As IoT devices are increasingly used in industrial equipment, home automation, and medical applications, the pressure to optimize power management for these devices is growing. This optimization must be achieved in a compact size without compromising heat dissipation or interfering with wireless communication.
There are various forms of IoT applications, typically referring to a smart connected electronic device that may be battery-powered and sends pre-computed data to cloud-based infrastructure. It utilizes a collection of embedded systems (e.g., processors, communication ICs, and sensors) to collect and respond to data and send it back to a central location or other nodes. This can be anything from a simple temperature sensor reporting room temperature to a device health monitor tracking the long-term health of expensive factory equipment.
Ultimately, developing these devices aims to solve specific challenges, whether automating tasks that typically require human intervention, such as home or building automation, or improving the availability and lifespan of devices in industrial IoT applications. If considering implementing status monitoring applications in architecture-based applications (e.g., bridges), safety can also be enhanced.Application ExamplesThe application areas for IoT devices are virtually limitless, with new devices and use cases being considered every day. Applications based on smart transmitters collect data about their environment to make relevant decisions regarding controlling temperature, triggering alarms, or automatically performing specific tasks. Additionally, portable instruments like gas meters and air quality measurement systems can provide accurate measurements to control centers via the cloud. GPS tracking systems are another application, such as tracking livestock (like cows) using smart ear tags. These are just a small part of cloud-connected devices, with other areas including wearable health applications and infrastructure inspection applications.Industrial IoT applications represent a significant growth area, part of the fourth industrial revolution centered around smart factories. Many IoT applications ultimately aim to achieve factory automation as much as possible, whether through the use of automated guided vehicles (AGVs), smart sensors (like RF tags or pressure gauges), or other environmental sensors deployed around the factory.
ADI identifies five primary areas of focus for IoT:
Smart Health – Vital sign monitoring applications supporting clinical and consumer applications.
Smart Factory – Focused on enhancing the responsiveness of factories and making them more flexible and efficient to build Industry 4.0.
Smart Buildings/Smart Cities – Utilizing smart sensing technology for building security, parking occupancy detection, and implementing temperature and electrical controls.
Smart Agriculture – Utilizing existing technologies to achieve automated agriculture and improve resource utilization efficiency.
Smart Infrastructure – Monitoring mobility and structural health based on condition monitoring technology.
Challenges in IoT DesignIn the ever-evolving field of IoT applications, what are the main challenges designers face? Most of these devices or nodes are installed post-facto or in hard-to-reach locations, making it impossible to power them. This means they must rely entirely on batteries and/or energy harvesting methods for power.Transmitting power around large factories can be costly. For example, consider powering a remote IoT node in a factory. If powering the device requires deploying new cables, not only is the implementation cost high, but it is also extremely time-consuming, so battery or energy harvesting methods are generally chosen to power these remote nodes.Relying on battery power necessitates adhering to strict power budgets to ensure that battery life is extended as much as possible, which will inevitably affect the total cost of ownership of the device. Another drawback of using batteries is the need to replace them after they are depleted. This includes the cost of the battery itself, as well as the high labor costs associated with replacing and disposing of old batteries.Additionally, the cost and size of batteries must be considered, which often leads to over-designing batteries to ensure they have sufficient capacity to meet the battery life requirements, typically requiring over 10 years. However, over-designing adds to the cost and size of the battery, so we must optimize the power budget while minimizing energy consumption as much as possible, keeping the battery size small while still meeting design requirements.For discussion purposes, we categorize power sources in IoT applications into the following three cases, which can be used individually or in combination as needed.
Devices using non-rechargeable batteries (primary batteries)
Devices requiring rechargeable batteries
Devices powered by energy harvesting
Primary Battery Applications
Everyone is familiar with various primary battery applications, also known as non-rechargeable battery applications. These are primarily used for applications that occasionally require power, meaning the device powers on occasionally and then returns to deep sleep mode, consuming very little power. The main advantages of using primary batteries are: they provide high energy density, simple design (since there is no need to include battery charging/management circuitry), and lower costs (because batteries are cheaper and fewer electronic components are required). They are very suitable for low-cost, low-power discharge applications; however, due to the limited lifespan of these batteries, they are less suitable for applications with slightly higher power consumption, and replacing batteries incurs additional costs for both the batteries themselves and the labor to replace them.
Imagine a large IoT device with many nodes. When you ask technicians to replace the battery of a device on-site, they typically replace all batteries at once to save on labor costs. Undoubtedly, this is a waste and exacerbates the global waste problem. More importantly, non-rechargeable batteries only provide about 2% of the energy used to manufacture the battery. About 98% of the energy is wasted, making the economic viability of this power source very low.Clearly, they do have a place in IoT-based applications. Their relatively low initial cost makes them very suitable for low-power applications. They offer a variety of types and sizes, and there is no need for additional electronic components for charging or management, making them a simple solution.From a design perspective, the key challenge lies in how to make the most of the power provided by these small power sources. To this end, we need to spend significant time developing a power budget plan to ensure that the battery’s lifespan is maximized, with a design goal generally set at 10 years.For primary battery applications, we can consider using two products from the micro-power product line: the LTC3337 micro-power coulomb counter and the LTC3336 micro-power buck regulator, as shown in Figure 1.Figure 1. Application circuit for LTC3337 and LTC3336.The LTC3336 is a low-power DC-DC converter with an input voltage of up to 15 V and programmable peak output current. The input can be as low as 2.5 V, making it very suitable for battery-powered applications. In no-load regulation, the quiescent current can be very low, only 65 nA. With continuous improvements in DC-DC converters, it can be easily set up and used in new designs. The output voltage can be programmed based on the connections of OUT0 to OUT3 pins.The companion device to the LTC3336 is the LTC3337, a micro-power primary battery health monitor and coulomb counter. This is another product that can be easily used in new designs, simply connecting the IPK pin according to the peak current requirement (in the range of 5 mA to 100 mA). Some calculations based on the selected battery are then filled in with the recommended output capacitance based on the selected peak current, as detailed in the datasheet.
Ultimately, finding suitable companion devices for IoT applications with limited power budgets. These products can accurately monitor the power usage of primary batteries and efficiently convert the output to usable system voltage.
Rechargeable Battery ApplicationsNow, let’s look at rechargeable applications. For IoT applications that require higher power or higher discharge rates, replacing primary batteries frequently is clearly not suitable, and rechargeable batteries will be a good choice. The initial cost of the battery and the charging circuitry makes the cost of rechargeable battery applications higher, but in high-discharge applications that require frequent discharging and charging, this cost is justified and can quickly be recouped.The initial charge of rechargeable battery applications may be lower than that of primary batteries, depending on the chemistry used, but in the long run, they are more efficient and generally waste less. Depending on the power demand, capacitors or supercapacitors can also be chosen for storage, but they are more used for short-term backup storage.Charging batteries involves several different modes and operating characteristics depending on the chemistry used. For example, the charging characteristic curve of a lithium-ion battery is shown in Figure 2. The bottom is the battery voltage, and the vertical axis represents the charging current.Figure 2. Relationship between charging current and battery voltage.When the battery is severely discharged, as shown on the left side of Figure 2, the charger needs to be smart enough to put the battery into a pre-charging mode, allowing the battery voltage to slowly increase to a safe level before entering constant current mode. In constant current mode, the charger inputs the set current into the battery until the battery voltage rises to the set float charging voltage.Both the set current and voltage depend on the type of battery used, with the charging current limited by the charging rate and the desired charging time, while the float charging voltage is based on the threshold for keeping the battery safe. System designers can help extend the battery’s lifespan by slightly lowering the float charging voltage based on system needs, just as with power considerations, it involves trade-offs. After reaching the float charging voltage, the charging current drops to zero, and the voltage is maintained for a period based on the termination algorithm.Figure 3 shows the behavior characteristic curves of three battery applications over time. The red line represents the battery voltage, and the blue line represents the charging current. It starts in constant current mode, with a maximum current of 2 A until the battery voltage reaches the 12.6 V constant voltage threshold. The charger maintains this voltage for the duration defined by the termination timer, which in this case is 4 hours. Many charger products support programming to set this time.Figure 3. Relationship between charging voltage/current and time.Figure 4 shows a good multifunctional buck battery charger (LTC4162) example, which can provide up to 3.2 A charging current, suitable for various applications, including portable instruments and applications requiring larger batteries or battery packs. It can also be used for solar charging.
Figure 4. LTC4162: 3.2 A buck battery charger.
Energy Harvesting ApplicationsWhen using IoT applications and their power sources, another option to consider is energy harvesting. Of course, system designers need to consider multiple factors, but the allure of free energy cannot be underestimated, especially for applications where power requirements are not too strict and installation locations are inaccessible (i.e., where technical maintenance personnel cannot reach).There are many different energy sources available, and it is not necessarily limited to outdoor applications. Solar energy, as well as piezoelectric or vibrational energy, thermoelectric energy, and even RF energy can all be harvested (though their power levels are low). Figure 5 shows the corresponding energy levels when using different harvesting methods.Figure 5. Energy levels available for various applications.As for the drawbacks, the initial costs are higher compared to the other power sources discussed earlier because harvesting components such as solar panels, piezoelectric receivers, or Peltier elements, as well as power conversion ICs and related enabling components, are required. Another disadvantage is that the overall size of the solution is larger, especially compared to power sources like button batteries. It is challenging to achieve a compact solution when using energy harvesters and conversion ICs.In terms of efficiency, managing low energy levels is also a challenge. Since many power sources are AC sources, rectification is required. We use diodes for rectification. Designers must consider the energy losses due to the characteristics of the components themselves. This impact diminishes with increased input voltage, but not always.Most of the devices discussed in energy harvesting come from the ADP509x product family and LTC3108, supporting a wide range of energy harvesting sources, providing multiple power paths and programmable charging management options, allowing for extremely high design flexibility. Various energy sources can power the ADP509x, but energy can also be extracted from the power source to charge batteries or power system loads. Any energy source can be used to power IoT nodes, from solar (indoor and outdoor) to thermoelectric generators (extracting heat from body heat in wearable applications or engine heat). Energy can also be harvested from piezoelectric sources, adding another layer of flexibility and being a nice way to extract energy from running motors.Figure 6. Functional block diagram of ADP5090 in energy harvesting applications.Another device capable of being powered by piezoelectric sources is the ADP5304, which operates at a lower quiescent current (typically 260 nA in no-load conditions), making it very suitable for low-power energy harvesting applications. The datasheet showcases a typical energy harvesting application circuit (see Figure 7), powered by a piezoelectric source to power an ADC or RF IC.
Figure 7. ADP5304 piezoelectric power application circuit.
Power ManagementWhen discussing applications with limited power budgets, power management should also be considered. Before examining different power management solutions, it is essential to perform power budget calculations for the application. This step is crucial as it helps system designers understand the important components used in the system and how much power each requires. This will influence their decision on whether to choose primary batteries, rechargeable batteries, energy harvesting, or a combination of these options.When researching power management, the frequency at which IoT devices collect signals and send them back to the central system or the cloud is another important factor that significantly impacts overall power consumption. A common approach is to adjust the duty cycle of power usage or extend the time intervals during which the device is awakened to collect and/or send data.
When attempting to manage system power usage, using standby modes for each electronic device (if available) is also a very useful tool.
ConclusionAs with all electronic applications, it is crucial to consider the power management aspect of the circuit early on. This is even more critical in power-constrained applications (such as IoT). Establishing a power budget early in the design phase helps system designers identify effective paths and suitable devices to address the challenges posed by these applications while still achieving high energy efficiency in compact solutions.View Previous Content↓↓↓
