jump to search box

Dynamic offset cancellation for MEMS accelerometers [electronic resource] / Pedram Lajevardi.

Availability

Online

At the Library

Other libraries

Author/Creator:
Lajevardi, Pedram.
Language:
English.
Publication date:
2012
Imprint:
2012.
Format:
  • Book, Thesis
  • 1 online resource.
Note:
Submitted to the Department of Electrical Engineering.
Note:
Thesis (Ph.D.)--Stanford University, 2012.
Summary:
Today's Micro-Electromechanical Systems (MEMS) accelerometers suffer from input offsets that drift with temperature and time. These devices are calibrated once after fabrication in order to cancel their offset. In order to cancel the offset drift over temperature, the temperature is swept during the post-fabrication calibration and the offset is recorded as a function of temperature. However, the residual offset drifts over the life-time of the device, and cannot be cancelled by a one-time factory calibration. With the continuously increasing precision requirements of accelerometers, this offset drift is emerging as an issue. This dissertation presents a new approach that dynamically measures and cancels a major part of the offset drift that is due to change in parasitic capacitance of the bondwires in a system-in-package-type MEMS accelerometer. This approach is based on modulation of the spring constant of the sensor element by applying a modulating electrostatic force. A prototype interface IC was fabricated in a 0.18-[mu]m 3-V CMOS technology, and was packaged and tested with a MEMS sensor element. The CMOS readout dissipates 3.1 mW, and has a noise-floor of 220 [mu]g/√Hz. The bandwidth of the interface is 1 kHz with a usable bandwidth of 200 Hz. The full-scale range is 9.14 g. The proposed scheme reduces the bondwire offset of a prototype by a factor of 112 (41dB).
Contributor:
Murmann, Boris, primary advisor.
Howe, Roger, advisor.
Wooley, Bruce A., 1943- advisor.
Stanford University. Department of Electrical Engineering

powered by Blacklight
© Stanford University. Stanford, California 94305. (650) 725-1064. Terms of Use | Copyright Complaints | Opt Out of Analytics
jump to top