Investigation of Power Saving Techniques in Signal Conditioning Circuits for Resistive Sensors

Abstract

In structural health monitoring (SHM), strain gauges are often used to sense important points of a system. However, as they are resistive sensors, a lot of current is required and that may reduce the autonomy of the device in charge of the SHM, when it is battery supplied. The aim of this project is to design an amplifying circuit for a resistive sensor (strain gauge) for structural analysis. More specifically, it will make up a part of a device with an analog-to-digital converter (ADC) module, a μC with a radio transmitter module and a memory. Also, it will have to be powered by a battery, also included in the device which makes energy saving vital due to the power consumption that these sensors can sometimes have. In the first place, a study has been done into the options available to make the battery last for a minimum of one year. As a consequence, it was decided not to have the system permanently active with a low voltage power, but to take samples at a certain frequency (1kHz due to bandwidth issues) and keep it in power-saving mode for the rest of the time. Once this was decided, a study into operational (OpAmp) and instrumentation (InAmp) amplifiers was carried out in order to, subsequently, choose two from each category that best adapt to the needs. In the end, AD8223 and INA827 were chosen as the instrumentation amplifiers and LT6230 and LTC6247 as operational amplifiers. Next, the circuits were chosen that later were tested to compare their results. More specifically, a circuit was made for each InAmp and two for each OpAmp: one with 2 amplifiers and one with 3. After designing and building a PCB board to test the circuits, it was measured, with the help of an oscilloscope, and the results from each one were taken, with special emphasis being given to settling time and the noise observed. Once the best option was found, and knowing the response time, the bandwidth of this circuit was reduced to minimize the noise produced at the output without slowing down the circuit. Finally, the board was connected to the wireless device where it will have to work in the future in order to check its correct working and data was collected during a few seconds