Filter
Top Products
P300H P300 Series Modem Installation and Operating Handbook Page 129
BERT:Sync OK Loss#:0 Rate:32000bps
Errors:328 Time:1.0min BER:1.7E-04
Test, Internal BERT, Results Menu
6.16.8 Test, Int' BERT, RESULTS Menu
Note in this screen the following keys are active:
Press 1 Inject single error (buffered for up to 10 rapid key presses)
Press 2 Start a test / Restart the current test (manual mode only)
Press 3 Stop the current test (manual mode only)
The fields on this screen are as follows:
BERT: “Sync OK” The test is running and the tester is `in sync`
“NO SYNC” The test is running but the tester is `out of sync`
“Reset” No test has been run since the unit was (re)configured
“Stopped” The test has been manually stopped (results are frozen)
Loss#: A count to 99 sync losses and recoveries. When the display reaches “>99" the
test is automatically stopped. The sync loss threshold is fixed at >25% errors over
1024 bits (equivalent to the Firebirds `Normal` threshold)
Rate: The test data rate. This accurately displays the rate at which the test is running
whether it is in the Main, ESC or Aux channels.
Errors: The cumulative count of errors during the test. If the count reaches 999,999 then
the test will stop and the results will be frozen. See Note 1 below.
Time: The cumulative test time, displayed as seconds to 1 minute, then decimal minutes
to 1 hour, then decimal hours to a maximum of 27.7 hours.
BER: The calculated BER over the entire duration of the test (ie average over test
period). It is displayed as follows:
When there are more than 100 errors the exact BER is displayed.
When there are less than 100 errors a
symbol prefixes the BER display
(approximately equal to, as the result is statistically poor).
When there have been NO errors the modem does not know if the BER is
perfect, or simply if the test has not been running long enough to receive an
error. In such a case it prefixes the display with a < symbol, and displays the
BER that it would calculate if ONE error had been received. This means that
as the test progresses without errors the displayed BER improves, but never
assumes it is zero (as a Firebird does).
Note 1:
The BER displayed by the Internal BERT is 100% accurate. However if you use the Internal BERT and in parallel monitor the same
received PRBS with an external Firebird, there will be a <0.1% discrepancy in the error count, but NOT the BER. For the purists amongst
you, and so you do not question the accuracy of the Internal BERT, the briefest useful explanation possible is as follows:
When the processor reads the BER counters once per second (errors, number of bits etc), it has to freeze the counts in order to read the
results accurately (in case the counters increment whilst being read). This reading of the counters takes less than 0.1% of the time, but
does mean that the BERT is `blind` for a small proportion of the time and some individual errors may not be counted. However the BER
is 100% accurate because as well as counting Errors, the BERT also counts PRBS bits received, and calculates the BER as <Errors> /