Optimising Bluetooth applications for battery power

Whether smart home sensors, wearable technology, or portable industrial sensors, the ability to connect to the consumer or Internet of Things (IoT) is a key aspect of device usefulness and operation. However, making and maintaining the connection consumes power from the tiny power sources that these devices rely upon.

Mark Patrick, Director of Technical Content for EMEA at Mouser Electronics further explores.

User anxiety is often driven by not having an accurate estimate of the remaining charge in the battery, which can lead to the device shutting down at inopportune times. While designers are trying to reduce device size and maximise battery life through efficient design, they are also attempting to provide accurate charge status without depleting the battery in doing so, giving users confidence that the device will continue to operate.

Low-powered wireless connected devices

Across the industrial and consumer sectors, there has been an incredible expansion of portable electronic devices. In the consumer world, individuals often possess multiple connected devices like smartphones, tablets, fitness trackers, and more. This same technology is also revolutionising modern industry.

The scale of the market is huge. According to a 2023 report by ABI Research for the Bluetooth SIG, around 5.4 billion Bluetooth-equipped devices were shipped in 2023, with a projected growth of 9% compound annual growth rate (CAGR) to 7.6 billion shipments per year by 2027 (Figure 1).

Figure 1: Bluetooth is the most common wireless protocol with billions of devices shipping annually. (Source: ABI Research)

These devices are revolutionising our lives, connecting to the wider world through wireless technology using the latest version of protocols such as Bluetooth, Bluetooth Low Energy, Bluetooth mesh, NFC, Thread, and Zigbee. But without this essential connectivity, their utility is, at best, severely hampered.

Every device also requires a power source to operate. For many personal portable devices, this may be an internal rechargeable battery, while for IoT nodes the battery is often a low-capacity coin cell.

Increasing battery utilisation and optimisation

Achieving a balance between energy consumption and device functionality is key. Equally important is providing an accurate estimate of the state of charge of the battery, without having a material impact upon battery drain.

Wireless connectivity is an essential feature, providing a variety of functionality depending on the end device. While many protocols are available for wireless connectivity – including NFC, Thread, and Zigbee – Bluetooth Low Energy is often used in low-power solutions. Unlike the permanently active connection associated with standard Bluetooth, Bluetooth Low Energy is active only when a connection is initiated, remaining in sleep mode at other times.

Devices such as simple sensors that transmit only small amounts of data periodically (for example, an ambient temperature gauge) will be active for just a few milliseconds once every few minutes or so. Using this approach, Bluetooth Low Energy achieves far longer battery life than standard Bluetooth.

Battery fuel gauging (Figure 2) is important in any wireless device to provide the user with an estimate of the remaining battery. Early battery fuel gauges used the principle of coulomb counting to measure the charge entering the battery during charging and assumed that all this charge would be available for powering a device.

However, this assumption is inaccurate due to charge acceptance inefficiencies, storage losses, and discharge losses. These factors result in less than 100% of the energy put into the battery being available for the device.

Figure 2: Users expect to have an accurate estimate of energy remaining in their rechargeable batteries. (Source: MarySan/shutterstock.com)

Additionally, early approaches often did not account for battery ageing. Often, the difference in capacity between a new and old battery can be significant; yet in both cases, the battery gauge would indicate ‘full’, while the available runtimes would be quite different.

Miniaturisation is a further challenge for designers. With each generation, devices get smaller and include ever greater capacity and functionality. The challenge is particularly acute with wrist-worn devices and even more so with hearables that must fit into the ear.

The restricted space available means that designers often prefer highly integrated solutions such as systems-on-chips (SoCs). The design also significantly restricts the space available for the battery, heightening the need for frugal power consumption and accurate battery gauging.

Nordic’s power management IC and wireless solution

Recognising the growing demand for intelligent and efficient power management, Nordic Semiconductor has developed the nPM1300 power management integrated circuit (PMIC). This advanced device includes two ultra-efficient buck converters, two load switches/LDOs, and integrated battery charging. It supports 800mA battery charging for Li-ion, Li-poly, and LiFePO₄ batteries.

To increase the accuracy of battery gauging in compact portable applications, the nPM1300 has a sophisticated fuel gauging mechanism that uses integrated battery current, voltage, and temperature measurements. Nordic’s advanced fuel gauge algorithm combines these measurements with a battery model to provide stable and precise state-of-charge predictions. The standard error remains below ±3% when the battery operates within its rated conditions, with accuracy maintained through temperature compensation across the battery’s operating temperature range.

The fuel gauge algorithm can be executed on the majority of Nordic host SoCs as well as other non-Nordic SoCs. To create the battery model, engineers can carry out an analysis of their battery using the nPM1300 Evaluation Kit (EK) (Figure 3) and the nPM Fuel Gauge Board extension.

Figure 3: Nordic nPM1300 evaluation kit, featuring the nPM1300 PMIC at its centre. (Source: Nordic Semiconductor)

The evaluation kit makes it easy for engineers to develop and configure the PMIC without the need for any coding, as the settings of the nPM1300 can be configured simply through an intuitive graphical user interface (GUI) before being ported to the SoC or microcontroller.

The nPM1300 PMIC streamlines system design by incorporating crucial functionalities needed for embedded Bluetooth Low Energy designs into a compact package. This integration enables improved battery life and more efficient charging, all while reducing the number of components required. The PMIC integrates a hard reset function, accurate battery fuel gauging, a system-level watchdog, power loss warning, and recovery from failed boot into a single, as well as five general purpose input outputs (GPIO) and three LED drivers into its compact package.

Typically, these functions are implemented separately in Bluetooth Low Energy-embedded designs, but the nPM1300 simplifies system design through integration, with configuration achieved through its I²C interface.

The nPM1300 is designed to provide highly efficient power regulation for Nordic’s nRF52 and nRF53 families, including the nRF52840 SoC (Figure 4). The nRF52840 is completely multi-protocol capable and offers support for Bluetooth Low Energy, Bluetooth mesh, Thread, Zigbee, 802.15.4, ANT, and 2.4GHz proprietary stacks.

Figure 4: Nordic’s nRF52840 is a highly featured, ultra-low power multi-protocol wireless SoC. (Source: Nordic Semiconductor)

The nRF52840 SoC is based upon a low-power 32-bit Arm Cortex-M4 CPU with 64MHz floating point unit. There are multiple on-board digital peripherals and interfaces such as high-speed SPI and QSPI for interfacing to external flash and displays, PDM and I²S for digital microphones and audio as well as a full-speed USB device for data transfer.

All peripherals have independent and automated clock and power management. This ensures that they are powered down when not required for task operation to keep power consumption to a minimum without the need to implement and test complex power management schemes. Peak current while transmitting is just 4.8mA, reducing to just 1.5μA in system ON mode.

Furthermore, through bypassing the initial internal regulator on the nRF52840's VDDH input and supplying it with regulated voltages directly from the nPM1300 PMIC, instead of directly from a battery of USB source, the system efficiency can be elevated, increasing the battery runtime by over 20%.

The device contains 1024KB (1MB) of flash memory and 256KB of RAM (Table 1) that can be used for code and data storage. Security needs are met through an on-chip Arm TrustZone CryptoCell cryptographic unit that delivers multiple cryptographic options that execute independently of the CPU.

Conclusion

Battery capacity has always been a challenge in smart devices, requiring designers to trade off between connectivity and runtime – a challenge exacerbated by inaccurate battery gauging.

Innovative new solutions, such the Nordic nPM1300 PMIC, have significantly addressed this challenge by supporting ultra-low-power operation, tight system integration, and accurate battery gauging. By combining the nPM1300 with an SoC, such as the nRF52840, engineers can design products with higher levels of functionality and accuracy whilst reducing the component count and physical size and meeting the latest IoT technical demands.