Choosing the best proportional-integral-derivative (PID) controller will depend on your specific needs. A PID controller can be in the form of a PI or PD controller, or only I or P. Not all applications require the use of all three parameters. The derivative control is the one most likely to be eliminated as it makes measurements based on system noise. Elimination can be done by setting unwanted parameters to zero.
Widely used as industrial controls, a PID controller calculates three separate parameters, characteristics, or measurable factors. Calculations of error values are made by taking the difference between a measured amount and the desired amount. Errors are minimized by adjusting the inputs to the control system.
In a PID controller, any proportional changes that are too large may cause system instability. If the same changes are too small, the system will not be responsive. The integral control measures the amount of error and attempts to minimize it. Derivative controls reduce the rate of change, but can slow response time and introduce more noise into the system.
To understand the control process, a good example is manually adjusting water temperature on a two-tap faucet. Both hot and cold water faucets are turned on and then adjusted by the user to the desired combined temperature. Adjustments must be made precisely or the user will go back and forth between water that is too hot or too cold. Full proportional controls eliminate on and off cycling in the system. A PID controller will automatically compensate when changes in the system are sensed.
The simplest control systems can be used for basic thermostat systems. A PID controller in an oven may work best with only proportional and integral controls. The derivative function could cause erratic changes from noise or electrical interference. Properly functioning, the control allows the oven to heat to the desired temperature and then cycles on and off to maintain it. Heating is slowed as the oven reaches the desired temperature to avoid going over the set point.
Basic on and off controls are fine in systems that do not require constant exact temperatures. Home heating and cooling units can use this, but better efficiency will be achieved with a proportional or PID controller. Industrial uses normally demand constant control for laboratory-type uses. Motion, temperature, and flow control requirements can all be met with PID functions. When a steady state error (SSE) is critical, all three controls — working together — will provide the desired outcome.
Factors that need to be considered are the type of input sensor to the system and the range of results allowed. Next, the output needs must be met. The outputs may be to an electromechanical relay, an analog receiver, or solid state relay (SSR). Finally, take into account the number of outputs required. PID controllers commonly come with a listing of all input and output types they work best with.