Power aspect of medical wearables

Although it must be noted that the powering of these devices could prove to be a daunting obstacle to overcome.

Each new generation of fitness band and smartwatch offers an elevated range of functionality - but often the feature sets incorporated into these wearable devices result in higher power consumption being witnessed. The same rules apply to medical devices too of course. The space constraints that wearables are subject to, both in the activity tracking and the patient monitoring arena, prevent the use of larger format batteries and, consequently, more innovative power management techniques need to be employed to keep system power budget levels in check.

If a wearable device is to be effective, it must be comfortable for the patient to wear over an extended period of time (as in most circumstances it may need to be worn constantly, so that the required data can be captured). It should not be too heavy that it leads to patient fatigue - this being especially true when considering elderly patients or infants.

Any wearable medical device will bring together various different key electronics components. These include: 

• A microcontroller unit (MCU), for data management and data processing purposes.
• A battery reserve of some form, for storing the energy needed to run it.
• Sensors, such as an accelerometer, a gyroscope, a heart rate monitor, etc.
• A communication interface, such as one supporting Bluetooth Low Energy (BLE) or Near Field Communication (NFC).
• Cryptographic mechanisms, to take care of data protection (to ensure that access can’t be gained to private medical information).

The use of MCUs and sensors defines the characteristics and capabilities of the device, and therefore the intended application for which it will be utilised. Clinical medical devices must be highly accurate and reliable, as well as being simple to manage and having highly efficient data protection capabilities.

They can have recording times that extend to a year or more. A system-level power management engineering approach to the design of such items will consist of several functional blocks, with grouped power domains based on shared needs. In this way, the best layout can be implemented for each domain in order to maximise efficient energy use (as illustrated in Figure 1).

Certain domains can be completely disabled when they are not being utilised, such as the RF element, or high duty cycle functions which are not time sensitive. The MCU at the heart of the system must merge ultra-low power management with the high integration of peripherals.

Figure 1: Typical internal circuit configuration of a wearable device with various power domains.

Battery technology

The battery chemistries most commonly used in wearable devices are lithium-ion (Li-Ion) and lithium-ion polymer (LiPo). Li-Ion batteries comprise of organic electrolytes and have nominal voltages from 3.2V to 4V (both primary and secondary). The two chemistries differ from each other in terms of their storage capacity - with Li-Ion being higher.

Li-Ion batteries have a relatively small impact on the environment and can boast modest dimensions and a lightweight construction. However, their capacity is directly linked to their size, and thus incurs a strong degradation in space-restricted designs.

Advances in material technology, such as the implementation of graphene, should ensure that storage levels far superior to what current Li-ion resources deliver can be benefited from in the future, with numerous research projects now underway all sharing this goal. The progression being made in relation to nanotechnology will allow supercapacitors to also be considered as a way of storing electrical energy for wearable use.

Ultra-low power MCUs

In a medical wearable context, the MCU must be able to consume currents in the order of nA while in standby mode, reaching a maximum of a few hundred µA when fully operational. The MCU should not merely deal with the data coming from the sensors, it also has to manage the power supply, delivering current only when a particular part of the system needs it, so that energy in not wasted unnecessarily.

Aimed at wearable device deployment, the MAX32660 MCU from Maxim Integrated (details shown in Figure 2) provides a compelling combination of strong operational performance and energy efficiency. It features a 32-bit ARM Cortex-M4 core with a built-in FPU processor, and offers access to sufficient memory to run advanced algorithms, as well as managing any sensors that are interfacing with it.

This MCU also boasts industry-leading power performance figures (reaching up to 50μW/MHz), in a compact WLP form factor package with external dimensions of just 1.6x1.6mm.

Figure 2: Block diagram for the MAX32660.

Microchip offers a wide range of ultra-low power 32-bit MCUs to meet the needs of the wearable medical market. This goes from the smallest SAM D MCUs, which are based on ARM Cortex-M0+ architecture, all the way to the ultra-low power SAM L series and the heightened performance PIC32MX XLP series.

These MCUs can provide power consumption per cycle of less than 35µA/MHz for active mode and draw just 200nA in sleep mode. They integrate a wealth of functionality - such as LCD ports, operational amplifiers, real-time clocks and mTouch sensing, plus USB and DMA interfaces. Low power peripherals are connected so as to allow almost no latency in data sharing without CPU interruption (Figure 3).

Figure 3: Microchip PIC32MX block diagram.

Further 32-bit MCU solutions used in the wearable sector include the Silicon Labs EFM32 Giant Gecko ARM Cortex-M3 based units. Giant Gecko MCUs are available with autonomous low-energy peripherals, including AES encryption for increased security, UART, a low energy sensor interface, and operational amplifiers (see Figure 4 for more details).

Figure 4: EFM32 block diagram.

Energy harvesting 

As we have already discussed, one of the main reasons why wearable devices are still not as widely used by the medical profession as they could be is because of the need to charge the battery regularly. The patient may forgot to do so, or it might be necessary for a caregiver to take responsibility for doing it. Through battery chemistries are improving, an alternative tactic could be to focus on energy harvesting technologies instead.

These collect energy from external sources (such a light from the sun, thermal gradients or movement) to charge the medical device - and can therefore operate indefinitely without the need for battery recharges.

Energy harvesting from human movement or heat is dependent on energy dissipated from the wearer. For the average person this figure is about 107J/day - which is equivalent to 20kg of batteries storing 2500mA/h and theoretically speaking would provide ample energy reserves to draw upon, if it could be done efficiently.

Thermo-electric energy transforms heat into electricity by exploiting the Seebeck effect, i.e. generating a potential difference proportional to the temperature gradient between the hot and cold sides. A pair of semiconductors that make up a Peltier cell would furnish the system design with the necessary elements. For wearable devices, the human body, which continually emits heat, could be used as the warm side, while the surrounding environment would serve as the cooler side needed for thermoelectric collection.

One of the main advantages of this set up would be the constant availability of energy during the day and the night, unlike solar energy harvested from photovoltaic cells (which can only take place in daylight hours).

Another form of energy harvesting is mechanically-oriented and relates to the vibration and movement of the patient’s limbs. Here piezo-electric elements would generate an electric current each time that a mechanical apparatus was operated. For energy harvesting in wearable devices, piezoelectric elements are often designed to produce electrical currents by using vibrations that occur when walking, breathing or moving the arms and legs.

The Energy Harvesting Solution to Go development kit from Wurth Electronics allows easy access to energy harvesting technologies, helping developers build power management circuits for wearable devices.

DC/DC converters

The DC/DC converter has the task of maintaining a constant voltage for all peripheral devices. Through the guidance of the MCU, it manages every portion of the energy stored in the accompanying Li-Ion batteries. These components are required to meet stringent specifications and at the same time consume as little energy as possible.

The LTC3107 from Linear Technology (as detailed in Figure 5) is a highly integrated DC/DC converter designed to prolong the life of the main battery in power-limited systems, in conjunction with an energy harvesting mechanism. Together with a small supplemental source of thermal energy, it can significantly extend battery life, thus reducing recurring maintenance costs related to battery replacement.

Figure 5: Typical application circuit for the LTC3107 DC/DC converter.

The continued miniaturisation trend in wearable medical technology, with smaller and more tightly integrated packages, requires advanced power management systems and, in many cases, the employment of some form of energy harvesting. The choice of digital circuitry, battery resources and accompanying DC/DC converter solutions are all fundamentals of the ultra-low power design concepts required.