How to Build Low-Power Sensor Interfaces for the Internet of Things

Power consumption is a vital element in the design of the Internet of Things. Being able to place battery-powered wireless sensors in many different places across the home, factory floor, or office and connect them to the Internet requires very-low-power architecture. Sometimes the sensor technology can limit any reduction in power, but the interface to the rest of the node is essential.

Newer technologies such as micro-machined sensing elements can help reduce the power consumption, but matching the performance and power consumption of an ADC, or having it integrated alongside the sensor, helps reduce the losses. The choice of ADC – usually successive approximation of sigma delta – also helps to reduce the power consumption.

One such example is the Silicon Labs Si8900/1/2 series of isolated monitoring ADCs, which are useful as linear signal galvanic isolators, level shifters, and/or ground loop eliminators used in many applications including power-delivery systems and solar inverters. These devices integrate a 10-bit SAR ADC subsystem, supervisory state machine and isolated UART (in the Si8900), I²C/SMbus port (in the Si8901), or SPI Port (in the Si8902) in a single package.

Figure 1: The Si8900 uses a proprietary CMOS process to combine a galvanic isolator with a data converter for a low-power sensor node.

Silicon Labs has used a proprietary CMOS isolation technology for a choice of 2.5 or 5 kV isolation ratings and integration with the SAR ADC along with a typical common-mode transient immunity performance of 45 kV/µs for robust performance in noisy and high-voltage environments.

Increasingly, the ADC can be integrated into the sensor itself. The ADT7410 from Analog Devices is a high-accuracy digital temperature sensor in a narrow SOIC package that integrates the sigma delta converter to provide a digital output.

Figure 2: The ADT7410 integrates a sigma-delta converter with a temperature sensor to give configurable resolution with low power.

The sensor contains a band gap temperature reference and a 13-bit ADC to monitor and digitize the temperature to 0.0625°C resolution. The ADC resolution, by default, is set to 13 bits (0.0625°C). This can be changed to 16 bits (0.0078°C) by setting Bit 7 in the configuration register (Register Address 0x03).

The ADT7410 is guaranteed to operate over supply voltages from 2.7 V to 5.5 V. When operating at 3.3 V, the average supply current is typically 210 μA. The device has a shutdown mode that powers it down and offers a shutdown current of typically 2 μA to reduce the power consumption of the node.

Sometimes the design needs more flexibility with a discrete ADC, but this does not mean having to increase the power consumption.

The ADC081C021 from Texas Instruments is a low-power, monolithic, 8-bit, analog-to-digital converter (ADC) that supports 100 kHz, 400 kHz, and 3.4 MHz modes. It operates from a single +2.7 to 5.5 V supply that also serves as the reference. The converter is based on successive approximation register architecture with an internal track-and-hold circuit that can handle input frequencies up to 11 MHz. The I²C interface links easily to a microcontroller in all three-speed modes and the device has an automatic power-down mode while not converting to reduce the power.

Normal power consumption using a +3 V supply is 0.26 W, falling to under 1 µW while not converting. These devices are optimized for battery-powered portable applications with a small form-factor and are well suited to sensor nodes in the Internet of Things.

Figure 3: The ADC081C021 uses a successive approximation converter with an automatic power-down mode when not converting to reduce the power.

One way to simplify the provision of a wireless sensor node is to have a proprietary architecture. Analog Devices uses a gateway device to connect to a number of highly-integrated wireless vibration sensors to provide a low-cost sensor network for a wide variety of industrial equipment applications. This uses a micro-machined MEMS 2-axis sensing element that is integrated alongside the processing power and the wireless transceiver. This reduces the power consumption both of the sensing element and of the node by eliminating the power losses that occur by going off chip to another device.

The combination of the ADIS16000 and ADIS16229 creates the sensing network, with the ADIS16000 providing the gateway function to manage the network of ADIS16229 sensor units.

Using a proprietary wireless protocol, one ADIS16000 can support up to six ADIS16229 nodes at one time in a local star network configuration. As the gateway node, the SPI interface of the ADIS16000 provides access to an addressable register map that manages configuration parameters (gateway and sensor node), remote alarm flags, and remote vibration data. The SPI interface of the ADIS16000 enables simple connection to most embedded processors, and its standard SMA connector supports direct connection to a wide variety of antennas. The ADIS16229 requires only an antenna and battery to start up and connect with the ADIS16000 to begin operation.

Figure 4: The combination of the ADI16000 gateway and the ADI16229 creates a dedicated wireless vibration sensor network that can be connected to the Internet.

The ADIS16229 iSensor combines dual-axis acceleration sensing with advanced time domain and frequency domain signal processing. Time domain signal processing includes a programmable decimation filter and selectable windowing function. Frequency domain processing includes a 512-point, real-valued Fast Fourier Transform (FFT); FFT magnitude averaging; and programmable spectral alarms. The FFT record storage system offers users the ability to track changes over time and capture FFTs with multiple decimation filter settings.

The dynamic range, bandwidth, sample rate, and noise performance of the ADIS16229 are well suited for a wide variety of machine health and production equipment monitoring systems. This device also provides a number of wireless configuration parameters, enabling a wide level of flexibility in managing the trade-off between battery life and communication frequency.

The mechanical part of this system includes two different frames, one fixed and one moving, that have a series of plates to form a variable, differential capacitive network. When experiencing the force associated with gravity or acceleration, the moving frame changes its physical position with respect to the fixed frame, which results in a change in capacitance. Tiny springs tether the moving frame to the fixed frame and govern the relationship between acceleration and physical displacement. A modulation signal on the moving plate feeds through each capacitive path into the fixed frame plates and into a demodulation circuit, which produces the electrical signal that is proportional to the acceleration acting on the device.

The ADIS16000 module comes in a 37.8 mm × 22.8 mm × 8.8 mm multichip module laminate (MCML) structure, and the ADIS16229 is available in a 37.8 mm × 22.8 mm × 13.5 mm MCML structure. Both have an SMA connector for simple antenna connection, have two mounting holes for simple installation, and support operation over a −40°C to +85°C temperature range. The ADIS16000 also includes a standard 1 mm, 14-pin connector for connecting to an embedded processor system. The ADIS16229 provides a lead structure that enables simple connection with battery leads.


Combining different ADC architectures alongside the sensors to give a digital output often provides a lower-power approach for developing a battery-powered node for the Internet of Things, but discrete ADCs are using techniques to also reduce the power consumption, giving the designer considerable flexibility. There is even the option of using a complete proprietary network to easily provide a low-power sensing network that can then be connected to the Internet for monitoring and control.