Figure 4.

The noise performance of the MEMS gravimeter over a period of 54 h. (a) Shows the normalised temperature variation of the shield, LED and Photodiode over this period. These temperatures were stabilised using a proportional integral derivative (PID) control mechanism. (b) Shows the tilt of two orthogonal axes of the device. The large variation of axis 2 was caused by the sinking of the device into the linoleum of the laboratory floor. (c) Shows the acceleration output of the device over this period. A multiple linear regression has been performed on this data to remove the effect of the tilt and a residual linear drift. It is not expected that the Earth tides could be observed with this noise level, but if a lower frequency device were inserted into the sensor then such an observation could be made.