., 227A, 227B, 227C and 227D can be dynamically reconfigured or reassigned by control logic 366 to connect regulator output 242 of buck regulator 116 of any of the spare regulator phases to either common regulator output V₁ or V₂.

For ease of illustration and discussion, phase-redundant voltage regulator apparatus 300 includes two output voltage levels, i.e., common regulator outputs V₁ and V₂, and four spare regulator phases 227A, 227B, 227C and 227D. In embodiments, however, any number of voltage levels and number of spare regulator phases can be specified for a particular electronic system.

The MPC 122 of each phase group is electrically coupled to each dedicated regulator phase, e.g., 126A, 126B, of the phase group. The MPC 122 is configured to transfer, to control logic 366, phase fault signals and a shared current (I_SHARE) phase control signal received from each dedicated regulator phase of a respective phase group. The MPC 122 is also configured to generate I_SHARE control signals that can be used to sequentially activate each dedicated regulator phase of the phase group for predetermined periods of time, which can be useful in managing controlled current-sharing between phases within a phase group.

Each spare regulator phase 227A, 227B, 227C and 227D includes output switching devices 217 and 219 and ORing devices 118 and 118A, which can be particularly useful for interconnecting the spare regulator phases 227A, 227B, 227C and 227D to common regulator outputs V₁ and/or V₂. According to embodiments, output switching devices 217 and 219 can be activated, and thus can connect regulator output 242 of buck regulator 116 to primary output 204 or secondary output 206 of a spare regulator phase, e.g., 227A. This activation can occur in response to receiving a phase S1 enable V₁ signal 316A or S1 enable V₂ signal 318A, respectively, for spare regulator phase 227A, for example. Spare regulator phases 227B, 227C and 227D can be similarly activated by using corresponding enable signals. These phase enable signals are received by phase-redundant controller 106, and supplied to output switching devices 217 and 219, respectively. (See FIG. 2 for further details).

Control logic 366 is electrically connected to the MPCs 122 of each of the phase groups 374 and 376. The control logic 366 is configured to receive I_SHARE phase control signals from, and exchange phase fault signals with the MPCs 122. Control logic 366 is also electrically connected to the spare regulator phases 227A, 227B, 227C and 227D, and is configured to assert phase enable signals to, transfer I_SHARE phase control signals to, and receive phase fault signals from the spare regulator phases. For example, S1 enable V₂ signal 318A can be asserted by control logic 366 to enable the secondary output 206 (V₂ output) of spare regulator phase 227A. Control logic 366 can also transfer I_SHARE phase control signals, e.g., V₁ I_SHARE S1 signal 304A or V₁ I_SHARE S2 signal 306A, through connection 380 or connection 382, respectively, to S1 I_SHARE signal 314A. Control logic 366 can also receive phase fault signals, e.g., V₁ N−2 fault signal 310A, from a spare regulator phase.

Control logic 366 can be particularly useful in monitoring voltage regulator phases for faults, and in response to detecting a faulty regulator phase, enabling and disabling various phases and supplying activated phases with I_SHARE control signals. According to embodiments, I_SHARE control signals received from a phase group can be transferred by the control logic 366 to a spare regulator phase, which can be useful in causing the spare regulator phase to be driven at a level consistent with active regulator phases of the phase group. Additional examples of functions performed by control logic 366 are further detailed in FIG. 4 and the associated text.

In embodiments, control logic 366 can include a non-volatile memory 368, which can be used for storing associations between activated spare regulator phases and phase groups having one or more faulty regulator phases. According to embodiments, the control logic 366 can be a microcontroller, custom integrated circuit (IC), programmable logic device (PLD), application-specific integrated circuit (ASIC), or the like. Non-volatile memory 368 can be a Flash memory, electrically erasable programmable read-only memory (EEPROM), or other type of memory device that does not lose data when a supply voltage is removed.

Embodiments can be useful in reducing voltage regulator packaging size and cost, relative to a voltage regulator apparatus that uses regulator phases dedicated to a particular voltage output, while providing reliable, robust power delivery. These efficiencies are gained through the use of shared spare voltage regulator phases in place of redundant voltage regulator phases that are dedicated to a particular voltage level.

In some embodiments, a common regulator input V_IN, as depicted in FIG. 3, can be connected to V_IN inputs 136 of all of the phase groups 374 and 376, as well as the to V_IN inputs 136 of all of the spare regulator phases 227A, 227B, 227C and 227D. In some embodiments, different common regulator inputs can be connected to the V_IN inputs 136 and used to supply a unique voltage to each phase group and spare regulator phase. According to embodiments the V_OUT output 148 of each of the regulator phases 326A, 326B within a phase group, e.g., 374, are all connected to the same common regulator output, e.g., V₁. In embodiments, the primary output 204 and secondary output 206 of each of the spare regulator phases 227A, 227B, 227C and 227D can be selectively connected to either common regulator output V₁ or common regulator output V₂, in response to enable signals, e.g., S1 enable V₁ signal 316A or S1 enable V₂ signal 318A, asserted by control logic 366 on spare regulator phases 227A, 227B, 227C and 227D.

According to embodiments, a quantity of “N+1” or “N+2” voltage regulator phases can be electrically connected in parallel, where “N” is a minimum number of phases needed to supply a specified current, and the additional one or two phases can be useful in replacing one or two failed regulator phases. In the case of failure or “faulting” of one or more redundant phases, faulty redundant phase(s) can be disabled in order to share a current load and to ensure uninterrupted power delivery.

FIG. 3 depicts a phase-redundant voltage regulator having no current sharing circuitry between the voltage regulator phases 326A and 326B. Embodiments are contemplated, however, where current sharing is maintained between the voltage regulator phases 326A and 326B. In some such embodiments, independent current throttle points could be implemented within each regulator phase 326A and 326B, to enable a voltage regulator phase to run at its current output limit while one or more other voltage regulator phases provide the additional current required to satisfy a total current load. This embodiment can enable simplified current-sharing, which can allow certain voltage regulator phases to run at full current load, while additional voltage regulator phases can run at lower current levels. In some embodiments, passive or “droop” sharing can be implemented, where current sharing can result from an phase-redundant regulator output voltage drooping below a specified reference voltage, and, in response, the various regulator phases electrically connected to that output boost their respective current outputs. This can result in the sharing of the load between the regulators. In some embodiments, active or “forced” current sharing can be implemented through the use of added current monitoring, control, and feedback loops.

It can be appreciated by those of skill the art of voltage regulator and electronic system design that the operations described in reference to FIGS. 4-6 can, according to some embodiments, be performed in somewhat different orders than are depicted in the figures, and discussed in the associated text. For example, the order of asserting phase enable signals and sending I_SHARE phase control signals, may, in some embodiments, be reversed from the order depicted in FIG. 5 and FIG. 6, within the sprit and scope of the present disclosure. It can also be understood that the operations described in reference to FIGS. 4-6 can, in order to maintain stable power delivery and minimize voltage transients such as spikes and ripples, be performed either simultaneously, or in a very rapid sequence. For example, the time between one operation being completed and a next operation being completed may be 5 ms or less, in the practice of the present disclosure.

FIG. 4 is a flow diagram depicting a process 400 for transferring, within a phase-redundant voltage regulator apparatus, a spare regulator phase between phase groups of regulators, according to embodiments consistent with the figures. The phase-redundant voltage regulator apparatus includes a plurality of spare regulator phases, e.g., 227A, FIG. 3, and a plurality of dedicated regulator phases, e.g., 126A, FIG. 3. The phase-redundant voltage regulator apparatus is consistent with the phase-redundant voltage regulators and apparatus depicted in and described in reference to FIGS. 1-3. The process 400 is generally executed through the use of control logic 366, FIG. 3, which is electrically connected and configured to monitor phase fault signals received from the phase groups, e.g., phase groups 374 and 376, FIG. 3.

Certain operations of the process 400 performed by control logic 366, FIG. 3, may be initiated external to a voltage regulator apparatus by system control function 108, FIG. 1. For example, the initiation of a phase transfer in operation 404, FIG. 4, or other “higher level” operations such as monitoring current loads, throttling a system function, and the like, as detailed in FIGS. 5 and 6, may all be either initiated and/or performed by system control function 108, FIG. 1.

FIG. 4 can be particularly useful in depicting a process 400 of transferring a spare regulator phase between phase groups. This process 400 is further referenced as a single operation, in both FIG. 5 and FIG. 6 and the associated descriptions, thus simplifying the illustration and discussion of the methods depicted in both FIG. 5 and FIG. 6. It can be understood that block 402 can represent an entry or transition point from the methods depicted in FIG. 5 and FIG. 6.

The execution of process 400 can provide, through the use of shared redundant regulator phases controlled and allocated by control logic, enhanced reliability of power delivery to electronic systems. When used in conjunction with a phase-redundant voltage regulator apparatus of an electronic system, process 400 can also provide for substantial decreases in voltage regulator cost, packing area, design complexity, and failure rate, through the use of shared, assignable redundant regulator phases. Such improvements can result in corresponding overall decreases in electronic system cost, complexity, and failure rate. The execution of process 400 can also result in a rapid, seamless transfer/reassignment of a spare regulator phase from one phase group to another phase group. This seamless transfer can result in consistent power delivery for an electronic system that is free of interruptions and transients. The process depicted and described with reference to FIG. 4 is generally consistent with the phase-redundant voltage regulator and apparatus with shared redundant spares depicted in and described in reference to FIGS. 1-3. It can be understood that process 400 details a series of operations that, when executed, result in the electrical transfer of a voltage regulator phase from a first or “present” phase group to a second or “new” phase group. The completed results of the execution of this process may be herein referred to in FIG. 5 and FIG. 6 as a “phase transfer.”

It can be understood that block 402 can represent an entry or transition point from the methods depicted in FIG. 5 and FIG. 6. The process 400 moves from start 402 to operation 404.

Operation 404 generally refers to the initiation of a phase transfer. According to embodiments, a phase transfer can be initiated in response to a variety of signals and/or conditions. For example, a fault signal generated by a failed/faulted phase can be responded to by control logic 366, FIG. 3, which initiates the phase transfer using the I_SHARE phase control, phase enable, and phase fault signals. A phase transfer can also be initiated by a change in system configuration, i.e., a number of phases required on a particular voltage domain, or as a result of a change of current load resulting from the throttling of at least one throttled electronic device within a voltage domain. A change in system configuration or a change of current load may be detected by system control function 108 or by control logic 366, FIG. 3, which can initiate the phase transfer. Once the phase transfer has been initiated, the process 400 moves to operation 406.

Operation 406 generally refers to asserting a phase enable signal that electrically couples an output of an available spare regulator phase to a common regulator output of a “new” phase group. Prior to coupling the available spare regulator to the new phase group, if the available spare regulator is presently coupled to the output of another “present” phase group, according to embodiments, it is first disconnected or uncoupled from the present phase group by de-asserting the phase enable signal coupling it to the present phase group. Following the possible disconnection, the available spare regulator is then coupled, through the assertion of a respective enable signal, to the new phase group. This decoupling/coupling process results in the output of the available spare regulator being electrically connected to the outputs of active regulators in the new phase group at a common regulator output. By way of example, if the present phase group is 374, the new phase group is 376, and the available spare regulator is spare regulator phase 227C, then the respective enable signals are S3 enable V₁ signal 316C and S3 enable V₂ signal 318C. The respective common regulator outputs are V₁ and V₂. Once the phase enable signal is asserted, the process 400 moves to operation 408.

Operation 408 generally refers to transferring an I_SHARE phase control signal from the new phase group to the available spare regulator phase. Prior to transferring an I_SHARE phase control signal from the new phase group to the available spare regulator phase, if the I_SHARE phase control signal from the present phase group is being transferred to the available spare regulator phase, this connection must be first disabled by control logic 366. Once this connection is disabled, then a connection between the new phase group and the available spare regulator phase can be established, by control logic 366, for the I_SHARE control signal. This disconnection/connection process results in the available spare regulator phase receiving an I_SHARE control signal from the new phase group that it is connected to, which allows it to be driven similarly to other active phases within the new phase group.

Following the above example of the present phase group being 374, the new phase group being 376, and the available spare regulator being spare regulator phase 227C, then the I_SHARE control signal from the present phase group is V₁ I_SHARE S1 signal 304A, the I_SHARE control signal from the new phase group is V₂ I_SHARE S1 signal 304B, and the control logic 366 establishes a new connection between 304B and S3 I_SHARE signal 314C after removing any existing connection between signal 304A and signal 314C. Once the I_SHARE phase control signal is transferred from the new phase group to the available spare regulator phase, the process 400 moves to operation 410.

Operation 410 generally refers to electrically connecting a phase fault signal from the available spare regulator to the MPC of the new phase group. In this example, S3 fault signal 312C is electrically connected, by control logic 366, to V₂ fault in signal 302B of MPC 122 of phase group 376. This MPC 122 can monitor the assigned spare voltage regulator phase 227C for faults, and communicate a fault to control logic 366. Once the phase fault signal is connected, the process 400 moves to operation 412.

Operation 412 generally refers to storing an association between the first used spare regulator phase and the common regulator output of the first phase group, into a non-volatile memory, e.g., 368, FIG. 3, within the control logic. In this example, the association includes information re: the phase group, i.e., “phase group 374” or the common regulator output V₁, and the spare, i.e., spare regulator phase 227A. This information is stored in non-volatile memory 368, which can be useful in preserving the linkage of the spare regulator phase 227A to the common regulator output V₁, in the event of a power failure. Upon restart of the voltage regulator following a power failure, control logic 366 can re-initialize the electrical connection of spare regulator phase 227A to the common regulator output V₁. Once the association has been stored, the process 400 ends at block 414. It can be understood that block 414 can represent an exit or transition point to return to the methods depicted in FIG. 5 and FIG. 6.

FIG. 5 is a flow diagram 500 depicting a process for reallocating regulator phases between phase groups of regulators, according to embodiments consistent with the figures. The phase-redundant voltage regulator apparatus includes a plurality of spare regulator phases, e.g., 227A, FIG. 3, and a plurality of dedicated regulator phases, e.g., 126A, FIG. 3, and a plurality of phase groups, e.g., 374 and 376, FIG. 3. The phase-redundant voltage regulator apparatus is consistent with the phase-redundant voltage regulators and apparatus depicted in and described in reference to FIGS. 1-3. The process 500 is generally executed through the use of control logic 366, FIG. 3, which is electrically connected and configured to monitor phase fault signals. Control logic 366 is responsive to a system control function 108, FIG. 1, and to monitored phase fault signals received from the phase groups, e.g., phase groups 374 and 376, FIG. 3.

FIG. 5 depicts a process 500 for reallocating regulator phases between phase groups of regulators. It can be understood, however, that the process depicted in FIG. 5 may be expanded to include reallocating multiple regulator phases between multiple phase groups of regulators, in accordance with the spirit and scope of the present disclosure.

The operational state of the system at start 502 includes no voltage regulator phase faults, and both phase groups 374 and 376, FIG. 3, are running normally with “N” phases being used to supply current to loads attached to common regulator outputs V₁ and V₂, FIG. 3, respectively.

The process 500 moves from start 502 to decision 504. At operation 504, a decision is made by system control function 108 regarding whether fewer than a previously specified number “N” of voltage regulator phases are presently needed to supply power within a particular “evaluated” phase group. This decision can be made, for example, by system control function 108 comparing the number N to a number of phases calculated, according to the present current load on the evaluated phase group, that are presently needed to satisfy the current load of the evaluated phase group. If fewer than N voltage regulator phases are presently required to satisfy the current needs of the evaluated phase group, the process moves to operation 510. If N or greater voltage regulator phases are presently required to satisfy the current needs of the evaluated phase group, the process moves to operation 506.

At operation 506, a decision is made by system control function 108 regarding whether or not to throttle a device or subsystem within the electronic system in order to free up one or more spare phases which can be subsequently transferred or re-allocated to one or more phase groups other than the evaluated phase group.

“Throttling” of an electronic system and/or system component can be understood to refer to a process of reducing the performance of the system or component in order to accordingly reduce its power consumption. Throttling can be a particularly useful technique for overall electronic system power management. Throttling can be initiated or controlled by a system control function, e.g., 108, FIG. 1, which can actively monitor and manage/control power consumption of various components and subsystems of an electronic system.

For example, a system control function 108 can monitor the power consumption of one or more CPUs in a computer system or server, and subsequently apply controls to reduce the clock frequency of those CPU(s), which can thereby limit their current consumption. Such clock frequency reduction can be performed, for example, when processors or other system components are executing a relatively light and/or non-critical workload. A throttling action or process can also include the selective and judicious power-down of presently unused system components. Such components can include, but are not limited to, network and wireless interfaces, I/O ports and data storage devices.

In the context of the present disclosure, throttling can be applied to a component or subsystem, e.g., a processor, processor card or processor cluster, in order to reduce the power demands of that component or subsystem on the evaluated voltage regulator phase group supplying its power. This can be useful in freeing up a spare voltage regulator phase which can subsequently be used immediately in another phase group, or marked as “unallocated” for future allocation and use in another phase group, when a need arises. According to embodiments, the decision regarding whether or not to throttle the electronic system can be coded into program instructions executed by system control function 108 and/or may be derived by interaction of system control function 108 with a system user or administrator. If the decision is made to throttle a portion or all of the electronic system, the process moves to operation 508. If the decision is made not to throttle a portion or all of the electronic system, the process returns to operation 504.

Operation 508 generally refers to throttling a portion or all of the electronic system. As described above, throttling a portion or all of the electronic system can be initiated and managed by system control function 108. Once the association has been stored into the non-volatile memory, the process 500 returns to operation 504.

At operation 510, a decision is made by system control function 108 as to whether to allocate available spare phases to a phase group other than the evaluated phase group or not. According to embodiments, the decision regarding whether or not to allocate available spare phases can be coded into program instructions executed by system control function 108 and/or may be derived by interaction of system control function 108 with a system user or administrator.

According to embodiments, at least one spare voltage regulator phase can be reallocated in response to commands received from a system control function 108. The set of spare voltage regulator phases can include voltage regulator phases that were designated as spare voltage regulator phases in response to a system throttle operation. According to embodiments, the set of spare voltage regulator phases can also include voltage regulator phases in excess of a number of voltage regulator phases specified as required for a voltage regulator phase. If the decision is not to allocate the available spare phases, the process moves to operation 512. If the decision is to allocate the available spare phases, the process moves to operation 514.

Operation 512 generally refers to “marking” spare phases as “allocated” for use in the evaluated phase group. It can be understood that such marking of spare phases can include storing, into a non-volatile memory 368 within the control logic 366, an association between a first portion of the set of spare voltage regulator phases and an “allocated” status. This marking of spare phases can be particularly useful in reserving these regulator phases for use with the evaluated phase group. Once the association has been stored into the non-volatile memory, the process 500 returns to operation 504.

Operation 514 generally refers to “marking” spare phases as “unallocated” for use in the evaluated phase group. It can be understood that such marking of spare phases can include storing, into a non-volatile memory 368 within the control logic 366, an association between a second portion of the set of spare voltage regulator phases and an “unallocated” status. This marking of spare phases can be particularly useful in allowing these regulator phases to be used with other phase groups, following, for example, the failure of one or more voltage regulator phases within that phase group. Once the association has been stored into the non-volatile memory, the process 500 moves to operation 516.

At operation 516, a detection and decision is made regarding whether a failure/fault is present in a phase group other than the evaluated phase group. As described above in reference to FIG. 3, control logic 366 can monitor and detect phase faults from the various dedicated and spare phases within a voltage regulator apparatus. According to embodiments, a phase fault signal from a first compromised phase group, i.e., phase group having a failed or faulted voltage regulator phase, can be a phase single-fault signal, a phase double-fault signal, or a spare phase fault signal. If a phase fault signal is detected, the process moves to operation 400. If a phase fault signal is not detected, the process returns to operation 516.

Operation 400 generally refers to transferring, in response to detecting the phase fault signal, at least one spare voltage regulator phase of the second portion of the set of spare voltage regulator phases to the first compromised phase group. The details of operation 400 are further depicted in FIG. 4 and discussed in the associated text. In some embodiments, the transferring can also include transferring at least one spare voltage regulator phase to a second compromised phase group. In embodiments, the transferring can include multiple independent phase transfers to various phase groups within an electronic system. Once the at least one spare voltage regulator phase has been transferred, the process 500 returns to operation 502.

It can be understood that the above process can be particularly useful in maintaining redundancy within a voltage regulator apparatus having multiple phase groups, without having to employ a dedicated set of spare voltage regulator phases or each phase group. Embodiments of the present disclosure can enable efficient use and reallocation of spare regulator phases to multiple phase groups within a voltage regulator apparatus. Other similar methods, within the spirit and scope of the present disclosure, may be used for a voltage regulator apparatus having a various numbers of available spare voltage regulators.

FIG. 6 is a flow diagram 600 depicting a process, consistent with the process, as depicted in FIG. 5 and discussed in the associated text, for reallocating regulator phases between phase groups of regulators, according to embodiments consistent with the figures. The phase-redundant voltage regulator apparatus is consistent with the phase-redundant voltage regulators and apparatus depicted in and described in reference to FIGS. 1-3. The process 600 is generally executed through the use of control logic 366, FIG. 3, which is electrically connected and configured to monitor phase fault signals.

FIG. 6 depicts a process 600 for reallocating regulator phases between phase groups of regulators. It can be understood, however, that the process depicted in FIG. 6 may be expanded to include reallocating multiple regulator phases between multiple phase groups of regulators, in accordance with the spirit and scope of the present disclosure.

The operational state of the system at start 602 includes no voltage regulator phase faults, and both phase groups 374 and 376, FIG. 3, are running normally with “N” phases being used to supply current to loads attached to common regulator outputs V₁ and V₂, FIG. 3, respectively.

The process 600 moves from start 602 to decision 604. At operation 604, a decision is made by system control function 108 regarding whether fewer than a previously specified number “N” of voltage regulator phases are presently needed to supply power within a particular “evaluated” phase group. This decision can be made, for example, by system control function 108 comparing the number N to a number of phases calculated, according to the present current load on the evaluated phase group, that are presently needed to satisfy the current load of the evaluated phase group. If fewer than N voltage regulator phases are presently required to satisfy the current needs of the evaluated phase group, the process moves to operation 610. If N or greater voltage regulator phases are presently required to satisfy the current needs of the evaluated phase group, the process moves to operation 606.

At operation 606, a decision is made by system control function 108 regarding whether or not to throttle a device or subsystem within the electronic system in order to free up one or more spare phases which can be subsequently transferred or re-allocated to one or more phase groups other than the evaluated phase group.

In the context of the present disclosure, throttling can be applied to a component or subsystem, e.g., a processor, processor card or processor cluster, in order to reduce the power demands of that component or subsystem on the evaluated voltage regulator phase group supplying its power. This can be useful in freeing up a spare voltage regulator phase which can subsequently be used immediately in another phase group, or marked as “unallocated” for future allocation and use in another phase group, when a need arises. According to embodiments, the decision regarding whether or not to throttle the electronic system can be coded into program instructions executed by system control function 108 and/or may be derived by interaction of system control function 108 with a system user or administrator. If the decision is made to throttle a portion or all of the electronic system, the process moves to operation 608. If the decision is made not to throttle a portion or all of the electronic system, the process returns to operation 604.

Operation 608 generally refers to throttling a portion or all of the electronic system. As described above, throttling a portion or all of the electronic system can be initiated and managed by system control function 108. Once the association has been stored into the non-volatile memory, the process 600 returns to operation 604.

At operation 610, a decision is made by system control function 108 as to whether to allocate available spare phases to a phase group other than the evaluated phase group or not. According to embodiments, the decision regarding whether or not to allocate available spare phases can be coded into program instructions executed by system control function 108 and/or may be derived by interaction of system control function 108 with a system user or administrator.

According to embodiments, at least one spare voltage regulator phase can be reallocated in response to commands received from a system control function 108. The set of spare voltage regulator phases can include voltage regulator phases that were designated as spare voltage regulator phases in response to a system throttle operation. According to embodiments, the set of spare voltage regulator phases can also include voltage regulator phases in excess of a number of voltage regulator phases specified as required for a voltage regulator phase. If the decision is not to allocate the available spare phases, the process moves to operation 612. If the decision is to allocate the available spare phases, the process moves to operation 614.

Operation 612 generally refers to “marking” spare phases as “allocated” for use in the evaluated phase group. It can be understood that such marking of spare phases can include storing, into a non-volatile memory 368 within the control logic 366, an association between a first portion of the set of spare voltage regulator phases and an “allocated” status. This marking of spare phases can be particularly useful in reserving these regulator phases for use with the evaluated phase group. Once the association has been stored into the non-volatile memory, the process 600 returns to operation 604.

Operation 614 generally refers to “marking” spare phases as “unallocated” for use in the evaluated phase group. It can be understood that such marking of spare phases can include storing, into a non-volatile memory 368 within the control logic 366, an association between a second portion of the set of spare voltage regulator phases and an “unallocated” status. This marking of spare phases can be particularly useful in allowing these regulator phases to be used with other phase groups, following, for example, the failure of one or more voltage regulator phases within that phase group. Once the association has been stored into the non-volatile memory, the process 600 moves to operation 616.

At operation 616, an analysis is performed by system control function 108 that compares voltage regulator load versus total current capacity of voltage regulators within the various phase groups within the voltage regulator apparatus. Based on the results of this comparison, one or more phase groups can be identified as having a relatively small load versus capacity discrepancy, relative to other phase groups. In embodiments, these phase groups can be designated as phase groups to receive the allocation of one or more available spare regulator phases. Once these these phase groups have been identified, the process 600 moves to operation 400.

Operation 400 generally refers to transferring at least one spare voltage regulator phase of the second portion of the set of spare voltage regulator phases to the various phase groups designated in operation 616. The details of operation 400 are further depicted in FIG. 4 and discussed in the associated text. In some embodiments, the transferring can also include transferring at least one spare voltage regulator phase to a second compromised phase group. In embodiments, the transferring can include multiple independent phase transfers to various phase groups within an electronic system. Once the at least one spare voltage regulator phase has been transferred, the process 600 returns to operation 602.

It can be understood that the above process can be particularly useful in maintaining redundancy within a voltage regulator apparatus having multiple phase groups, without having to employ a dedicated set of spare voltage regulator phases or each phase group. Embodiments of the present disclosure can enable efficient use and reallocation of spare regulator phases to multiple phase groups within a voltage regulator apparatus. Other similar methods, within the spirit and scope of the present disclosure, may be used for a voltage regulator apparatus having a various numbers of available spare voltage regulators.

According to embodiments, a current delivery capacity of each regulator phase within a phase group can be specified so as to result in the delivery of a specified cumulative output current of the phase group following a failure of one regulator phase within a phase group. In some embodiments, the per-phase current delivery capacity of each regulator phase within a phase group can allow for delivery of a specified cumulative output current of the phase group following a failure of at least two regulator phases within a phase group.

The descriptions of the various embodiments of the present disclosure have been presented for purposes of illustration, but are not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The terminology used herein was chosen to explain the principles of the embodiments, the practical application or technical improvement over technologies found in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A phase-redundant voltage regulator apparatus comprising: a plurality of regulator phases, each including a regulator electrically coupled to receive, at a regulator input, an input voltage and provide, at a regulator output, a respective output voltage;a set of phase groups of the plurality of regulator phases including a first and a second phase group, each phase group including; a common regulator input electrically interconnected to regulator inputs of regulators of the phase group;a common regulator output electrically interconnected to regulator outputs of regulators of the phase group;at least one redundant regulator phase;at least one dedicated regulator phase of the plurality of regulator phases;at least one spare regulator phase of a set of spare regulator phases;a multi-phase controller (MPC) electrically coupled to each dedicated regulator phase of the phase group, the MPC configured to transfer, to control logic, phase fault signals and a shared current (ISHARE) phase control signal received from each dedicated regulator phase of a respective phase group; andthe set of spare regulator phases of the plurality of regulator phases including a first and a second spare regulator phase, each spare regulator phase including: a secondary output ORing device electrically coupled and configured to limit current flow into a secondary output of the spare regulator phase;a first output switching device configured to, in response to a first phase enable signal, electrically couple the regulator output of the spare regulator phase to a first common regulator output; anda second output switching device configured to, in response to a second phase enable signal, electrically couple the regulator output of the spare regulator phase to a second common regulator output; andcontrol logic, electrically connected to: MPCs of each phase group of the set of phase groups, the control logic configured to receive ISHARE phase control signals from, and exchange phase fault signals with the MPCs; andspare regulator phases of the set of spare regulator phases, the control logic configured to assert phase enable signals to, transfer ISHARE phase control signals to, and receive phase fault signals from the spare regulator phases;the control logic configured to, in response to receiving a phase fault signal from an MPC, electrically interconnect a spare regulator phase to a phase group that includes a failed regulator phase.

2. The phase-redundant voltage regulator apparatus of claim 1, each dedicated regulator phase of the plurality of regulator phases further including: a phase-redundant controller (PRC) configured to monitor current at the regulator input and further configured to monitor current and voltage at the regulator output;an output ORing device configured to limit current flow into a primary output of a respective dedicated regulator phase; andan input protection device configured to provide, in response to a control signal from the PRC, input overcurrent protection and output overvoltage protection to the respective dedicated regulator phase.

3. The phase-redundant voltage regulator apparatus of claim 1, wherein the MPC of each phase group is further configured to: receive a feedback output voltage and receive a respective detected current signal from each dedicated regulator phase of the phase group;generate ISHARE control signals to sequentially activate each dedicated regulator phase of the phase group for predetermined periods of time, the ISHARE control signals managing controlled current-sharing between phases; andmaintain, following a failure of one or more regulator phase of the phase group, current-sharing between all active regulator phases of the phase group.

4. The phase-redundant voltage regulator apparatus of claim 1, wherein each ISHARE control signal is an analog signal that represents, through a voltage, a required current output level of at least one regulator phase.

5. The phase-redundant voltage regulator apparatus of claim 1, wherein the output ORing device and the secondary output ORing device are each selected from the group consisting of: an N-channel field-effect transistor (NFET), a P-channel field-effect transistor (PFET), an NPN transistor, and a PNP transistor.

6. The phase-redundant voltage regulator apparatus of claim 1, wherein a regulator serial interface of an MPC is coupled to a system control function through a serial control bus selected from the group consisting of: an Serial Peripheral Interface (SPI) interface, a Power Management Bus (PMBus) interface, and an Inter-Integrated Circuit (I2C) interface.

7. The phase-redundant voltage regulator apparatus of claim 1, wherein the first phase group of the set of phase groups is configured to maintain current sharing.

8. A method for reallocating a set of spare voltage regulator phases between phase groups of voltage regulator phases, the method comprising using control logic that is responsive to a system control function and responsive to monitored phase fault signals received from the phase groups to: store, into a non-volatile memory within the control logic, an association between a first portion of the set of spare voltage regulator phases and an “allocated” status;store, with the control logic, into the non-volatile memory within the control logic, an association between a second portion of the set of spare voltage regulator phases and an “unallocated” status;detect, with the control logic, a phase fault signal from a first compromised phase group of the phase groups; andtransfer, in response to detecting the phase fault signal, at least one spare voltage regulator phase of the second portion of the set of spare voltage regulator phases to the first compromised phase group.

9. The method of claim 8, wherein the phase fault signal is selected from the group consisting of: a phase single-fault signal, a phase double-fault signal, and a spare phase fault signal.

10. The method of claim 8, wherein the at least one spare voltage regulator phase is reallocated in response to commands received from a system control function.

11. The method of claim 8, wherein the transfer of the at least one spare voltage regulator phase to the first compromised phase group includes transferring at least one spare voltage regulator phase to a second compromised phase group.

12. The method of claim 8, wherein the set of spare voltage regulator phases includes voltage regulator phases that were designated as spare voltage regulator phases in response to a system throttle operation.

13. The method of claim 8, wherein the set of spare voltage regulator phases includes voltage regulator phases designated as spare voltage regulator phases in response to a system throttle operation.

14. The method of claim 8, wherein the set of spare voltage regulator phases includes voltage regulator phases in excess of a number of voltage regulator phases specified as required for a voltage regulator phase.

15. A method for reallocating a set of spare voltage regulator phases between phase groups of voltage regulator phases, the method comprising using control logic that is responsive to a system control function and responsive to monitored phase fault signals received from the phase groups to: store, into a non-volatile memory within the control logic, an association between a first portion of the set of spare voltage regulator phases and an “allocated” status;store, with the control logic, into the non-volatile memory within the control logic, an association between a second portion of the set of spare voltage regulator phases and an “unallocated” status;detect, with the control logic, a phase fault signal from a first compromised phase group of the phase groups; andtransfer, in response to detecting the phase fault signal, at least one spare voltage regulator phase of the second portion of the set of spare voltage regulator phases to the first compromised phase group.

16. The method of claim 15, wherein the phase fault signal is selected from the group consisting of: a phase single-fault signal, a phase double-fault signal, and a spare phase fault signal.

17. The method of claim 15, wherein the at least one spare voltage regulator phase is reallocated in response to commands received from a system control function.

18. The method of claim 15, wherein the transfer of the at least one spare voltage regulator phase to the first compromised phase group includes transferring at least one spare voltage regulator phase to a second compromised phase group.

19. The method of claim 15, wherein the set of spare voltage regulator phases includes voltage regulator phases that were designated as spare voltage regulator phases in response to a system throttle operation.

20. The method of claim 15, wherein the set of spare voltage regulator phases includes voltage regulator phases designated as spare voltage regulator phases in response to a system throttle operation.