13/40

SOFTWARE TECHNIQUES FOR COMPENSATING ST7 ADC ERRORS

2.3 HISTOGRAM TECHNIQUE

A histogram arranges several conversion results in increasing order and calculates the

number of occurrences of each value. The technique is to discard the least occurring values.

We assume that the least occurring values are the effect of a momentary disturbance in the

application.

Figure 6. Overview of Histogram technique

Example

Figure 6 shows an example histogram. It gives the number of occurrences of a conversion

versions. Each output value is arranged in a bar chart showing the number of occurrences.

(for example value=409 received 12 times out of 32 conversions).

The principle is to reject the values with the smallest number of occurrences, assuming them

to be result of noise etc.

This technique has limitations in practical usage. The disadvantage is more RAM requirement

because all the results are stored in an array. It also requires lot of time for post processing the

data. The processing of data involves finding the total number of occurrences of any value and

then discarding the data which occurred less than a specified number of time.

Number of

Conversions

Digital Output

409

410

411

412

408

3

1

12

11

5

Rejected values