MM-IM User Guide
13
Note: The call to Master Control is effectively just an ‘interrupt’, it is not a data
transfer call
. Typically, to acknowledge the interrupt and find out the cause for the
alarm, Master Control would pre-empt its normal polling schedule and poll the calling
device immediately. It would then be able to retrieve the alert queue and forward any
alarms or events to the operators’ consoles in the normal way. Remember, without
this mechanism for pre-empting the poll schedule, it could be hours or even days
before the next scheduled poll goes out to the MM-IM – too long to wait if an alarm
condition has occurred.
Now the plant is monitored using plant inputs mapped to a combination of Medina points
of type AI, DI and PI. It follows that alarms and events must be supported on these points
and clearly the point type will determine the exact nature of the alarm or event. The
following alarms and events can be configured:
Each input of type AI can be configured with two pairs of high and
low alert setpoint levels - Hi, Lo and HiHi, LoLo, plus two Rate of
Change (RoC) alert setpoint levels - RoC Positive and RoC Negative.
Each input of type DI can be configured with one alert setpoint level
Each input of type CI can be configured with a Hi setpoint level
Any Medina input point (any AI/AID, any DI/DID or any PI/PID) can be independently
configured to raise either an Alarm or an Event against each of its configured setpoint
levels, or to ignore the setpoints (alerts disabled).
In addition to the alert setpoint levels so far discussed, measures such as level and time
deadbands can also be configured in many cases to reduce the incidences of false alarms
and hence reduce the overall alarm rate. This is an important factor in large installations
which can otherwise become flooded with very large numbers of duplicate alerts.
Full details of the various alerting functions supported by the MM-IM are presented below.
Analogue Inputs
HiHi, Hi, Lo and LoLo Level Alerts
Illustrated in Figure 1, each Medina AI point can be configured with two pairs of high and
low alert setpoint levels:
Historically, all transactions on a telemetry network must always be initiated by a single master (i.e. the Data Gatherer)
and addressed to a particular slave (using an address field embedded in the protocol). The master then waits for the polled
slave to recognise its own address and respond. Since only one poll goes out from the master at a time, and given that
only the addressed slave is allowed to respond to that poll, any possibility of conflict on the network is avoided.
To Next Page
Loading...