AnsweredAssumed Answered

ADXL372 peak noise for FFT

Question asked by naohisa on Apr 19, 2018
Latest reply on Apr 19, 2018 by naohisa

Hi,

 

I am experimenting with ADXL372.

Since there are questions about measurement data, please let me.

 

Question:

 

Although I look at the output (x, y, z) of acceleration data using FFT,

peak noise occurs at a specific frequency.

 

Detailed setting and measurement result are below.

ODR = 3200Hz

BANDWIDTH = 1600Hz

FIFO = OFF

LPF = DISABLE

HPF = DISABLE

DATA_READY = ON

MODE = Measurement Mode

 

 

As a result of consideration, 

It turned out that peak noise occurred at frequencies of ODR/5 and 2*ODR/5.

Since ODR is 3200 Hz this time, it can be confirmed that peak noise occurs at 640 Hz and 1280 Hz.

 

For the result,

we experimented with changing the measurement mode and filter(1600Hz, 800Hz, 400Hz...),

but peak noise occurs at the same frequency.

 

The object to be measured was a vibration testing machine,

trying to set the mounting position is horizontally or sideways.

The recall rate of the phenomenon is 100%.

 

By the way, the same phenomenon was confirmed also in the experiment with the evaluation board (EVAL-ADXL372Z).

It is the result of FFT analysis (x, y, z) when ODR is set at 6400 Hz.

 

Even in the experimental results on the evaluation board,

it was confirmed that peak noise occurs at frequencies of ODR/5(1280 Hz) and 2*ODR/5(2560 Hz).

I think that this is the characteristic of the sensor itself, but is the recognition correct?

 

Regards,

Naohisa

Outcomes