Wireless Technology for Home Automation Can Save Energy

Today, most engineers spend a lot of time focusing on producing energy efficient products. However, in the larger scheme of things, even using energy efficient products in a very energy inefficient home or office will not have much impact on overall energy consumption. Home and building automation is where properly integrated technology can greatly impact total energy use.

Until recently, a really energy efficient home or building had to be designed from the ground up, incorporating all the cabling, wired sensors, and wired controls needed to enable the smart use of energy. This has been the stumbling block for many years since it does not support the retrofit market.

That is where low cost, reliable wireless technology comes into play. Several building automation standards are emerging and competing for acceptance, promising to save energy as well as enhance our lifestyles, comfort, safety, and health. Wireless home and building automation is one area that is sure to grow.

This article describes some of the emerging wireless home and building automation standards and the devices used to make them work. Parts described here as well as development tools are available on the DigiKey website.

Early attempts

The first attempt at wireless connectivity for home automation came from BSR, who later became X-10. These modular blocks plugged into outlets and allowed remote control and timed events to dictate when something turned on or off, as well as controlling the intensity level for dimming of incandescent bulbs.

By using the power lines as the signaling medium, X-10 devices did not require any new communication wires and thus called themselves wireless. Although crude by current standards, these devices were somewhat effective, though there were issues when a controller was on a different phase of the power line than the controlled device.

Two other factors did not allow X-10 devices to be real contenders in an effective energy-managed system. First, the slow communications rate (less than 300 baud), and second, the fact that communication could only take place from a controller to a device and not back to the controller. This meant a control system lacked any knowledge regarding the real state of the system at any time. How can you do load balancing, scheduling, and peak demand load processing if you do not know what is on and what is off?

Still, one-way timed and sequenced events can do a lot to save energy. A single push button when leaving your house can make sure the coffee pot is off, the lights are off, the stereo and TV are off, and all is secure. The same holds true when entering – a single switch can turn on all the lights for safety.

The real benefits of home automation only arrive when intelligence is built into the structure. In order to accomplish this, embedded wireless sensors, actuators, user interfaces, displays, and processors all need to work together in a unified and standardized way. Harmony only occurs when everyone is in tune. That is where the emerging protocols hope to make inroads.

Still pulsing after all these years

Power-line modulation as a form of ‘wireless’ is still being used, and, when combined with RF technology, creates a low-cost retrofit solution. For example, INSTEON™ modules work with 131.65 kHz modulated power lines as well as 904 MHz RF to create a dual-band mesh network topology where the power line and RF combine to solve some of the inherent X-10 problems. Useful for lighting control, appliance control, heating and cooling, nodes can be controllers, repeaters, or end devices, and the unit is backward compatible with X-10.

Others like RFXCOM, WGL & Associates, and Z-Wave still offer X-10 support and support for hybrid mixed-network topologies. Training modules for Sigma Designs' Z-Wave solutions are available online from DigiKey. The company’s chips and modules like the ZM310-CME1 (see Figure 1) with its ten I/Os, UART, SPI, Triac driver, and A/D inputs are well positioned to take on many of the battery-powered sensor networks and actuator controllers for smarter homes and buildings.

The ZM310-CME1 also supports international frequency standards for Z-Wave such as the European 868.4 MHz, the US 908.4 MHz, and Australian 921.4 MHz bands, making it usable throughout the international marketplace.

Figure 1: The 8051-based ZM3102N Z-Wave module implements the Z-Wave protocol and provides Flash and user SRAM for OEM home automation applications. (Courtesy of Sigma Designs, Inc.)

New kids on the block

There is a new generation of RF wireless protocols and standards targeting home and building automation as well as applications for fitness, health, media control, lifestyle, and more. By trying to tie it all together, the hope is to have multiple partners who develop interoperable systems.

One protocol that is starting to gain traction as semiconductor manufacturers and product design companies alike begin to embrace it is the ANT+ protocol and standard.

The ANT and ANT+ movement started as a proprietary wireless sensor movement aimed at coordinating and gathering data from exercise equipment, heart rate monitors, and other health and wellness applications. The small and simple stack size, coupled with the ability to create complex network topologies, have created a development organization – the ANT+ Alliance, an indirect subsidiary of Garmin – that specifies data communications and profiles for interoperability. ANT+ can implement peer-to-peer, star, tree, and mesh network topologies using its basic nodes and hubs (see Figure 2).

Figure 2: From point-to-point to complex network topologies, the ANT+ protocol enables a very flexible deployment of network topologies including the ability to provide for alternative routing if a key node becomes inaccessible. (Courtesy of Dynastream Innovations, Inc.)

ANT+ provides support for up to 2³² addresses with a 64-bit network security key. It supports both unidirectional as well as bi-directional communications, and is designed to run for approximately four years on a standard CR2032 coin cell.

A key benefit is the simplified OSI network model, which allows simpler and less code-intensive software stacks (see Figure 3). As such, semiconductor suppliers who are just now developing ANT+ applications will most likely provide the software IP for free to designers who use their chips.

Figure 3: The simplified networking model for the ANT+ protocol means that smaller MCUs with less code space can be used to keep costs low, especially for small, low-power sensors and actuators. (Courtesy of Dynastream Innovations, Inc.)

The 300 members of the ANT+ Alliance include several semiconductor device manufacturers who are starting to offer direct application support to designers wishing to incorporate ANT+ into their products. ANT+ is well suited for many home and building automation applications since simple sensors embedded in the structure can be low power, small, and use small processors without a lot of Flash or RAM.

The Panasonic ENW-89827A2JF is an example of an integratable module that can be quickly dropped into your application (see Figure 4). A member of the company’s PAN1327 family, the 2.4 GHz surface-mount module runs from 1.7 to 4.8 volts and implements Bluetooth and ANT protocols while supporting data rates up to 4 Mbits/sec.

Dynastream also offers an ANT+ transceiver with its 2.4 GHz ANTC782M5IB module, providing up to a 1 Mbit/sec usable data rate. These modules are based on chips from TI like the CC2567, CC2570, and CC2571 ANT RF network processors.

Figure 4: The Panasonic ENW-89827A2JF is an example of the company’s PAN1327 family of embeddable RF modules for ANT+. (Courtesy of Panasonic)

The big kahunas

The most widely deployed non-Wi-Fi Personal Area Network (PAN) devices so far are Bluetooth. However, headphones are now coming with the newest Bluetooth Low Energy (LE) part of the v4.0 spec (see Low-Energy Bluetooth Gets Personal). As a result, Bluetooth LE may be a formidable contender for home and building automation, especially when coupled with a cell phone.

On the other hand, many semiconductor manufacturers are banking on the emerging popularity of ZigBee®. Dozens of semiconductor manufacturers provide chips that are customized for ZigBee designs.

Playing a role similar to the Wi-Fi Alliance, the ZigBee Alliance is responsible for the IEEE 802.15.4 standard specification to which ZigBee compliant devices must adhere. Like other wireless standards, ZigBee operates in the UHF ISM bands providing usable data rates up to 250 Kbits/sec.

The flexible architecture of the protocol allows it to operate in peer-to-peer, star, tree, and mesh configurations, requiring every network to have a coordinator controller at its heart.

Flexible routing, the ability to handle complex and potentially vast networks, and the large number of available devices are all plusses for ZigBee. Unfortunately, the networking stack for ZigBee is a large chunk of code to paste into an MCU. Therefore either scaled back profiles or larger MCUs may be needed for embedded ZigBee applications (see Figure 6).

Figure 6: ZigBee provides a lot of flexibility and interoperability, but uses a larger and more sophisticated stack that must live inside the local MCU’s code area. (Courtesy of ZigBee Alliance)

Fortunately, most semiconductor manufacturers provide royalty- or license-free stack code as long as you are using their chips.

Case in point is the Microchip MRF24J40, a single chip 802.15.4 2.4 GHz transceiver with integrated MAC/PHY functionality which acts as an RF peripheral via its SPI interface. Like other power optimizing devices, the MRF24J40 uses both a 20 MHz crystal for operations and a 32.768 kHz crystal for real-time clock and low-power sleep modes. Like other chip makers, Microchip makes available a free downloadable stack for ZigBee as well as Microchip’s proprietary MiWi and MiWi P2P protocol stacks.

Atmel also provides support for ZigBee with its RF micros and RF transceivers such as the AT86RF231-ZU single chip 2.4 GHz transceiver, which also uses SPI as the external MCU interface. In addition to ZigBee, the AT86RF231 also supports RF4CE (a remote control subset of ZigBee), WirelessHART (an industrial automation protocol), 6LoWPAN (IP V6 in a wireless format), as well as supporting proprietary wireless ISM band development.

NXP is also a player in the ZigBee space and features the JN5148-001-X MCU and transceiver in a single chip. The low power on-chip 32-bit RISC controller works tightly with the on-chip 802.15.4 RF section and a wide range of peripherals to provide a single-chip RF solution for ZigBee and other 2.4 GHz wireless links.

An interesting feature of the JN5148 is that instead of just using a 32.768 kHz crystal for low-power operation, the chip can operate down to 4 kHz, which lowers power even more when RTC accuracy is not required. External interrupts can be used for event detection, keeping power as low as possible until the microcontroller needs to wake up. A protocol stack library is also available to speed development.

Figure 7: The NXP RF transceivers feature a rich mix of peripherals as well as the 802.15.4 transceiver and 32-bit RISC controller. (Courtesy of NXP Semiconductors)

Others sure to emerge

The protocols mentioned here are some of the most commonly talked about and supported protocols so far for automation and energy management, but these are by no means the only choices. Others are sure to emerge, and the use of other existing wireless protocols like Wi-Fi may even flourish.

Wi-Fi may not be the most cost-effective solution, and from a performance point of view it may require more setup time when waking from a low-power sleep state, but some may find it desirable to work within an existing Wi-Fi network.

Either way, relying on semiconductor manufacturers to offer up-to-date interoperable solutions simplifies the learning curve and the tools required by design engineers. This also simplifies compliance going forward, especially as specifications evolve. A case in point is the new ZigBee Smart Energy V2.0 specification, which defines an IP-based protocol aimed at energy delivery, usage, and monitoring. As chips become available, that offer silicon-level support for the new protocol, engineers should find it straightforward to take advantage of the new capabilities.

This brings up another interesting point. You do not have to implement the entire protocol for any of these standards. If you prefer, you can create your own higher-level protocols and data structures that ride over the basic transport layer. This is one way to shrink the otherwise full-blown networking model to make it more suitable for embedded applications.

On the downside, doing so may prevent your device from communicating with other devices that do adhere to the standard. If interoperability is important, then implementing the full stack is the way to go.