06-25-2015 12:24 AM - edited 06-25-2015 12:42 AM
Your data has a gigantic DC offset, meaning that most of the result will be at zero frequency. Set the y-axis of your graph from 0..0.1 and you'll see the other frequencies. Alternatively, you could set the mapping of your Y axis to logarithmic.
Your data also seems to be in segments different offsets and contains some random spikes. Maybe you should analyze it in segments. Also, your data variation is relatively small compared to the digital resolution, meaning the data is quantized to relatively few levels.
Maybe you could record the data with the offset removed and with more gain, appropriate to the data range.
What does the data represent and what are you trying to get out of it. Just blindly throwing an FFT at something does not magically solve things. What's the reason for the transform?
06-25-2015 12:50 AM
Thanks very much for your help..
So, my block diagram is no problem, and the problem is the data, right? Okay..
Actually, i want to compare my fft result from Senscope ( software from the sensor vendor ) with labview. The data come from vibration sensor. I get those data by shaking the sensor ( ???) . Maybe you have the similar data , so i could try on my own?
NB: Hello everyone, i need real sensor data in excel format . If anyone has, please share to me...
06-25-2015 01:46 AM
i try with another data, and success !!! ![]()
Thankyou