I have built a testing setup that uses EVAL-ADXL001 as a calibration for the acceleration applied. A general view of the setup is shown in the figure below. Basically it is constructed of a speaker that serves as a shaker. A plastic fixture (white plate and holder in the picture) that holds a metal (brown) box is fixed to the speaker. The metal box contains the EVAL-ADXL001 with input BNC for powering the accelerometer and for viewing the signal on the scope.
The acceleration I would like to be able to measure are around 0.5-1G. The problem is that the noise level is still relatively high. I am using a 22nf for C2 and 300Ohm for R1 in order to remove high frequencies noise. In addition I am using batteries to power the device (They are shown in the picture above).
The results below are what I measure. The signal is for sin vibration at 50Hz. The PTP amplitude is 50mV. With 4.8V powering the device this is about 2G. The noise PTP is about 20mV.
If I use averaging, that is offered by the scope, the noise is reduced as shown in the figure below.
Here the PTP is 5mV, that is about 0.2G. This is nice but I rather not use average to get the results I need.
Any thoughts how to reduce the noise are welcomed!
Maybe placing the batteries in a faraday cage. I saw power sources with noise as low as 4uV. would this help?