Embedded Technologies for the Smart Grid
SoCs Driving Higher Levels of Integration in Smart Appliances
Ubiquitous connectivity is being realized from handheld communication devices to smart appliances within our homes and working environments. IC manufacturers are driving advances in semiconductor processes and integration in the form of System-on-Chip (SoC) platforms.
RUFINO OLAY AND REGHU RAJAN, MICROSEMI
Page 1 of 1
In recent years, SoCs have increased in complexity from purely analog or digital devices to mixed signal ICs that can be field upgraded to adjust to evolving communication protocols or regional deployment requirements. The next evolution of the SoC is to include an RF subsystem onto the mixed signal controller IC platform while ensuring shrinking power consumption budgets are met. Energy harvesting methods ensure that remote or self-powered devices are capable of increasing the duration of operation in low-power modes.
Designing for an Evolving Smart Grid
The term “smart” is increasingly used to describe an intelligent bi-directional communication between our community, environment and products. For example, an electric, water or natural gas smart meter is but one component of a smart grid system providing numerous functions such as:
• Accurate real-time load data
• Secure two-way communication to a host network or the Internet
• Ability to connect/disconnect nonessential loads dependent on preset parameters or user feedback
• Rugged, reliable, low-power operation
With the increased functionality and capability of smartphones, consumers are becoming more comfortable with monitoring and controlling greater aspects of their lives. The rise of smart appliances is a natural extension of the control that consumers are demanding to not only make their lives easier by allowing them to accomplish more with their precious time, but to also control their financial expenditures. Simple push-to-start or timer-based systems are evolving into sophisticated controller-based systems with high levels of connectivity. Energy consumption can be monitored and usage adjusted to align with more favorable time-varying pricing. As a result, appliance manufacturers will have to invest in telecommunications and software development utilizing platforms that allow them to quickly get to market while providing a flexible path for upgradability due to evolving communication protocols.
With this in mind, companies must carefully approach the design of their smart appliances from an architectural level in order to take into account current and future requirements. Semiconductor advances in the form of smaller process geometries and packaging technologies coupled with mixed signal capabilities, higher reliability and decreased power consumption are giving engineers more flexibility to realize their designs.
Putting the “Smarts” into Next Generation Appliances
Mixed signal SoCs are progressively becoming the platform of choice for complex system design and provide a single system development platform for both the software and hardware engineering teams.
Figure 1 shows the major subsections of a non-volatile flash process based intelligent mixed signal SoC, which includes an embedded ARM Cortex-M3 microcontroller, programmable analog and FPGA logic fabric. The combination of a standard MCU and programmable logic with the FPGA provides the ability to tackle system-level algorithms such as power management, communication interfaces, encryption for data security and more.
Mixed Signal SoC Platform.
Numerous analyses can be performed on load monitoring data such as power factor and harmonic content. Algorithms such as Fast Fourier Transform (FFT) can be cycle intensive and therefore implementation within the FPGA fabric is recommended to free up the Cortex-M3 microcontroller for other high-level system tasks. In general, partitioning the computationally intensive algorithms into the FPGA fabric allows for parallel processing techniques that historically have been suitable for FPGA implementation, which reduces cycle times and power requirements.
Secure and Low Power Connectivity
In smart appliances, connectivity is an integral function. Unfortunately, no single communication protocol has yet been widely standardized, thus the ability to customize to regional and market needs is key. Multi-protocol support can be realized with different RF front ends or TCP/IP. To save space, protocol stacks can be stored in memory and loaded into the Cortex-M3 microprocessor during initial set-up or field upgraded in the field as necessary.
Data integrity and security are of utmost importance in a two-way communication system. Of first consideration is the hardware. During initialization, the configuration data must be stored on-chip to avoid unauthorized access to the boot-code. This makes it virtually impossible to steal the configuration data as compared to external memory bit stream interception. Secondly, an Advanced Encryption Standard (AES) of at least 128-bit cryptographic keys provides another layer of protection.
Short-range sensor networks are widely used today for wireless communication in factories, industrial complexes, commercial and residential buildings, agricultural settings and urban areas, where they improve manufacturing efficiency, safety, reliability, automation and security. They can be used in ambient/environmental monitoring, building automation and security, access control, structural health monitoring, tire pressure monitoring systems, tank level monitoring, wireless cold chain tracking for pharmaceutical shipments, and flexible smart cards for embedded autonomous sensors, to mention a few. Until recently, almost all sensor networks have used costly wired data communications and power connections. Moving to wireless protocols eliminates the data communications wiring, but still requires power sources. Batteries provide an alternative, but replacing them when they wear out can be expensive, especially when sensors are installed behind walls or in other similarly unreachable locations.
An important component to an ultra-
low-power Wireless Sensor Network (WSN) is efficient energy storage and management. Micro-power batteries, such as thin-film batteries, have greatly advanced in technology in recent years along with micro-power management solutions. Advances in ULP technology have replaced the need for AA or AAA batteries to much lower battery capacities and sizes. Hence small, flexible and “smart” wireless sensors with long battery life are a reality.
A new class of wireless sensors powered by harvested energy that does not need battery replacement is the latest in WSN technology for sensing and monitoring hard-to-reach environments and applications where energy can be harvested. Wireless sensors working on harvested energy have a set of needs that are more stringent than regular wireless sensors, such as low peak power, ultra-low standby current etc., apart from general low power consumption. This is a relatively new field within WSN and has wide-spread applications including medical, M2M, military and other research areas.
The technology and design considerations on the short-range radio transceiver play a key role in efficiency of such low power wireless sensors. The requirements on the transceiver to fit applications mentioned above can be categorized as shown in Figure 2.
Requirements for an ultra-low power radio transceiver.
The power supply requirement of the transceiver is a key factor in the wireless sensor design and application. Since most ULP sensors run from tiny batteries and energy harvesting sources, sub 2V supply voltages are highly desired as most sensors run out of a single cell depending on battery chemistry. Radio transceivers that work down to 1.1V give additional flexibility to sensor design and reduce power management constraints. The supply voltage, power amplifier (PA) energy consumption (at comparable range) and link data rate are often ignored when comparing different solutions. However, all three have a substantial impact. A radio operating at 2.5V consumes twice as much power as a radio with the same current consumption but operating at 1.25V. Operating at higher voltage is only required when output power in excess of 5 dBm is needed. This is not the case for short-range applications, as output power is rarely over 0 dBm. Low supply voltage is an easy way to reduce power consumption at the system level, but it requires an RF IC designed for low voltage operation.
Radios consume certain current when in transmit or receive operation. In real wireless networks radios seldom operate continuously, instead they operate in small time slots to save power and to not use up the radio spectrum. This is called duty-cycling and results in times when a radio consumes little or no power and other times when a radio is operating and consumes significant current. The absolute value of current when a radio is operating is the peak current and these instantaneous currents are often much higher than the reported average current of the system due to time-averaging. Radios that have high peak currents (even though average currents might be lower) can impose constraints on power management circuits, especially in applications that use harvested energy. Hence radios with low peak currents are desirable.
This constraint is even more important for wireless sensors that run from harvested energy sources. Often energy harvester transducers have higher output impedance than batteries. The micro-power management layer between the transducer and the sensor converts the supply characteristics including source impedance. Therefore, the low peak current consumption in the radio transceiver reduces constraints on the power supply of the wireless sensor.
For a radio transmitter, the power consumption of the PA can be very large. Many 802.15.4 or Bluetooth radios consume 25-40 mW for a 25-meter free-space range, wasting over 95 percent of it.
The principal parameter from a transmitter PA point of view comes from the receiver. Its sensitivity defines, for a given range, how much power must be radiated. Most radios fall into the -85 dBm to -95 dBm sensitivity range, resulting in a factor 10 in PA power consumption. The three main factors impacting power consumption are receiver sensitivity, carrier frequency and output impedance. They are additive, and together can represent over two orders of magnitude in PA power consumption variation for an identical range.
The choice of carrier frequency is also an important parameter for the transceiver. The two available options within industrial, scientific and medical (ISM) radio bands are 2.4 GHz or sub-GHz frequencies. Some of the factors to consider with this choice are:
• Power consumption
• Data rates
• Antenna size
• Interoperability (standards)
• Worldwide deployment
Wi-Fi, Bluetooth and ZigBee technologies are heavily marketed 2.4 GHz protocols used extensively in today’s markets. However, for low power and lower data rate applications, such as wireless sensors, wireless medical monitoring, home security/automation and smart appliances/metering, sub-GHz wireless systems offer several advantages, including longer range for given power, reduced power consumption and lower deployment and operating costs. Therefore sub-GHz carrier frequencies have certain advantages over 2.4 GHz in terms of range and signal quality.
As radio waves pass through walls and other obstacles, the signal weakens. Attenuation rates increase at higher frequencies, therefore the 2.4 GHz signal weakens faster than a sub-GHz signal. 2.4 GHz radio waves also fade more quickly than sub-GHz waves as they reflect off dense surfaces. In highly congested environments, the 2.4 GHz transmission can weaken rapidly, which adversely affects signal quality.
Even though radio waves travel in a straight line, they do bend when they hit a solid edge like the corner of a building for example. As frequencies decrease the angle of diffraction increases, allowing sub-GHz signals to bend farther around an obstacle, reducing the blocking effect.
The Friis Equation demonstrates the superior propagation characteristics of a sub-GHz radio, showing that path loss at 2.4 GHz is 8.5 dB higher than at 900 MHz. This translates into 2.67x longer range for a 900 MHz radio since range approximately doubles with every 6 dB increase in power. To match the range of a 900 MHz radio, a 2.4 GHz solution would need greater than 8.5 dB additional power.
Besides the need for higher power for the same link budget, the 2.4 GHz band has higher chances of interference. The airways are crowded with colliding 2.4 GHz signals from various sources, such as home and office Wi-Fi hubs, Bluetooth-enabled computer and cell phone peripherals and microwave ovens. This traffic jam of 2.4 GHz signals creates a lot of interference. Sub-GHz ISM bands are mostly used for proprietary low-duty-cycle links and are not as likely to interfere with each other. The quieter spectrum means easier transmissions and fewer retries, which is more efficient and saves battery power. Figure 3 shows an example of an architecture used for a smart appliance based on a ULP wireless sensor for connectivity where power consumption is critical.
Control & RF subsystem for smart appliances.
Semiconductor advances are allowing designers to deploy increasing amounts of control and connectivity options. The trend toward further field upgradable SoC platforms with integrated RF subsystems provides hardware and software architects a platform on which to co-develop system features and algorithms necessary to support the process of developing highly intelligent smart appliance and smart grid products.
Aliso Viejo, CA.