Background: I'm fairly new to signal processing and sensors, etc. I'm working on my thesis, which involves the use of an MEMS acceleration sensor as a vibration measuring device (for rotating machinery). I managed to extract the data from the sensor and save it to my computer. Now, I'm doing post-processing on the data with python. Basically I'm getting accelerations in X, Y ,Z axis and the time (milli seconds).
Problem: the signal is noisy and I'm getting strange results when I perform FFT to it.
I already did detrend (to remove the huge spike at 0 Hz) and used a window function (on the detrended data) to get a cleaner Fourier Transformation but still I'm not sure if what I'm getting is normal.
Above is data I get from the sensor. Blue: original data, Orange: after detrend, Green: after applying the window function. I applied the window function to all 3 but only showed it on the X
Above is the FFT of the data, with only unto 4 Hz on the frequency axis. Blue: FFT on detrended signal, Orange: FFT on the windowed signal
FFT of same data but full frequency axis
The measurements were taken while the sensor was not moving with a sampling rate of 1 kHz. Also, using the sensor's low pass filter of 184Hz.
Am I on the right path? Are these results to be expected? Or is there something wrong with my work?
I hope you guys can help me. Thank you in advance.