Use Energy Saving Ideas When Designing ICs

From building automation to factory floors, companies are developing “smart” solutions to conserve energy. Many of these solutions are the result of very creative thinking on the part of integrated-circuit (IC) designers. Deep application understanding, breakthrough intellectual property (IP), and efforts to trim wasted power all contribute in major ways to more responsible energy management. Although applications of today’s ICs vary in function and purpose, they all follow a common theme: Use only the power that’s needed for the required functionality.

For example, take the many applications, such as alarms and security sensors, which need to operate for years on a single battery. Understanding the specifications of available batteries, the acceptable battery swap-out rate, the security level, and the monitoring frequency requirement can result in a wiser decision on the appropriate intelligence layer to add to the application’s circuitry. Energy-conservation issues can be resolved by carefully defining the device’s logic capability—controlling what the device will or will not do and wisely establishing power domains.

Typically, choices include integrating a state machine with well-defined functions and/or embedding a microprocessor that can flexibly handle a lot of “what-if” inputs and scenarios. Depending on the specific application, either option can define acceptable standby modes, the timing of power-hungry activities, and operation-mode selection. Such flexibility puts a lot of power control into the hands of the designer and user alike—including defining the boundaries of the application’s required functionality.

From another perspective, further functional integration can include the elimination of inefficient hardware and systems that served certain roles in the application. For instance, simple communication from an IC to an industrial network can be further leveraged with additional capabilities in monitoring, controlling, and billing (e.g., efficiently invoicing end customers for energy usage). In effect, those previous subsystem hardware and software components could be re-engineered or eliminated altogether, putting their task burdens on the lower-cost silicon IC.

By sufficiently understanding the application’s environment dynamics, complementary blocks can be added to the IC to eliminate inefficiency. For example, a high/low-output power mode cell can be included with an on-chip transmitter. The user can then choose the most efficient setting for reliable transmission outside the chip.

Breakthrough Intellectual Property

Embedded cells on silicon constitute a major part of a company’s IP and product-portfolio improvement efforts. Continual R&D and other IP-acquisition activities put a company on the path toward discovering revolutionary solutions that can enhance energy efficiency and functionality. An example is ON Semiconductor, which invented a circuit and operating mode for its RF products that would specifically address power-consumption issues during a duty-cycled receive operation.

Rather than waiting for a lengthy 500-μs to 1-ms industrystandard receive wake-up time, the “Quick Start” oscillator was developed to provide a very fast RF crystal startup time. At just 15 μs, the crystal starts up. Within just 130 μs, the integrated receiver is ready to receive or check for the presence of an RF signal. The result was a more than 75% reduction in wait time, which directly multiplies the life of a low-power application battery.

As a companion to the oscillator innovation discussed above, the company developed the Sniff Mode programmable duty cycle. Here, low-end, dedicated, embedded microprocessor logic was used to control physical IP, thereby lowering power consumption in both the chip and the external peripherals that it drives. Programming the duty cycle assists in optimizing power conservation and overall system performance, as it shuts down non-critical features during “off-time.”

Trimming Waste

Obviously, an inefficient system wastes energy. The same holds true for ICs. One example is in adjustable impedance matching. By developing an on-chip EEPROM array that stores register settings, these values can trim a low-noise amplifier’s impedance match to an externally connected RF receiver antenna for improved sensitivity.

In addition, physical objects like walls, equipment, and storage tanks can cause RF transmission interference. If these objects are mobile or if the transceiver often changes location, maintaining clear transmission can be quite challenging. On the other hand, boosting power to a fixed transmit level for the highest interference scenario causes waste when there are fewer obstacles present. To resolve this dilemma, an IC can be programmed with an adjustable turnon energy threshold that automatically and dynamically enables the part to discern a good signal from noise. (Sniff Mode incorporates this feature.)

Rather than boosting a transmitted RF signal high enough to get a message completely through the data channel, one can try enhancing the receiver sensitivity or reliability. The information can then be accessed more readily with the same signal strength. One suggestion in the RF world is to incorporate a “clock and data recovery” circuit, which synchronizes the data-processing clock with the incoming data. Another example would be to implement antenna spatial diversity by using two independent receive channels. To avoid wasting power, a two-channel receive system can reduce the need to re-transmit the same message.

Simply choosing an IC over a discrete alternative can save on power consumption. An inherent feature of an IC is its small physical size, as it is composed of even smaller cells or blocks that perform the same functions as their discrete counterparts. Yet they can provide less heat loss and more efficient and reliable inter-block operation. With an IC, fewer external connections are usually required. As a result, overall power loss is reduced.

With all else being equal, such as operational speed and functionality, smaller silicon geometries tend to require lower currents to perform the same circuit tasks. In manufacturing operation environments, a small circuitry footprint also tends to enable higher operational efficiencies and less material consumption during printed-circuit-board (PCB) production (e.g., making the PCBs smaller).

A power budget is an ever-present constraint. To conserve energy, a chip designer can leverage the existing energy supply by fully understanding and dealing with the application requirements. Alternatively, he or she can innovate new circuits that intelligently take power conservation to the next level. Another approach is to simply trim out the waste where possible.

Paul Pulley is a product manager from ON Semiconductor’s Custom Industrial Products business unit. He holds an MBA and BSEE and is PDMA certified as a new product development professional. Pulley has worked for ON Semiconductor for eight years.