CS5571

24

DS768PP1

3/25/08

10:56

Figure 20 illustrates a noise histogram of 4096 samples with the input signal adjusted to almost the exact

center of a code with DITHER on. Figure 21 illustrates a noise histogram of 4096 samples with the input

signal at the center of a code with DITHER off.

Notice that with a signal at the center of a code that the converter outputs the same code over 96% of the

time. Figures 22 and 23 illustrate the noise histogram, DITHER on and then DITHER off with the input

signal at a code boundary. Notice that in the DITHER off case the converter only exhibits two codes of

noise.

3.9 DITHER

From the performance plots, one should conclude that the best AC performance for small signals occurs

with DITHER on. For capturing multiple samples and performing averaging, DITHER should also be on

because the dither will randomize the quantization noise of the converter and provide improved accuracy.

However, if only one conversion is to be taken on a DC input, DITHER should be set to off. With DITHER

off, the converter exhibits only two codes of noise if the signal is at any point other than the exact center

of a code. This means that with DITHER off the converter will nominally yield over 32,000 noise-free

counts.

Output (Codes)

3

2624

9

0

500

1000

1500

2000

2500

3000

3500

4000

4500

-2

-1

0

Output (Codes)

12

751

709

0

500

1000

1500

2000

2500

3000

3500

4000

4500

75

3940

81

00

-2

-1

0

Output (Codes)

12

Figure 20. Noise Histogram, 4096 Samples

Dither On, Code Center

Figure 21. Noise Histogram, 4096 Samples

Dither Off, Code Center

58

0

1953

66

2019

0

500

1000

1500

2000

2500

3000

3500

4000

4500

-2

-1

0

Output (Codes)

12

Output (Codes)

2050

2046

-2

-1

0

1

2

0

500

1000

1500

2000

2500

3000

3500

4000

4500

00

0

Figure 22. Noise Histogram, 4096 Samples

Dither On, Input at Code Boundary

Figure 23. Noise Histogram, 4096

Samples

Dither Off,