Microcontroller Power Efficiency Optimization for Ultra-Low-Power Designs

By: Jukka Eskelinen, product marketing director for tinyAVR series

Kim Meyer, Field Applications Engineer

Atmel Corporation

Whether for consumer, industrial, or medical applications, power optimization is typically achieved by reducing active processing time and extending processor sleep mode time. However, with the advent of ultra-low power applications, this approach is no longer sufficient. Single-cell operation, charging and discharging near battery thresholds, the need to control motors and/or high-brightness LEDs, and reducing device form factors and cost, trends have changed the way developers optimize power consumption.

For electric toothbrushes, PMPs, remote controls, wireless sensors, and other portable and hand-held devices, power management must be incorporated into all levels of the system. Power regulation can be achieved system-wide by optimizing power consumption through efficient cell voltage conversion, utilizing multiple current modes, introducing intelligent battery management, and employing energy-saving techniques at the application level.

Efficient voltage conversion

Many ultra-low power applications are moving towards single-cell architectures to reduce device cost, size and weight. These three elements are also key to the success of battery-powered portable applications. In many cases, the battery is even heavier than all the other components plus the PCB. Also, standard AA or AAA batteries are usually the largest components on the PCB. Reducing the power supply to a single battery is attractive because it simplifies the battery holder design and makes the overall structure of the product lighter.

However, the design of single-cell power supplies also brings a variety of new challenges to designers. Although when fully charged, the voltage range of a single cell is usually 1.2V-1.5V, but in fact even if the battery voltage drops below 1V, there is still a considerable amount of energy available. An MCU with a supply voltage of 1.8V requires at least two batteries to operate in series. Some applications, such as driving high-brightness LEDs with large forward voltages, require as many as four batteries. In order to drive motors, LEDs, and even the processor itself from a single battery, a regulator must be used to boost the existing voltage to a suitable level. However, a boost regulator costs almost as much as an MCU and takes up a lot of PCB space. In addition, some regulators must be controlled by the MCU, further increasing the complexity of the design.

The seamless operation of an integrated self-managing boost regulator within the MCU not only avoids most of the cost and space issues associated with external regulators, but its MCU also provides higher effect. For example, the integrated regulator ATtiny43U (see Figure 1) is capable of boosting voltages as low as 0.7V, enabling discharge closer to the limit of battery capacity than technologies supported by other types of implementations. An integrated regulator also achieves fairly low reactive current (typically 1µA for the ATtiny43U) and can automatically start once there is enough voltage (1.2V for a full battery or nearly complete charge).

Microcontroller Power Efficiency Optimization for Ultra-Low-Power Designs

Figure 1. Integrating a boost regulator enables the ATtiny43U to operate from a single cell with voltages as low as 0.7 V, effectively driving load currents up to 10mA. Also, it allows discharge closer to the limit of the battery’s capacity than other types of implementations.

In addition, this regulator supports all battery technologies, giving designers full freedom to choose the best battery for a specific application. The battery voltage range is 0.7V-1.8V, and developers can use 1.6V alkaline or silver oxide batteries, 1.5V lithium batteries, 1.4V zinc-air batteries (Zinc-Air), and 1.2V NiMH and NiCd batteries, etc. .

Boost and Low Current Modes

High current capability without external drive circuitry is also important for many applications. The ATtiny43U’s boost regulator has up to 30mA current drive capability, enabling direct control of bright LEDs and small motors. Since the regulator is an integrated part of the MCU, it can be optimized for the architecture to maximize efficiency. For example, Figure 2 shows the conversion efficiency of the ATtiny43U for a specific load current based on residual charge.

Microcontroller Power Efficiency Optimization for Ultra-Low-Power Designs

Figure 2. An integrated boost regulator optimized for its MCU architecture maximizes conversion efficiency at different loads and supply voltages. Integrated regulators also reduce board space requirements and overall system cost by eliminating the need for external regulators.

As shown, high current operation is less efficient than low current operation. However, most high current applications do not require continuous operation in high current mode. For example, an electric toothbrush or a camera only occasionally turns on the motor. If their architecture is locked in high-current mode, these devices will operate very inefficiently even when only a small amount of power is required; that is, the regulator will suffer from low-efficiency characteristics under high-current operation. provide low current.

To maintain efficiency, the MCU must be able to support multiple operating modes. Thus, when the device requires a high current and a tightly regulated Vcc, the MCU and regulator work in Regulated Mode. On the other hand, when the motor or other peripherals are idle and the load current drops below 0.6mA, the regulator automatically switches to Low Current Mode to more effectively regulate power consumption.

Microcontroller Power Efficiency Optimization for Ultra-Low-Power Designs

Figure 3. This figure shows a typical output voltage curve of a boost converter at various loads. At light or no load (green curve), the measured transition time (rising voltage) is in the hundreds of microseconds, while the idle time (falling voltage) is in seconds. Note that this change occurs when the MCU is in power saving mode or consumes very little power. In the main operating mode, the active regulation mode, the output voltage remains stable (3V +/- 100mV) (red curve).

In addition, at light or no load, the converter in regulation mode will periodically reach its low duty cycle limit. By automatically switching to low-current mode, the converter stops switching, minimizing power consumption while still remaining active (see Figure 3). This change in output voltage occurs when the MCU is powered down or consumes very little power. In the main working mode, namely Active Regulated Mode, the output voltage remains stable (3V±100mV). Also note that the typical conversion voltage will vary with battery power consumption (see Figure 4). The regulator is an independent subsystem that does not require active management by the MCU. However, for designers who need more direct control of the boost regulator, some features can be controlled in software.

Microcontroller Power Efficiency Optimization for Ultra-Low-Power Designs  

Figure 4. The typical transition range between Active Low Current Mode and Active Regulated Mode depends on the available input voltage and load.

Since the actual efficiency depends on the application, it makes no sense to integrate all the passive components related to power regulation. For example, cost is the dominant factor in some markets, while in others, the most important driver may be longevity. Rather than being forced to use passive components optimized for other markets, or products that are satisfactory but not optimal for all applications, developers should choose passive components that provide the best balance for their application. And this can be done with just a few components (ie, an Inductor, two bypass capacitors, and a Schottky diode).

Whether for consumer, industrial, or medical applications, power optimization is typically achieved by reducing active processing time and extending processor sleep mode time. However, with the advent of ultra-low power applications, this approach is no longer sufficient. Single-cell operation, charging and discharging near battery thresholds, the need to control motors and/or high-brightness LEDs, and reducing device form factors and cost, trends have changed the way developers optimize power consumption.

For electric toothbrushes, PMPs, remote controls, wireless sensors, and other portable and hand-held devices, power management must be incorporated into all levels of the system. Power regulation can be achieved system-wide by optimizing power consumption through efficient cell voltage conversion, utilizing multiple current modes, introducing intelligent battery management, and employing energy-saving techniques at the application level.

Efficient voltage conversion

Many ultra-low power applications are moving towards single-cell architectures to reduce device cost, size and weight. These three elements are also key to the success of battery-powered portable applications. In many cases, the battery is even heavier than all the other components plus the PCB. Also, standard AA or AAA batteries are usually the largest components on the PCB. Reducing the power supply to a single battery is attractive because it simplifies the battery holder design and makes the overall structure of the product lighter.

However, the design of single-cell power supplies also brings a variety of new challenges to designers. Although when fully charged, the voltage range of a single cell is usually 1.2V-1.5V, but in fact even if the battery voltage drops below 1V, there is still a considerable amount of energy available. An MCU with a supply voltage of 1.8V requires at least two batteries to operate in series. Some applications, such as driving high-brightness LEDs with large forward voltages, require as many as four batteries. In order to drive motors, LEDs, and even the processor itself from a single battery, a regulator must be used to boost the existing voltage to a suitable level. However, a boost regulator costs almost as much as an MCU and takes up a lot of PCB space. In addition, some regulators must be controlled by the MCU, further increasing the complexity of the design.

The seamless operation of an integrated self-managing boost regulator within the MCU not only avoids most of the cost and space issues associated with external regulators, but its MCU also provides higher effect. For example, the integrated regulator ATtiny43U (see Figure 1) is capable of boosting voltages as low as 0.7V, enabling discharge closer to the limit of battery capacity than technologies supported by other types of implementations. An integrated regulator also achieves fairly low reactive current (typically 1µA for the ATtiny43U) and can automatically start once there is enough voltage (1.2V for a full battery or nearly complete charge).

Microcontroller Power Efficiency Optimization for Ultra-Low-Power Designs

Figure 1. Integrating a boost regulator enables the ATtiny43U to operate from a single cell with voltages as low as 0.7 V, effectively driving load currents up to 10mA. Also, it allows discharge closer to the limit of the battery’s capacity than other types of implementations.

In addition, this regulator supports all battery technologies, giving designers full freedom to choose the best battery for a specific application. The battery voltage range is 0.7V-1.8V, and developers can use 1.6V alkaline or silver oxide batteries, 1.5V lithium batteries, 1.4V zinc-air batteries (Zinc-Air), and 1.2V NiMH and NiCd batteries, etc. .

Boost and Low Current Modes

High current capability without external drive circuitry is also important for many applications. The ATtiny43U’s boost regulator has up to 30mA current drive capability, enabling direct control of bright LEDs and small motors. Since the regulator is an integrated part of the MCU, it can be optimized for the architecture to maximize efficiency. For example, Figure 2 shows the conversion efficiency of the ATtiny43U for a specific load current based on residual charge.

Microcontroller Power Efficiency Optimization for Ultra-Low-Power Designs

Figure 2. An integrated boost regulator optimized for its MCU architecture maximizes conversion efficiency at different loads and supply voltages. Integrated regulators also reduce board space requirements and overall system cost by eliminating the need for external regulators.

As shown, high current operation is less efficient than low current operation. However, most high current applications do not require continuous operation in high current mode. For example, an electric toothbrush or a camera only occasionally turns on the motor. If their architecture is locked in high-current mode, these devices will operate very inefficiently even when only a small amount of power is required; that is, the regulator will suffer from low-efficiency characteristics under high-current operation. provide low current.

To maintain efficiency, the MCU must be able to support multiple operating modes. Thus, when the device requires a high current and a tightly regulated Vcc, the MCU and regulator work in Regulated Mode. On the other hand, when the motor or other peripherals are idle and the load current drops below 0.6mA, the regulator automatically switches to Low Current Mode to more effectively regulate power consumption.

Microcontroller Power Efficiency Optimization for Ultra-Low-Power Designs

Figure 3. This figure shows a typical output voltage curve of a boost converter at various loads. At light or no load (green curve), the measured transition time (rising voltage) is in the hundreds of microseconds, while the idle time (falling voltage) is in seconds. Note that this change occurs when the MCU is in power saving mode or consumes very little power. In the main operating mode, the active regulation mode, the output voltage remains stable (3V +/- 100mV) (red curve).

In addition, at light or no load, the converter in regulation mode will periodically reach its low duty cycle limit. By automatically switching to low-current mode, the converter stops switching, minimizing power consumption while still remaining active (see Figure 3). This change in output voltage occurs when the MCU is powered down or consumes very little power. In the main working mode, namely Active Regulated Mode, the output voltage remains stable (3V±100mV). Also note that the typical conversion voltage will vary with battery power consumption (see Figure 4). The regulator is an independent subsystem that does not require active management by the MCU. However, for designers who need more direct control of the boost regulator, some features can be controlled in software.

Microcontroller Power Efficiency Optimization for Ultra-Low-Power Designs  

Figure 4. The typical transition range between Active Low Current Mode and Active Regulated Mode depends on the available input voltage and load.

Since the actual efficiency depends on the application, it makes no sense to integrate all the passive components related to power regulation. For example, cost is the dominant factor in some markets, while in others, the most important driver may be longevity. Rather than being forced to use passive components optimized for other markets, or products that are satisfactory but not optimal for all applications, developers should choose passive components that provide the best balance for their application. And this can be done with just a few components (ie, an inductor, two bypass capacitors, and a Schottky diode).

Smart battery management

Accurately estimating remaining energy is an important factor in maximizing TV power usage. For example, rechargeable batteries require close monitoring and charging control within a set range to ensure safe battery usage and the longest possible lifespan. The more accurate the estimate of the remaining charge, the closer the battery can get close to its limit capacity to safely charge and discharge without worrying about damage to the battery from overcharging and discharging.

While more sophisticated control of battery charge and discharge means more energy is available for the battery to last longer, this type of control lacks flexibility and can severely limit the battery technologies a processor can support. For example, batteries with different chemistries have different safe charge and discharge voltage thresholds, and if the MCU has a fixed threshold or is limited in how the threshold is configured, it can become a technical barrier to the battery that the MCU can effectively manage. As a result, developers may be forced to use a specific battery based on the chosen MCU, rather than choosing the most suitable battery technology.

For applications where batteries must be replaced, the flexibility to support rechargeable batteries is critical. Rechargeable batteries have very different thresholds than disposable batteries and can compromise their overall charge capacity if drained excessively. The resulting reduction in usage time is likely to be classified as a device failure rather than a battery failure. The firmware of the ATtiny43U utilizes the built-in ADC to monitor the battery voltage and decide when to put the device into Stop Mode, thereby completely draining the primary battery while ensuring that the rechargeable battery will receive the best possible charge over multiple charge cycles. long usage time.

While the automatic shutdown of the processor can protect the rechargeable battery, a sudden power loss may not be acceptable from an application point of view. For example, turning off the camera suddenly leaves the lens exposed, making it vulnerable to damage. Therefore, designers can accurately estimate the remaining energy through an important power management element. For example, using the 10-bit ADC of ATtiny43U to measure the battery voltage at regular intervals can achieve the aforementioned purpose. Using this approach, there is an opportunity to put individual devices into a safe configuration before the device shuts down.

High power efficiency at the application level

Many applications incorporate an MCU as a secondary processor to the host processor, offloading tasks such as Display refresh, keyboard monitoring, small motor work, and intelligent battery management. The advantage of using a secondary processor is that the MCU can perform these functions with greater power efficiency than the application processor. For example, an application processor that monitors the keyboard must be woken up frequently to perform tasks. And because the power consumption of the MCU in working mode is less than that of the application processor, using the MCU to monitor the keyboard and update the Display can make the application processor sleep continuously for a longer time, thereby saving considerable energy.

Of course, processing efficiency also has a major impact on power efficiency, as the more work the MCU can perform per cycle, the faster it can go into sleep mode. Increasing the clock frequency increases power consumption, so a more efficient MCU architecture can support a dynamic operating frequency and execute instructions in a single cycle, and perform automated peripheral management.

Ultra-low-power MCUs also require multiple sleep modes. For example, a sensor application could monitor the temperature until it exceeds a threshold. If the entire MCU is in working mode during the monitoring period, more energy is consumed than is actually required. Different sleep modes are supported, allowing developers to shut down different parts of the device for better power savings (see Table 1).

Microcontroller Power Efficiency Optimization for Ultra-Low-Power Designs

Table 1. Ultra-low-power MCUs have multiple sleep modes, so developers can configure an consumes MCU.

There are several architectural innovations in the ATtiny43U architecture that developers can use to improve power efficiency in active and sleep modes:

Precise supply voltage: While the MCU can accept a single voltage supply, it may architecturally have multiple different internal voltages. Such a design approach results in low power efficiency because the dynamic power is higher than expected. When all analog peripherals, flash, EEPROM, and RAM operate at the same voltage, the overall power consumption of the device is reduced.

Minimize leakage current: Temperature, supply voltage and process technology all affect leakage current. Rather than modifying existing architectures to operate at lower voltages, ultra-low-power MCUs must be designed from the ground up with power efficiency in mind, as exemplified by Atmel’s picoPower AVR microcontroller family .

Brown-Out Detection (BOD): While zero-power brown-out detectors do not consume power, they are also slow to respond and may take a full millisecond to detect a voltage below the threshold, This creates a risk for the MCU. In contrast, the “sleep BOD” can detect an undervoltage condition within 2 microseconds while consuming only 20μA. Since the MCU does not need under-voltage protection in deep sleep mode, the sleep BOD can be turned off at this time, and zero power consumption can be achieved. Using this approach, developers can achieve low power consumption and fast response at the same time.

Digital Input Interrupt Register (DIDR): Multiple inputs to peripherals (such as ADCs) to increase design flexibility for small pin-count devices. However, transistors containing input buffers will leak current when loaded with voltages in the Vcc/2 range. At this time, if a dedicated input interrupt register is used, and a disable bit is added to each analog input, the developer can disable the input buffer individually to avoid leakage.

Clock Gating: Clock gating techniques can reduce the switching frequency of any clock domain. Any unused clock can be gated to avoid unnecessary power consumption.

Power-Saving Registers: While the various sleep modes simplify power management, they tend to only enable or disable the entire peripheral portion. In this way, even if only one peripheral is used, the other peripherals must be in working condition. The Power Reduction Register allows developers to control the switching of each peripheral module completely independently. Disabling a peripheral module in active mode can reduce overall power consumption by 5-10%; in idle mode, it can save 10-20%.

Flash Sampling: Traditional flash designs are designed to remain active in active mode. However, at lower clock rates, flash read times will be less than clock cycles. The flash sampling technique is to have the flash sample the contents of the array at a speed of the order of 10ns and then disable it immediately, thereby reducing the average power consumption.

Fast wake-up: If the system wakes up slowly, it has to stay in working mode for a longer time to accommodate longer delays and prevent interruption of real-time event processing. In other words, the faster the MCU wakes up, the longer it stays in sleep mode.

When evaluating different ultra-low-power MCU specifications, developers must be clear-headed to ensure comparison of equivalent measurements. For example, consideration should be given to:

Efficiency over a range: Efficiency specifications are usually given based on the best measured (sweet spot) results of the MCU rather than results on load current and voltage. The typical operating range of an application may place it on a lower efficiency curve. Additionally, efficiency must be estimated over the entire voltage drop range of the cell.

Safe battery operating range: Although the power consumption of the MCU may be relatively small, if the voltage and temperature cannot be measured accurately enough, the battery limits can be exceeded, resulting in battery damage and reduced life time. Accuracy is a key factor when determining how much battery energy a device can safely use.

Regulator inefficiency: MCUs without a boost regulator have higher efficiency specifications because the switching losses are hidden in the external regulator. Also, in single-cell designs, if the MCU does not have an integrated regulator, remember to factor in the cost and design complexity of an external boost regulator.

Efficiency over the device’s entire usage range: The MCU may be very efficient when driving high currents, but it will be very inefficient when driving low currents unless it has multiple operating modes. Therefore, if the application does not often require high current capability, the overall efficiency is reduced.

· Whether the specifications are measured with a single or multiple batteries: Some MCU specifications will vary with the number of batteries used. For example, if there are multiple batteries, the internal boost regulator can be avoided, thereby improving efficiency. Conversely, when only a single battery is used, various specifications (such as wake-up time) obtained with multiple batteries may be degraded.

Maturity of the development environment: Achieving ultra-low power requires innovation at the architectural layer. Ultra-low-power MCUs based on new architectures often offer at best limited design tools that are still in development. Since software development is one of the most important cost factors, the stability, integrity and functionality of design tools play a pivotal role in helping developers manage power consumption efficiently and get products to market quickly.

Figure 5. Using demonstration tool kits such as the STK600 and ATtinyx3U top-level modules, developers can measure power efficiency under real operating conditions. These tool kits allow developers to fully use the features of the ATtiny43U and Ateml’s rich and mature development tool kit to test single-cell operation, obtain power curves for high-brightness LEDs, and adjust power thresholds to fully utilize the maximum capacity of the battery within a safe range.

One of the ways to determine how “ultra-low” the power consumption of an MCU is is to measure it yourself. Demo Kits provide a proven means to test the efficiency of an MCU under real operating conditions and to take advantage of its feature set. For example, as long as the ATtinyx3U top module (top module) is connected to the STK600 development board, developers can fully use the functions of the ATtiny43U and Atmel’s comprehensive development tool suite (see Figure 5). Using this module, developers can test the limits of single-cell operation, program power consumption profiles while directly driving high-brightness LEDs, and drive the auto-shutdown and power-up functions of the integrated boost regulator to adjust power thresholds , make full use of the maximum capacity of the battery within a safe range.

Summary of this article

The single-cell design eliminates the need for a backup battery load, which tends to be the heaviest and largest component in an ultra-low power system. MCUs with integrated on-chip regulators and configurable modes can effectively bridge the gap between the MCU’s extremely small supply voltage and the typical output voltage of standard single-cell technology, allowing developers to adapt existing load conditions and battery voltages. Power consumption is minimized. With only one battery, no external regulators required, and with battery draw down to 0.7V and high current capability for LEDs and small motors, designers can Compact battery powered device.

By: Jukka Eskelinen, product marketing director for tinyAVR series

Kim Meyer, Field Applications Engineer

Atmel Corporation

The Links:   NL4864HL11-02A B104SN01-V2 IGBT-CENTER

Related Posts