Microchip Technology’s dsPIC33FJ12GP family represents the world’s smallest digital signal controllers (DSCs), and is therefore well suited for a new class of smart-sensor applications in which digital techniques enhance the performance of a variety of electronic sensors. One intuitive way to understand the impact of smart sensors might be related to the argument for continuing to develop piloted aircraft, even though the technology exists for unmanned aerial vehicles (UAVs). Specifically, a human pilot interpreting and communicating information can have a greater tactical benefit than a UAV that simply collects and transmits massive amounts of reconnaissance data.

Article Tools

• 

In a similar way, smart sensor technology uses a DSC to process raw data from simple sensors into a form more easily manipulated by the host processor. This not only promises to improve the precision, accuracy and reliability of the data collected by the sensor, but also promises to improve the general performance of the host system.

Leonard Elevich, principal applications engineer with Microchip, states there are two main reasons for this system-level performance improvement. First, because DSCs feature configurable hardware, they can execute signal-processing algorithms much faster than a conventional processor or MCU. Second, using DSCs to perform signal processing offloads this same task from the host system’s central processor, enabling the latter to service other tasks more quickly.

The signal processing algorithms for sensors fall within two general categories: those that improve performance and those that improve reliability. To improve performance of an individual sensor, a DSC might actually process information from another sensor. For example, if high accuracy is required from a photodiode in an unregulated thermal environment, it will be necessary to implement some sort of temperature compensation scheme. A simple reading from a thermistor coupled to the photodiode could provide the temperature data needed for the algorithm.

Figure. The small form factor of the dsPIC33FJ12GP DSC family enables digital signal processing to be applied directly to the output of a low-cost sensor.

Another example of improving sensor performance involves the elimination of signal noise. This noise may originate in the environment surrounding the sensor, or it may originate within the sensor itself. For example, the DSCs can be used to implement 50 Hz/60 Hz notch filters to eliminate power line noise.