I was measuring a 10, 20, 40 and 60 Hz wave by putting the ADXL355 on a shaker.
The signal was also measured by a different accelerometer of higher reliability and showed the expected frequency in the fft. The ADXL355 however showed frequencies in the fft, that are off from what I expected. The data from the ADXL355 was around 37% higher compared to the true data, so when I applied a 10 Hz wave to the Accelerometer, I got 13.3 Hz as a response. When I applied 40 Hz I got 55.31 Hz as a response.
During the experiment I was reading out the FIFO and didn't get any FIFO_FULL. I was using the Accelerometer with 4 kHz ODR, so according to this my sampling frequency was 4 kHz as well (is this assumption correct?). I set the highpassfilter on 247e-3 Hz.
I have attached the ffts of the ADXL355 data and the data fromt the 2nd accelerator.
What could cause such an error?