USD

Enhance Battery Life in Wearables Through Efficient Timekeeping During Idle States

By Stephen Evanczuk

Contributed By Digi-Key's North American Editors

For users, the battery life of wearables and other personal electronic devices is a critical factor in purchase decisions. To maximize battery life, developers typically take advantage of the extended idle time for these devices, placing microcontrollers and other power consuming components in a low-power sleep state until user interaction is required. Even in the lowest power sleep states, however, systems require an accurate real-time clock (RTC) to maintain wall clock time and manage scheduled events.

Although developers have had a number of options to support accurate timekeeping in sleep states, few of those options can satisfy emerging requirements for reducing both power consumption and design size.

This article shows how developers can use a power efficient RTC chip from Maxim Integrated in combination with an ultra-low-power microcontroller to extend battery life in wearables, Internet of Things (IoT) devices, and other size- and power-constrained products.

Basic timekeeping

RTCs serve a fundamental function in most designs that need to interact with users or other systems according to real-world clocks and calendars. At its core, an RTC combines a crystal oscillator circuit with a series of registers that hold date and time data accumulated from a countdown chain (Figure 1).

Diagram of RTC timekeeping circuitFigure 1: In a basic RTC timekeeping circuit, a crystal oscillator drives a countdown chain that updates registers holding date and time values. (Image source: Maxim Integrated)

From this basic design, RTC devices have evolved to offer a range of features designed to meet varying application needs for timekeeping accuracy and functional capability. Today, developers can find RTC devices that support different operating voltages, internal memory capacities, and an expansive set of features well beyond just wall clock time and date.

In a growing number of applications, however, timekeeping capabilities are not the sole deciding factor for RTC device selection. As designers respond to demand for smaller, battery-powered products like wearables, the impact of timekeeping on overall system power consumption has gained increase attention. Because systems need to maintain wall clock time even in their lowest power sleep states, designers recognize that timekeeping current optimization has emerged as a critical requirement for these products. At the same time, any useful timekeeping solution needs to meet tight constraints for design simplicity and footprint.

Microcontroller RTC tradeoffs

For some applications, designers might not choose to add a separate RTC device at all but simply rely on the RTC functionality built into many microcontrollers. Of course, not all microcontrollers include a built-in RTC. Those that do typically require periodic recalibration of RTC output to meet the application timekeeping accuracy requirements. Besides the need for additional hardware and software to perform this recalibration, clock errors can accumulate, resulting in incorrect data timestamps before errors meet the threshold for recalibration.

Although these errors can be corrected by synchronizing the device's time with the network, best practices for low-power design call for minimizing network connections to reduce the time that power-hungry radio transceivers are active. Clearly, use of a microcontroller's integrated RTC functionality can present many tradeoffs to developers working to build accurate, low-power designs.

Ultra-low-power microcontrollers such as those in Maxim Integrated’s Darwin family address these concerns with features and capabilities designed specifically for low-power operation (see Build More Effective Smart Devices: Part 1 – Low-Power Design with MCUs and PMICs). For example, in its lowest power "backup" mode with its RTC enabled and no SRAM retention, the ultra-low-power Maxim Integrated MAX32660 Darwin microcontroller consumes about 630 nanoamperes (nA) with a 1.8 volt supply. In backup mode (and in all its operating modes), the RTC circuit accounts for 450 nA, which is less than many standalone RTC devices.

For developers working to maximize battery life, the MAX32660 offers an even lower power option. With RTC disabled in its lowest power backup mode (no SRAM retention), the MAX32660 uses only 200 to 300 nA. The apparent discrepancy between this value and the difference between RTC-enabled backup mode current (630 nA) and the RTC circuit current (450 nA) relates to varying activity in circuits involved in these specific operating states. Of course, this approach means that designers need to find an external RTC device able to operate more accurately and at a lower current than the microcontroller's RTC.

The availability of the Maxim Integrated MAX31341B low-power RTC lets developers take full advantage of the very lowest power modes available in advanced microcontrollers while meeting clock accuracy requirements, despite extended offline operation.

Efficient timekeeping

The Maxim Integrated MAX31341B addresses the growing need for small ultra-low-power RTC devices in battery-powered space-constrained designs. Unlike earlier RTCs, the MAX31341B consumes a miniscule 180 nA during its fundamental timekeeping operations while integrating essential functionality in a small wafer-level package (WLP) measuring 2 millimeters (mm) x 1.5 mm (Figure 2).

Diagram of Maxim Integrated MAX31341BFigure 2: The Maxim Integrated MAX31341B consumes 180 nA timekeeping current while integrating full RTC functionality in a 2 mm x 1.5 mm package. (Image source: Maxim Integrated)

Along with accurate date and time data, the MAX31341B provides time-based alert features used in many applications. On-chip control logic manages a countdown timer and a pair of alarms that can generate output interrupts through the device's ØINTA and ØINTB pins. Developers can reconfigure the device to use ØINTA as the CLKIN input for an external clock to drive the RTC counter. Similarly, ØINTB can be used as CLKOUT to output a square wave at a programmable output frequency set through register settings to the desired divide-by counter.

The device can also be programmed to generate interrupts in response to inputs on its D1 digital input pin or AIN analog input pin. For its analog input, interrupts are generated when the signal on AIN ascends or descends through one of four programmed threshold values (1.3 volts, 1.7 volts, 2.0 volts, 2.2 volts). When operating in this mode, the MAX31341B could signal a host processor, for example, when the RTC supply voltage has fallen below threshold or has been restored, allowing the host to take appropriate action.

The AIN input also plays an important role in the MAX31341B's power management capability, which provides a means to maintain power to the device if the primary voltage supply source becomes unavailable or falls below threshold. With the MAX31341B, developers simply add an external voltage source such as a rechargeable battery or supercapacitor to the hardware design. The corresponding software setup is equally straightforward, only requiring a bit to be set in the device's power management register to configure the device for automatic power management.

When programmed in this mode, the MAX31341B AIN pin serves as the output of a trickle-charge chain comprising a selectable diode, Zener, and choice of three internal resistor paths for setting the desired charging current level (Figure 3).

Diagram of Maxim Integrated MAX31341B RTCFigure 3: The Maxim Integrated MAX31341B RTC integrates a trickle-charge chain that allows developers to programmatically configure the chain and charging current level. (Image source: Maxim Integrated)

In normal operations in this mode, the device draws the trickle-charge current, typically at microamp (µA) levels, from the primary voltage source, VCC. At the same time, the MAX31341B monitors both VCC and the backup source, using the AIN port to follow the backup supply voltage level. If VCC falls below the measured voltage at the AIN pin, the MAX31341B automatically disables the trickle-charge chain and switches its power source to the backup through AIN.

Development support

Maxim Integrated supports designers interested in the MAX31341B's hardware configuration and programmable capability with its MAX31341EVKIT, an evaluation board and accompanying evaluation software application. As illustrated in the evaluation kit's schematic, developers implement the backup hardware design by simply connecting the MAX31341B directly to a backup voltage source such as Eaton’s KW-5R5C334-R supercapacitor (Figure 4).

Diagram of Maxim Integrated MAX31341EVKIT board schematicFigure 4: This section of the Maxim Integrated MAX31341EVKIT board schematic demonstrates that for a timekeeping voltage supply backup, the MAX31341B simply needs a direct connection from its AIN pin to a rechargeable voltage source such as the Eaton KW-5R5C334-R supercapacitor used on the evaluation board. (Image source: Maxim Integrated)

Running on a personal computer connected via USB to the MAX31341B RTC evaluation board, the evaluation software provides a set of tabs for monitoring device timekeeping results and for setting interrupts and registers. Using this software, developers can set the device to operate in power management mode and explore the device's options for configuring the trickle-charge path (Figure 5).

Image of Maxim Integrated MAX31341B RTC evaluation kit softwareFigure 5: The Maxim Integrated MAX31341B RTC evaluation kit software provides a series of menus for setting device registers and programming special features such as power management mode and its trickle-charge chain configuration. (Image source: Digi-Key)

As suggested in the schematic shown in Figure 4, a system design built with the MAX31341B RTC is nearly as simple as the functional block diagram of the hardware interface (Figure 6).

Diagram of Maxim Integrated MAX31341B RTCFigure 6: Developers can add the Maxim Integrated MAX31341B RTC to their system designs with little more than a crystal oscillator, an optional backup voltage source, and a few passive components. (Image source: Maxim Integrated)

As with the backup voltage source, integration of the required external crystal requires no additional components. Unlike earlier RTC devices, the MAX31341B allows use of crystals with an equivalent series resistance (ESR) of up to 100 kilohms (kΩ), enabling use of a wider choice of crystals than possible with those earlier devices.

On the host side, the MAX31341B provides a simple I2C serial interface for interaction with a processor such as the Maxim Integrated MAX32660 Darwin microcontroller. Using this interface, software code running on the host needs only a few instructions to manage MAX31341B operations and access time and date data sequentially or in a single burst.

Using the MAX32660 and MAX31341B, developers can implement ultra-low-power designs able to meet the needs of many applications that rely on accurate timekeeping. In practice, RTC clock errors arising from typical crystal oscillators could lead to problems for some applications, particularly those that need to work across wide temperature ranges.

In tuning fork crystal oscillators used in typical RTC designs, the error rate, expressed as parts per million (ppm), increases as temperature drops or rises relative to the temperature turnover point (the point where the change in error rate goes to zero). For most 32 kilohertz (kHz) crystals, the temperature turnover point ranges between 20°C and 30°C. Outside of this range, a typical crystal exhibits a temperature coefficient between -0.02 to -0.04 ppm/°C2, resulting in double-digit error rates at the high and low temperatures users are likely to encounter.

For example, the datasheet for ECS’s ECS-.327-6-12-TR crystal used on the MAX31341EVKIT evaluation board specifies nominal values for turnover temperature and temperature coefficient of 25°C and -0.03 ppm/°C2, respectively. In turn, the error rate of the MAX31341B RTC clock follows those characteristics, as shown in Figure 7.

Graph of Maxim Integrated MAX31341B RTC clock errorFigure 7: The Maxim Integrated MAX31341B RTC clock error is determined by the performance of the external crystal oscillator, falling away from the crystal's temperature turnover point at a rate determined by the crystal's temperature coefficient. (Image source: Maxim Integrated)

Even at the 20 ppm error rate exhibited in more extreme temperatures, the corresponding clock error amounts to only about one minute per month. The impact of this error rate, of course, can vary significantly for a personal fitness wearable compared to, say, a structural integrity monitor embedded in a bridge. For less critical data, periodic corrections using network resources might suffice. For critical applications, designers might need to compensate for RTC error in timestamps associated with critical data, or use a temperature compensated crystal oscillator (TCXO) such as SiTime’s SIT1552AI-JE-DCC-32.768E, which is specified with 5 ppm stability over its full temperature range of -40°C to +85°C.

Conclusion

Current consumption during extended idle periods has emerged as a significant factor limiting battery life in small, space-constrained devices such as wearables and other mobile products. During these periods, these systems generally require the ability to accurately maintain the current time and date even as most of its components enter low-power sleep states. In using its integrated real-time clock functionality, an ultra-low-power microcontroller may be unable to achieve its lowest level of power consumption.

Designed specifically to provide a lower power solution, an RTC device from Maxim Integrated lets developers maintain accurate timekeeping functions at nanoampere levels. As a result, other system components can sleep in their lowest operating modes during idle periods to maximize battery life in mobile designs.

Disclaimer: The opinions, beliefs, and viewpoints expressed by the various authors and/or forum participants on this website do not necessarily reflect the opinions, beliefs, and viewpoints of Digi-Key Electronics or official policies of Digi-Key Electronics.

About this author

Stephen Evanczuk

Stephen Evanczuk has more than 20 years of experience writing for and about the electronics industry on a wide range of topics including hardware, software, systems, and applications including the IoT. He received his Ph.D. in neuroscience on neuronal networks and worked in the aerospace industry on massively distributed secure systems and algorithm acceleration methods. Currently, when he's not writing articles on technology and engineering, he's working on applications of deep learning to recognition and recommendation systems.

About this publisher

Digi-Key's North American Editors