To help in the detection of a communication failure between OPC servers, a watchdog timer can be created to periodically receive data through OPC Mirror and signal an alarm if it does not arrive.
To set up a watchdog timer for OPC Mirror
From the OPC Mirror installation area, copy the file OPC_WDOG.fhx into the \DVData\ImportExport folder. The default location for the OPC Mirror installation area is C:\Program Files\FRS\OPC Mirror.
Import the OPC_WDOG.fhx file into the DeltaV configuration database using the standard import procedure.
If a warning about there being no SCHEMA block or the controller not existing appears, you can ignore these messages.
This creates a control module named OPC_WDOG in plant area AREA_A. The module’s function block diagram contains a comment that explains how to use it. A copy of those instructions is presented below.
To use the watchdog timer control module
This module produces an active alarm condition when a chosen parameter has not changed value for a configurable period of time. Presumably there is an OPC client that is writing values to parameters in this module or some other module.
Choose a parameter that is expected to change each time the client application writes that parameter. If there is no parameter guaranteed to change within a known period, have the client write a parameter dedicated for this purpose, say a counter which increments each time it is written.
The parameter named HEARTBEAT in this module is an external reference parameter. Configure the path for this parameter to be the parameter you expect to change each time it is written.
Select the OND1 block, then change the TIME_DURATION parameter (seconds) to a value longer than the time the heartbeat parameter is expected to change within.
The alarm parameter COMM_ALM becomes active if the timer times out because the heartbeat parameter has not been written for time duration.
The protocols used by OPC are inherently non-deterministic. You must configure the loss of communication alarms and watchdog timers in order to be alerted when communication failures between servers occur.