SNR improvement in MEMS electrothermal displacement sensors

Electrothermal displacement sensors are doped silicon resistors used for nanopositioning purposes. The displacement precision of the nanopositioners is strongly dependent on the signal to noise ratio (SNR) of the sensors. The sensitivity of the electrothermal sensor is proportional to the temperature of the sensor. However, higher temperature increases the thermal noise of the device. We propose new methods both in the device and the readout circuit to improve the SNR. The design of a multiple sensor architecture is reported in this work. Through experiments we show that with averaging multiple sensor signals the overall SNR is improved by the number of averaged sensor outputs. The experiments are carried out on SOI (Silicon on Insulator) MEMS (Microelectromechanical Systems) nanopositioners. By combining three sensor outputs, we achieved a 4 dB enhancement in SNR. Furthermore, the sensitivity of commonly used Wheatstone bridge and transimpedance amplifier (TA) readout circuits are investigated. We show analytically and experimentally that TA circuit offers four times higher sensitivity.