Agilent Technologies J-BERT N4903, Руководство пользователя

Страница: 179 / 434

Quite often, we do not need to measure the exact BER, but can stop the
measurement if we are sure that the BER is above or below a threshold.
In a jitter tolerance test, for example, we need just to assure that the
device under test operates with a BER better than let us say 10
-12
;
whether the true BER is 1.1 × 10
-13
or 2.7 × 10
-15
is irrelevant.
To abort the measurement for a single point and proceed to the next,
we need two limits that tell us whether the BER is above or below the
given threshold.
These limits have been calculated from the error probability density
functions applicable to BER measurements. The equations were solved
for a level of confidence of 95%. The calculations have led to the
following table:
Table 20
95% confidence level lower limits, BER
> 10
-12
95% confidence level upper limits, BER
< 10
-12
Min number of
errors Max number of
compared bits (×
10
12
)
Max number of
errors Min number of
compared bits (×
10
12
)
1 0.05129 0 2.996
2 0.3554 1 4.744
3 0.8117 2 6.296
4 1.366 3 7.754
5 1.970 4 9.154
6 2.613 5 10.51
7 3.285 6 11.84
The BER threshold of 10
-12
and hence the setting of the multiplication
factor to 10
12
is just an example-by changing the exponent, the table
applies analog to other thresholds.
The following figure shows a plot of this table.
Advanced Analysis
5
Agilent J-BERT N4903 High-Performance Serial BERT 179
The Uncertainty Band