6. Redundancy with NX3030 CPU
311
Redundancy %Q
memory offset
reserved for
diagnostics
%Q redundant memory
offset reserved for
diagnostics initial
address
Redundancy %Q
memory length
reserved for
diagnostics
%Q redundant memory
offset reserved for
diagnostics size
Table 6-3. NX4010 parameters
I/O Drivers Configuration
The configuration of I/O drivers, at first, isn’t different in relation to a non-redundant CPU.
What can be observed is that some I/O drivers have commands which allow its use in a redundant
CPU, but it doesn’t imply in configuration differences. These commands, normally, must be executed
in the NonSkippedPrg program. E.g. a MODBUS RTU master driver in a RS-485 serial network
must be disabled in a non-Active CPU using the code inserted by the user in NonSkippedPrg. More
information regarding administration of MODBUS driver in a redundant system can be found in the
MODBUS Instances Managing in Redundant System section.
In the case of PROFIBUS network, there are also special different commands for the CPUs in Active
and Non-Active states. In this case, however, the redundancy management executes such commands
automatically, without any user management.
To configure PROFIBUS I/O remotes, including remotes and I/O modules, see NX5001 Modules
Configuration section from this manual.
MainTask Configuration
The configuration screen associated to the only task of a redundant CPU, called MainTask, which is
cyclic, can be accessed through a click on the MainTask in the Device Tree.
Two parameters must be adjusted on this screen:
ï‚· The MainTask cycle time
ï‚· Watchdog time
Furthermore, the screen shows an estimative of the necessary time to manage the redundancy,
calculated by MasterTool. Such estimative is only reliable after the project is complete, with all
POUs developed and redundant memory areas defined.
Several considerations must be taken in order to adjust correctly the MainTask cycle time:
ï‚· The cycle time must be sufficiently low to allow the proper process control, taken in account all
control feedback times
ï‚· The cycle time must be high enough for allowing, at least, the sum of the following times:
o The NonSkippedPrg and ActivePrg POUs maximum execution time, together
o The necessary time to manage the redundancy (redundancy overhead)
ï‚· Besides this, the cycle time must have an additional looseness necessary for the other processes
execution times (PROFIBUS communication, Ethernet communication with SCADA systems,
etc.)
MasterTool has conditions of calculating the necessary time for redundancy management
(redundancy overhead), after the project is finished (all developed POUs and redundant memory
areas defined).
Regarding the NonSkippedPrg and ActivePrg POUs execution maximum time, they are possible to
be measured after these POUs are already developed. Initially, MasterTool estimates 10ms for these
two POUs maximum time, together, but the user must revise this field afterwards, when measuring
using the final project.