CDBCapture+

DS327F1

5

Using the same sine wave acquired data, a Fre-

quency Domain Analysis can be performed and is

shown in Figure 4, below. The Frequency Domain

Analysis is performed by selecting "FFT" under

the TEST window. The results of this analysis are

displayed to the left and bottom of the graph. The

graph shows the magnitude vs. frequency spectrum

from DC to one half the sampling frequency, with

the number of frequency bins equal to one half the

number of samples taken. The fundamental and

first four harmonics are defined by numbers on the

graph while the fundamental frequency and its

magnitude are shown below the graph. Four basic

statistics relating to the FFT are shown to the left of

the graph, these being Signal to Peak Noise (S/PN),

Signal to Noise and Distortion (SINAD), Signal to

Distortion (S/D), and Signal to Noise Ratio (SNR).

The software allows the user to select the number

of samples (binary numbers) to be used in acquir-

ing the data for the FFT. Note: if the number of

samples selected is too small or if the input fre-

quency is set too low, it may not be possible for the

FFT to discern the fundamental from the DC bins.

In this case the software will display the following

recommendation: "The test signal frequency falls

within the DC bins and is therefore indistinguish-

able. Either increase the number of samples or in-

crease the frequency of the test signal.

Then

acquire and try the FFT test again."

The software offers the user the option of perform-

ing a FFT frequency analysis without having to ap-

ply a sine wave signal input. By grounding the

input and telling the computer what the maximum

A/D count is for a full scale signal, the software

will sample the grounded input and assume a full

scale signal was applied, providing a theoretical

Signal to Noise Ratio based on the internal noise of

the system and converter.

The display for a

grounded input FFT is shown in Figure 5.

Figure 4. Frequency Domain Analysis