GOLDENEAGLE AM 4.0.x user manual
– 07/2012
Page 18
Head Office
: Parc d’activites Kennedy - 20, avenue Neil Armstrong – F-33700 Bordeaux-Merignac (France)
Tel +33 (0)5 57 928 928
– Fax +33 (0)5 57 928 929 –
–
www.audemat.com
5. REMOTE APPLICATION PRESENTATION
5.1.
Introduction
5.1.1. Working principle
The unit is designed to automatically monitor a programmed list (radio stations) by comparing measurements
from an internal tuner to user-defined reference settings. In case of failure, alarms are generated and sent by
SNMP Traps or mail. The unit communicates with the outside world via IP network (Ethernet or modem). It
comes with a client-server application serving as graphic user interface that makes it possible to remotely
configure the unit, view measurements and event logs as well as view real time measurements and listen to
the audio.
5.1.2. Functions
Cyclical monitoring on multiple programs
Cyclical monitoring is done by setting the tuner successively on each set frequency for acquisition of
measurements and by then comparing the programmed reference values with the newly acquired data.
Please note that measurements are collected only once per frequency which explains why a very brief
anomaly will not be detected. Each monitoring element is successively tested according to the hierarchical
test order. With this order; lower level tests will stop when a higher level error has been detected. This
method prevents useless alarms from being sent.
Measured Value
Monitoring Parameters
Alarms
Values
Recorded
RF Level
-
Nominal level
-
Low threshold 1
-
Low threshold 2
-
High threshold
-
hysteresis
-
timeout
-
short intermittent errors
RFLOW2
RFLOW1
RFHIGH
yes
AF Level
-
Nominal level
-
Low threshold
-
High threshold
-
hysteresis
-
timeout
AFLOW
AFHIGH
yes
Triggering alarms:
As monitoring starts, each test is set to
‘NO_ERROR’ status.
This status is then reevaluated with each monitoring cycle as the tests are performed again. When a
measured value is out-of-tolerance, the state switches to
‘ERROR’, but the alarm is not triggered
immediately. For each parameter
, the “timeout before alarm” parameter can be configured: it represents a
delay during which an error state is tolerated before the alarm is triggered. Once this length of time has been
r
eached, the state switches to ‘ALARM’, and corresponding actions are then taken (notification, placement of
the
event in the event log…)
Switching from ‘ERROR’ to ‘NO_ERROR’ is immediate; however, it takes 2 confirmations to go from
‘ALARM’ to ‘NO_ERROR’, the test must be error free for 2 consecutive cycles.
The hysteresis is an additional factor to be taken into account before the unit returns to normal state after a
‘level’ type threshold violation. This value adjusts the triggering threshold after an alarm begins so that
fluctuations near the threshold level will not continuously trigger alarms. For instance, if we set the lower
level threshold at 40, and the hysteresis at 5, the alarm activates at a level lower than 40, and stays on until
the level rises above 45.
