WO2016087002A1

WO2016087002A1 - Voltage regulator mechanism, electronic device, method and computer program

Info

Publication number: WO2016087002A1
Application number: PCT/EP2014/076684
Authority: WO
Inventors: Rickard EVERTSSON; Benny JÖNSSON; Daniel Nielsen; Tomas Johansson
Original assignee: Telefonaktiebolaget Lm Ericsson (Publ)
Priority date: 2014-12-05
Filing date: 2014-12-05
Publication date: 2016-06-09

Abstract

A voltage regulator mechanism arranged to adapt voltage level for supply to an entity of an electronic apparatus is disclosed. The voltage regulator mechanism comprises a controller arranged to, at each reboot of the entity, determine whether the reboot is due to an instability condition, and if so the controller registers an instability condition and updates an adaptive voltage scaling based on registered instability conditions. An electronic device, such as a communication device, comprising such a voltage regulator mechanism, a method of adapting a voltage level for supply to an entity of an electronic apparatus, and a computer program for implementing the method are also disclosed.

Description

VOLTAGE REGULATOR MECHANISM, ELECTRONIC DEVICE, METHOD AND COMPUTER PROGRAM

Technical field

The present invention generally relates to a voltage regulator mechanism, an electronic device comprising such a voltage regulator mechanism, a method of adapting a voltage level for supply to an entity of an electronic apparatus, and a computer program for implementing the method. Background

Embedded systems comprise different electronic components. Typically, they consist of one or many central processing unit, CPU, cores, memories, and/or complex application specific integrated circuits, ASICs, used to specific functionalities. There may also be many regulators used to power each of these parts.

Each regulator is configured to provide accurate and appropriate voltage levels in order to power the connected devices in its different modes. If the connected devices require high performance the voltages are increased and when the devices require low performance the regulators may be decreased. These modes are used to boost performance when needed and to save power when not needed.

When a device, such as an electronic communication device or element, is mass produced each device or element may be different due to silicon process variation. Due to this process variation some devices may operate at lower base voltage than others. To handle these differences, the regulator for each device may be tuned according to AVS (Adaptive Voltage Scaling) technology. The technology is sometimes also referred to as Dynamic Voltage Scaling. The AVS (Adaptive Voltage Scaling) technology could also take temperature, and the life time of a device into account. In the latter case it may adjust the voltage levels as the device gets older, e.g. based on run time, according to a scheduled AVS level.

The AVS setting trims voltage levels for devices but also retains a margin large enough to cover all kind of known liabilities, such as IR-drop (current-resistance-drop), regulator inaccuracies, board layouts, temperatures, silicon process variation, ageing etc. Apart from this, the margin is traditionally also assigned to be large enough to handle some unknown liabilities, such as external components and insufficient board design.

The above AVS Settings are traditionally assigned to cover a device's complete life cycle. However, the margin can somewhat be reduced if e.g. temperature and ageing margins are handled separately. Power consumption may be an important figure, especially for products powered by batteries to maximize time between recharge.

Traditionally, a conservative approach was taken to cover every device, e.g. ASIC instance, in every foreseeable but at the same time it is desired to apply a low voltage to keep the power consumption low. This is complex and despite the conservative approach it may still cause instability due to power integrity.

Power integrity issues are difficult and time consuming to investigate since they often do not appear to have a rational explanation causing bit flips, deadlocks and unexplained watchdog resets. Devices experiencing this may be so deadlocked that they cannot even wake up from a system interrupts intended to override other interrupts.

If a device is a rare corner case sample where the certain circumstances together with the AVS Settings causes instability, there is traditionally no way to recover from this. The same AVS Settings are applied at each boot and the device may continue to experience instability. It is therefore a desire to provide a solution providing a balance between power consumption and avoiding such instability.

Summary

The invention is based on the understanding that adapting the margin for AVS based on occurred instability events may provide for lower voltage and thus lower power consumption for most individual electronic devices.

According to a first aspect, there is provided a voltage regulator mechanism arranged to adapt voltage level for supply to an entity of an electronic apparatus. The voltage regulator mechanism comprises a controller arranged to, at each reboot of the entity, determine whether the reboot is due to an instability condition, and if so the controller registers an instability condition and updates an adaptive voltage scaling based on registered instability conditions.

A reboot may be considered to be due to an instability condition for each reboot where no user or system instruction caused the reboot in a controlled way. A user instruction causing reboot in a controlled way may include when a user has turned off the electronic apparatus and then turned on the electronic apparatus. A system instruction causing the reboot in a controlled way may include action due to a software update requiring reboot or another planned software induced reboot.

The controller may be arranged to check a stored log about the reboot against a database for the determination whether the reboot is due to an instability condition. The controller may be further arranged to update the adaptive voltage scaling based on any one of lifetime of the entity and temperature of the entity. The controller may be arranged to base the update of the adaptive voltage scaling on the determination whether the reboot is due to an instability condition and the any one of lifetime of the entity and temperature of the entity, where the any one of lifetime of the entity and temperature of the entity is used as an upper bound for increasing the adaptive voltage scaling.

The controller may be arranged to increase the adaptive voltage scaling when the number of registered instability conditions reaches a predetermined threshold.

The controller may be arranged to register the instability condition in a nonvolatile memory.

According to a second aspect, there is provided an electronic device comprising a voltage regulator mechanism according to the first aspect.

The electronic device may be a communication device arranged for cellular communication comprising a transceiver, a processor, input and output interfaces and the voltage regulator mechanism.

According to a third aspect, there is provided a method of adapting a voltage level for supply to an entity of an electronic apparatus. The method comprises determining at each reboot of the entity whether the reboot is due to an instability condition, and if the reboot is due to an instability condition, registering the instability condition; and updating an adaptive voltage scaling based on registered instability conditions.

A reboot may be considered to be due to an instability condition for each reboot where no user or system instruction caused the reboot in a controlled way. A user instruction causing reboot in a controlled way may include when a user has turned off the electronic apparatus and then turned on the electronic apparatus. A system instruction causing the reboot in a controlled way may include action due to a software update requiring reboot or other planned software induced reboot.

The determining whether the reboot is due to an instability condition may comprise checking a stored log about the reboot against a database.

The method may comprise updating the adaptive voltage scaling based on any one of lifetime of the entity and temperature of the entity. The method may comprise basing the update of the adaptive voltage scaling on the determination whether the reboot is due to an instability condition and the any one of lifetime of the entity and temperature of the entity, where the any one of lifetime of the entity and temperature of the entity is used as an upper bound for increasing the adaptive voltage scaling.

The method may comprise increasing the adaptive voltage scaling when the number of registered reboots due to an instability condition reaches a predetermined threshold.

The registering the instability condition may comprise registering the instability condition in a non-volatile memory.

According to a fourth aspect, there is provided a computer program comprising instructions which, when executed on a processor of an electronic apparatus, causes the electronic apparatus to perform the method according to the third aspect.

Brief description of the drawings

The above, as well as additional objects, features and advantages of the present invention, will be better understood through the following illustrative and non-limiting detailed description of preferred embodiments of the present invention, with reference to the appended drawings.

Fig. 1 schematically illustrates an arrangement with a power source, an entity to be powered and a voltage regulator mechanism according to an embodiment.

Fig. 2 is a block diagram schematically illustrating a communication device according to an embodiment.

Fig. 3 is a diagram illustrating voltage versus lifetime of an entity powered by a power source and principles of adaptive voltage scaling.

Fig. 4 is a diagram illustrating voltage versus lifetime of an entity powered by a power source and principles of adaptive voltage scaling.

Fig. 5 is a flow chart illustrating a method according to an embodiment.

Fig. 6 is a flow chart illustrating a method according to an embodiment.

Fig. 7 schematically illustrates a computer-readable medium and a processing device. Detailed description

This proposal gives the possibility not only to tune AVS (Adaptive Voltage Scaling) settings based on known liabilities but also on experienced instability issues as well. The instability may be caused by a number of reasons including both corner cases of known liabilities and unknown liabilities which are hard to predict. If the voltage levels are too low for a specific device, it is likely that this device will experience power related instability issues and at some point of time crash and reboot due to a freeze / deadlock. Each fatal error such as a watchdog timeout or another abnormal reset may be monitored and noted in a persistent storage memory, e.g. a non- volatile memory. The AVS Settings will be adjusted based on these instabilities and AVS margins may be increased until a limit has been reached, as will be further elucidated below.

At each boot the number of fatal errors and abnormal reboots may be checked and used as input when configuring adjusted voltage levels for the regulators.

Fig. 1 schematically illustrates an arrangement with a power source 102, an entity 104 to be powered and a voltage regulator mechanism 100 according to an embodiment. The voltage regulator mechanism 100 comprises circuitry 106 for inputting power from the power source 102 and outputting power at a regulated voltage to the entity 104. The entity 104 may for example be a processor or controller comprised in an electronic device, e.g. a communication device for cellular communication such as a small base station or mobile terminal, where power consumption issues are important. The circuit 106 is controlled by an AVS manager 108, which is a controller or processor providing control signals to the circuit 106. The AVS manager 108 is arranged to receive information from the entity 104 about reboots and associated data, such as data about watchdog timeout or another abnormal reset. The voltage regulator mechanism may comprise a database 110 holding information about instability conditions mapped on data that may be received from the entity 104. The AVS manager 108 may operate according to what is demonstrated with reference to Figs 5 and 6 below.

A voltage regulator mechanism 100 as demonstrated above may be used in any electronic device or apparatus having elements or entities that are powered using AVS. Such electronic apparatuses may for example be computers, media players, digital cameras, and communication devices. The communication devices may for example be arranged for cellular communication, point-to-point communication or arranged for broadcasting (transmitters or receivers). An example of an electronic apparatus will be given with reference to Fig. 2.

Fig. 2 is a block diagram schematically illustrating a communication device 200 according to an embodiment. The communication device comprises an antenna arrangement 202, a receiver 204 connected to the antenna arrangement 202, a transmitter 206 connected to the antenna arrangement 202, a processing element 208 which may comprise one or more circuits, one or more input interfaces 210 and one or more output interfaces 212. The interfaces 210, 212 can be user interfaces and/or signal interfaces, e.g. electrical or optical. The communication device 200 may be arranged to operate in a cellular communication network, and may for example be a base station or a mobile station. The processing element 208 can fulfill a multitude of tasks, ranging from signal processing to enable reception and transmission since it is connected to the receiver 204 and transmitter 206, executing applications, controlling the interfaces 210, 212, etc. The processing element 208 is here illustrated as a central processing resource for the communication device 200, but in the context of the invention should be considered to represent processing elements in any of the elements of the communication device 200. Thus, a voltage regulator mechanism, VRM, 214, e.g. as the one demonstrated with reference to Fig. 1, is provided to one or more of these processing elements for providing a voltage level according to the approaches demonstrated herein.

Fig. 3 is a diagram illustrating principles of adaptive voltage scaling. The diagram shows voltage versus lifetime of an entity powered by a power source. The thin line 300 illustrates the voltage requirement for normal operation, which is not known for each individual entity, of e.g. a produced batch, but some properties are known such as average initial voltage requirement for a batch, average ageing behavior, etc. Traditionally, an AVS margin is set such that enough voltage is provided to at least a majority of individuals of the entity, which may be set as the dashed line 302. The generous margin in most cases leads to unnecessary power consumption. Here, it is instead suggested to apply a leaner margin, as illustrated by the solid line 304. When the applied voltage level according to the solid line 304 meets and/or is below the required voltage as of the thin line 300, one or more instability events may occur. The approach is then to, based on the occurred instability events, increase the applied voltage such that a voltage margin, still relatively lean compared with the traditional margin as shown by the dashed line 302, again is established. This is kept until further instability events occur, wherein voltage is increased again. Here, it can also be seen that some log about occurred instability events may be kept, wherein if new instability events occur rather soon after a voltage increase, as intended to be illustrated by the rather close in time groups of instability events to the right, the voltage increase may be assigned based on this, and, at the last voltage increase shown in the diagram is intended to illustrate, be a larger increase than an applied voltage increase, e.g. as of the first and second voltage increases illustrated in the diagram. Fig. 4 is a diagram illustrating principles of adaptive voltage scaling. As in Fig. 3, the thin line 400 illustrates the voltage requirement for proper operation, which is not known for each individual entity, the dashed line 402 illustrates an applied voltage according to a traditionally applied voltage margin, and the solid line 404 illustrates applied voltage according to the suggested approach. The first, second and third voltage increases are similar to those discussed with reference to Fig. 3, and the topic of this diagram is the last voltage increase. Here, some instability events occur, wherein a voltage increase is applied. However, here the voltage increase is bounded by the traditional AVS margin scheme, wherein the voltage is set at maximum to the voltage level as indicated by the dashed line 402. That is, the suggested approach may be bounded by the traditional AVS scheme such that power consumption never exceeds the one of the traditional AVS scheme. This is also a safeguard for preventing unlimited voltage increase in case instability events occur frequently, and may be misinterpreted as caused by insufficient voltage margin. At worst, the voltage level then never becomes higher than for the traditional AVS scheme. In most cases, a significant amount of power is saved since the voltage level can be kept lower than for the traditional AVS scheme.

Fig. 5 is a flow chart illustrating a method according to an embodiment. The method assumes a voltage regulator mechanism arranged to adapt voltage level for supply to an entity of an electronic apparatus. The voltage regulator mechanism comprises a controller arranged to, at each reboot of the entity, perform the method. The method includes to determine 500 whether an adaptive voltage scaling margin should be updated, e.g. at reboot if the reboot is determined to be caused by an instability event that may be caused by insufficient voltage. The method then includes applying 502 an adaptive voltage scaling accordingly. Thus, the determination 500 comprises determining whether the reboot is due to an instability condition, and if the reboot is due to an instability condition the controller registers the reboot is due to an instability condition and updates of the adaptive voltage scaling by an increase based on registered reboots due to an instability condition.

Fig. 6 is a flow chart illustrating a method according to an embodiment. The method comprises to check 600 whether a reboot has occurred. If the reboot has occurred, it is checked 602 whether an instability event has occurred, and whether this instability event is likely to have occurred due to insufficient voltage. Here, it can be noted that this check 602 may alternatively be performed prior to the reboot as well, e.g. instantaneously as the instability event occurs, or be part of a process induced by the instability event. For example, the process may include an occurrence of an unexpected reboot, wherein an error handler provides an event log, and a further reboot is performed for putting the entity working again, wherein the checking 602 may be considered to include both the providing of the event log and checking the event log as demonstrated below.

The check 602 may be performed by checking an event log against a database, where the database comprises collected knowledge about indicators on instability upon insufficient voltage. If it is found that the instability event is likely due to insufficient voltage, the reboot is registered 604 as an instability event. Optionally, it is checked 605 whether a predetermined number of instability events have occurred. If not, the method returns to check for new reboots. If the predetermined number of instability events have occurred, or alternatively without this check 605, the method proceeds with assigning 606 AVS voltage with an increased voltage. Optionally, it is checked 607 whether the assigned AVS voltage exceeds a scheduled level, i.e. corresponding to traditional AVS margin as discussed with reference to Fig. 4. If the assigned AVS voltage exceeds the scheduled level, the scheduled AVS voltage is applied 609. If the assigned AVS voltage does not exceed the scheduled level, or alternatively without the check 607, the assigned AVS voltage is applied 608.

The methods according to the present invention are suitable for implementation with aid of processing means, such as computers and/or processors, especially for the case where the AVS manager or the processing element 208 demonstrated above comprises a processor handling AVS. Therefore, there is provided computer programs, comprising instructions arranged to cause the processing means, processor, or computer to perform the steps of any of the methods according to any of the embodiments described with reference to Figs 3 to 6. The computer programs preferably comprise program code which is stored on a computer readable medium 700, as illustrated in Fig. 7, which can be loaded and executed by a processing means, processor, or computer 702 to cause it to perform the methods, respectively, according to embodiments of the present invention, preferably as any of the embodiments described with reference to Figs 3 to 6. The implementation of the above demonstrated approaches is particularly suitable for software implementation for several reasons. One reason is that update of the mechanism for determining whether an instability event is likely to be because of shortage of voltage is facilitated by e.g. a software update. Another reason is that checking of event logs against a database of likely events is a task that is particularly suitable for performing by processing means executing software for e.g. parsing the event log and checking a database. The computer 702 and computer program product 700 can be arranged to execute the program code sequentially where actions of the any of the methods are performed stepwise. The processing means, processor, or computer 702 is preferably what normally is referred to as an embedded system. Thus, the depicted computer readable medium 700 and computer 702 in Fig. 7 should be construed to be for illustrative purposes only to provide understanding of the principle, and not to be construed as any direct illustration of the elements.

Claims

1. A voltage regulator mechanism arranged to adapt voltage level for supply to an entity of an electronic apparatus, wherein the voltage regulator mechanism comprises a controller arranged to, at each reboot of the entity, determine whether the reboot is due to an instability condition, and if so the controller registers an instability condition and updates an adaptive voltage scaling based on registered instability conditions.

2. The voltage regulator mechanism of claim 1, wherein a reboot is considered to be due to an instability condition for each reboot where no user or system instruction caused the reboot in a controlled way.

3. The voltage regulator mechanism of claim 2, where a user instruction causing reboot in a controlled way includes when a user has turned off the electronic apparatus and then turned on the electronic apparatus.

4. The voltage regulator mechanism of claim 2 or 3, where a system instruction causing the reboot in a controlled way includes action due to a software update requiring reboot or another planned software induced reboot.

5. The voltage regulator mechanism of any one of claims 1 to 4, where the controller is arranged to check a stored log about the reboot against a database for the determination whether the reboot is due to an instability condition.

6. The voltage regulator mechanism of any one of claims 1 to 5, wherein the controller is further arranged to update the adaptive voltage scaling based on any one of lifetime of the entity and temperature of the entity.

7. The voltage regulator mechanism of claim 6, wherein the controller is arranged to base the update of the adaptive voltage scaling on the determination whether the reboot is due to an instability condition and the any one of lifetime of the entity and temperature of the entity, where the any one of lifetime of the entity and temperature of the entity is used as an upper bound for increasing the adaptive voltage scaling.

8. The voltage regulator mechanism of any one of claims 1 to 7, wherein the controller is arranged to increase the adaptive voltage scaling when the number of registered instability conditions reaches a predetermined threshold.

9. The voltage regulator mechanism of any one of claims 1 to 8, wherein the controller is arranged to register the instability condition in a non- volatile memory.

10. An electronic device comprising a voltage regulator mechanism according to any one of claims 1 to 9.

11. The electronic device of claim 10, being a communication device arranged for cellular communication comprising a transceiver, a processor, input and output interfaces and the voltage regulator.

12. A method of adapting a voltage level for supply to an entity of an electronic apparatus, the method comprising

determining at each reboot of the entity whether the reboot is due to an instability condition, and

if the reboot is due to an instability condition, registering the instability condition; and

updating an adaptive voltage scaling based on registered instability conditions.

13. The method of claim 12, wherein a reboot is considered to be due to an instability condition for each reboot where no user or system instruction caused the reboot in a controlled way.

14. The method of claim 13, where a user instruction causing reboot in a controlled way includes when a user has turned off the electronic apparatus and then turned on the electronic apparatus.

15. The method of claim 13 or 14, where a system instruction causing the reboot in a controlled way includes action due to a software update requiring reboot or other planned software induced reboot.

16. The method of any one of claims 12 to 15, wherein the determining whether the reboot is due to an instability condition comprises checking a stored log about the reboot against a database.

17. The method of any one of claims 12 to 16, comprising updating the adaptive voltage scaling based on any one of lifetime of the entity and temperature of the entity.

18. The method of claim 17, comprising basing the update of the adaptive voltage scaling on the determination whether the reboot is due to an instability condition and the any one of lifetime of the entity and temperature of the entity, where the any one of lifetime of the entity and temperature of the entity is used as an upper bound for increasing the adaptive voltage scaling.

19. The method of any one of claims 12 to 18, comprising increasing the adaptive voltage scaling when the number of registered reboots due to an instability condition reaches a predetermined threshold.

20. The method of any one of claims 12 to 19, wherein the registering the instability condition comprises registering the instability condition in a non-volatile memory.

21. A computer program comprising instructions which, when executed on a processor of an electronic apparatus, causes the electronic apparatus to perform the method according to any of claims 12 to 20.