Microcontroller
From Capsil Wiki
Overview
The heart of any wireless sensor node is its microcontroller MCU. It provides the computational capability to the sensor, but it differs from standard central processing units CPUs in its focus on self-sufficiency, power efficiency, and low cost. Two of the most commonly utilized MCUs are Atmel's 8-bit range processors and Texas Instruments MSP 430 family of 16-bit RISC processors. Atmel's MCUs are based on a Harvard architecture -- programs and data stored separately -- with a RISC instruction set. They have clock speeds up to 20 MHz with 20 MIPS. However, many applications run the MCUs at lower speeds in order to achieve lower power consumption. They also feature full integration onto a single die of the Flash, EEPROM and SRAM. The architecture is optimized for high level programming languages, especially C.
Atmel MCUs have been successfully used a wide on wireless sensor platforms including DSYS25, MICAz, BTnode, and Tyndall Mote. Atmel's range of 32 RISC MCU's have also been utilized in sensors. The AMAT91 which is based on the ARM7TDMI? processor core was used by University of Edinburgh in the ProSpeckz sensor.
Texas Instrument's MSP430 CPU family uses a von Neumann architecture -- ccommon memory space in which both program instructions and data are stored -- with various memory and peripheral configurations. The MSP430 is designed specifically for ultra-low-power applications, using a flexible clocking system and wide variety of operating modes designed to reduce power consumption, thereby extending battery life. The current draw when in sleep/power down mode is 0.1-μA, 0.8-μA in standby and 250-μA in operating mode when running at 3V. The MSP430 can operate down to 1.8V, further improving a superior power specification. The MSP430 is used on the SHIMMER, Telos, BSN node and Tmote wireless sensor nodes among others.
