1. Open Loop Tuning (Process Reaction Curve): In this method, the controller is placed in "Manual" mode, effectively breaking the feedback loop. A step change is applied directly to the manipulated variable (e.g., opening a valve by 10%), and the resulting response of the controlled variable is recorded over time.
• The resulting graph is the "Process Reaction Curve."
• From this curve, parameters like the process gain, dead time, and time constant are calculated.
• These values are then used in the Ziegler-Nichols open-loop formulas to set the PID gains.
2. Closed Loop Tuning (Ultimate Gain Method): Methods like the Ziegler-Nichols Ultimate Gain (or Oscillatory) method require the system to remain in "Auto" mode. The integral and derivative actions are turned off, and the proportional gain is increased until the system reaches sustained oscillations. This is inherently a closed-loop procedure.
3. Other Methods: CHR (Chien, Hrones, and Reswick) and GA (Genetic Algorithms) are refinement or optimization techniques that can be applied to either type of data, but the "Process Reaction Curve" is the definitive "Open Loop" fundamental method taught in control theory.