User guide

The Serial BERT allows you to run tests from 150 Mb/s up to 12.5Gb/s
(depending on the options with which you bought the instrument). See “Error
Detector Bit Rate” on page 179 for more information.
• Does your device require differential inputs?
As complementary outputs, Data/Clock Out and Data/Clock Out may satisfy
this requirement. See “Diagram 2. Connections for Differential Inputs” on page
30.
• Is the output of your device noisy or very small in amplitude?
You may need to manually set up a sampling point. See “Sampling Point Setup
- Concepts ” on page 193.
• Do you want to check the BER in bursts (for example, for loop testing
applications)?
The Serial BERT provides capabilities that help you set up and run burst tests.
See “Introduction to Burst Sync Mode” on page 211.
• Are you aware of how the sync mode setting can affect the way errors are
measured and displayed?
See “What Type of Synchronization Should You Use?” on page 208.
Test Times and Confidence Levels
A true BER measurement must be statistically valid. Because it is not possible to
predict with certainty when errors will occur, your device must be tested long
enough to have confidence in its BER performance. The table below lists the test
times required for different BERs at specific bit rates.
N O T E
The test times in this table are valid for continuous measurements. If measurements
are made in bursts, the test times will be longer (continuous test time divided by
the burst duty cycle times the burst sync ratio). See “Introduction to Burst Sync
Mode” on page 211 for background information.
The test times for 95% confidence level are:
Table 3
BER STM-64/
OC-192
(9.95328 Gb/
s)
STM-16c/
OC-48c
(2.48832 Gb/
s)
STM-4c/
OC-12c
(622.08 Mb/s)
STM-1/OC-3
(155.52 Mb/s)
1E-14 ~ 8.4 hours ~ 1.4 days ~ 5.6 days ~ 22.4 days
1E-13 ~ 50 minutes ~ 3.3 hours ~ 13 hours ~ 2.2 days
Planning the Test 1
Agilent J-BERT N4903B High-Performance Serial BERT 27