All switching is done without any bump on the loop controller outputs. Managing the slave loop operating mode management is identical to managing the process loop operating mode: this loop behaves as if it were alone.
There are, however, mechanisms specific to the master loop:
-
Manual mode and Auto/Manu switching are identical to those of the process loop.
-
If the loop controller runs in the Automatic mode (by default), two different situations may be observed:
-
The slave loop controller runs in Automatic mode and uses the Remote setpoint. The cascade is closed in this case, which means the master loop truly runs in Automatic mode.
-
The slave loop controller runs in Automatic mode and uses the local setpoint. Auto-tuning is running. The loop controller runs in Tracking mode. In this case, the master loop controller runs in Tracking mode.
The point of this procedure is to avoid any bump when cascade closes. You may encounter several situations:
-
If the slave loop controller uses the local setpoint, the master loop controller follows the slave’s local setpoint.
-
However, should the slave loop controller have an integral action, the master loop controller follows the slave loop process value.
-
Finally, should the slave loop controller not have an integral action, the slave loop controller must be either a P or a PD. The master loop output is calculated and optimized to prevent any bump when cascade close (depending on the controller output and its parameters).
When system is cold-started, the master loop controller always starts in Automatic mode. However, it is possible to re-configure the initial operating mode for the slave loop controller and setpoint type for each loop (Remote/Local).
Preventing Further Changes to Outputs
The Master loop characteristically presents one additional functionality: when the master is in Automatic mode and the slave output begins to saturate, the master loop "freezes" (prevents) any changes to the master output, in the direction which saturates the slave. This function limits saturation of the master’s integral action. Therefore, it is enabled only if the master loop controller has an integral action.
For instance, when the slave loop controller runs in Automatic mode, it is configured in reverse action and its output reaches the upper limit. In order to dissociate the output from its limit by acting on the setpoint, you must lower said setpoint. The master output is now slanted in a rising, growing direction.
Cascade Loop with a PID Loop Controller
The following schema shows a cascade loop with a PID loop controller.
A cascade loop generally consists of 2 process loops with some restrictions and additional functions.
The OUT_MAN master loop output is the remote setpoint for the slave loop setpoint branch. The OUT_MAN output is therefore expressed in the slave loop scale. It is subject to the slave loop Setpoint branch level limit.
The slave loop controller may be of one of the two following types: Split Range or Hot/Cold.
The Auto/Manual operating mode and the value for the master loop manual command are not accessible from the Control Expert process control screens. They may however be accessed via the user program.
Restrictions are as follows:
-
No totaling on the master loop process value branch.
-
No Feed forward branch on slave loop.
-
No ON/OFF loop controller on any loop.
-
The slave loop Setpoint branch is a single branch, with no scaling.
-
Only one model loop controller may be configured on the master or slave loop.
