I’m using Arduino 33 Sense IMU to measure vibrations and I have tweaked it up to about 400Hz. However, the object (motor) I’m measuring can have a vibration frequency of up to 10kHz.
My question is; how does the frequency limitation on the IMU affect the measurement? Will the m/s^2 be correct for a e.g. a second or will I miss out on data?
I’m having hard time testing this due to I don’t have a test equipment.
That particular product uses an LSM9DS1 MEMS IMU.
When you set the frequency, you are setting the frequency cutoff of the internal anti-aliasing filter.
There should be minimal aliasing, but the bandwidth will be limited to the 408Hz, assuming you have set the ODR (Output Data Rate) and filter to the maximum.
If you increase the ODR but set the anti-aliasing filter to a lower number, the output will roll off at a lower frequency than 408Hz.
If you have a large amount of high frequency vibration (which seems unlikely in most situations) it may come across as greatly attenuated aliased signal and/or a DC shift.
If you have no idea whatsoever what signal there is, it's probably a good idea to somehow borrow something (perhaps a calibrated piezo sensor) that will tell you what is there in the high-frequency part of the spectrum. It's possible a microphone could be of help if you can't do that. You'll need to sample at typical audio sampling frequencies, 48kHz is good, but definitely well over 20kHz to get 10kHz bandwidth.
You will alias. When you sample a signal, you must sample faster than twice the highest frequency content in that signal. Otherwise, the higher frequencies will sort of disguise themselves as lower frequencies.
