You can tune a process controller, but you can't tuna fish. To efficiently manage their controllers, however, many users need some assistance.

To help calculate tuning parameters, Control Engineering senior editor Dave Harrold recently compiled "Process Controller Tuning Guidelines" to serve as an introduction or refresher for users.

For the complete "Process Controller Tuning Guidelines," visit the Online Extra section at www.controleng.com. Recent articles in Control Engineering on improving control loop performance include: "Turn Problem Loops into Performing Loops," Feb. '99, p. 77; "Assessing Control Loop Performance" May '99, p. 91; and "Open-Loop Response Testing Improves Process," May '99, p. 99. These are also available online.

While calculating tuning values requires extensive knowledge of a specific controller's algorithms, there are three basic algorithm types: ideal, parallel, and series-the most common algorithm in analog and digital controllers.

While technical volumes contain many tuning methods, software from Techmation (Scottsdale, Ariz.) and ExperTune (Hubertus, Wis.) can also help collect and analyze data. In software's absence, worthwhile tuning guidelines include:

• The "Ultimate Sensitivity Method" that uses varied gain settings, setpoints, process variable and output responses, and damping results to achieve a critically damped controller response. Users can also determine "usable" controller settings;
• The Reaction Curve (Open Loop) Method that makes step changes to a steady-state process and records PV results, including loop deadtime, loop time constant, and loop signatures for future reference;
• Using medium response tuning constants for integrating processes—such as level, batch temperature, batch pressure, and pH tanks—in which only one controller output setting produces a stable, balanced PV. An open-loop test is the best way to gather data for integrating processes; and
• Cascade control loops, which can effectively control slow moving variables, such as tightly maintaining boiler levels using a level controller cascaded to a flow controller. Cascade control strategies require secondary loop dynamics to be five times faster than the primary loop.

To control all loops in a process, users should work from the raw material to the final product starting with flows and then pressures, levels, temperatures, and finally whatever else remains. Control loop tuning begins with ensuring that each loop component can perform at its best and then patiently applying proven tuning methods. The payoff for optimum performance of all control loops can range from 5% to 25% improvements in quality and productivity.

For more information, visit www.controleng.com/freeinfo