Controllers

The previous article Controlled System, dealt with the controlled system - the part of the system which is controlled by a controller. This article is about the controller. The controller is the device in a closed-loop control that compares the measured value (actual value) with the desired value, and then calculates and outputs the manipulated variable. The above section showed that controlled systems can have very different responses. There are systems which respond quickly, systems that respond very slowly and systems with storage property.
For each of these controlled systems, changes to the manipulated variable y must take place in a different way. For this reason there are various types of controller each with its own control response. The control engineer has the task of selecting the controller with the most suitable control response for the controlled system.

Control response

Control response is the way in which the controller derives the manipulated variable from the system deviation. There are two broad categories: continuous-action controllers and non-continuous-action controllers.

Continuous-action controller

The manipulated variable of the continuous-action controller changes continuously dependent on the system deviation. Controllers of this type give the value of the system deviation as a direct actuating signal to the manipulating element. An example of this type of controller is the centrifugal governor (Picture 1). It changes its moment of inertia dependent on speed, and thus has a direct influence on speed.

Picture 1: Centrifugal governor as Continuous-action controller

Non-continuous-action controller

The manipulated variable of a non-continuous-action controller can only be changed in set steps. The best-known non-continuous-action controller is the two-step control that can only assume the conditions "on" or "off". An example is the thermostat of an iron (Picture 2). It switches the electric current for the heating element on or off depending on the temperature.

Picture 2: Thermostat of an iron as Non-continuous-action controller

Time response of a controller

Every controlled system has its own time response. This time response depends on the design of the machine or system and cannot be influenced by the control engineer. The time response of the controlled system must be established through experiment or theoretical analysis. The controller is also a system and has its own time response. This time response is specified by the control engineer in order to achieve good control performance.

The time response of a continuous-action controller is determined by three components:

>> Proportional component (P component)
>> Integral component (I component)
>> Differential component (D component)

The above designations indicate how the manipulated variable is calculated from the system deviation.

Proportional controller

In the proportional controller, the manipulated variable output is proportional to the system deviation. If the system deviation is large, the value of the manipulated variable is large. If the system deviation is small, the value of the manipulated variable is small. As the manipulated variable is proportional to the system deviation, the manipulated variable is only present if there is a system deviation. For this reason, a proportional controller alone cannot achieve a system deviation of zero. In this case no manipulated variable will be present and there would therefore be no control. The output response of the proportional controller for proper input is shown on Picture 3.

Picture 3: Proportional Controller response (input and output)

Integral-action controller

An integral-action controller adds the system deviation over time, that is, it is integrated. For example, if a system deviation is constantly present, the value of the manipulated variable continues to increase as it is dependent on summation over time. However, as the value of the manipulated variable continues to increase, the system deviation decreases. This process continues until the system deviation is zero. Integral-action controllers or integral components in controllers are therefor used to avoid permanent system deviation. The output response of the integral-action controller is shown on Picture 4.

Picture 4: Integral-action Controller response (input and output)

Differential-action controller

The differential component evaluates the speed of change of the system deviation. This is also called differentiation of the system deviation. If the system deviation is changing fast, the manipulated variable is large. If the system deviation is small, the value of manipulated variable is small. The response of the differential-action controller is shown on Picture 5. A controller with D component alone does not make any sense, as a manipulated variable would only be present during change in the system deviation. A controller can consist of a single component, for example a P controller or an I controller. A controller can also be a combination of several components - the most common form of continuous-action controller is the PID controller.

Picture 5: Differential-action Controller response (input and output)

Technical details of controllers

In automation technology controllers are almost exclusively electrical or electronic. Although mechanical and pneumatic controllers are often shown as examples in text books, they are hardly ever found in modern systems. Electrical and electronic controllers work with electrical input and output signals. The transducers are sensors which convert physical variables into voltage or current. The manipulating elements and servo drives are operated by current or voltage outputs. Theoretically, there is no limit to the range of these signals. In practice, however, standard ranges have become established for controllers:

1. For voltage: 0 ... 10 V; -10 ... +10 V;
2. For current: 0 ... 20 mA; 4 ... 20 mA;

Internal processing of signals in the controller is either analog with operational amplifier circuits or digital with microprocessor systems:

>> In circuits with operational amplifiers, voltages and currents are processed directly in the appropriate modules.
>> In digital processing, analog signals are first converted into digital signals. After calculation of the manipulated variable in the microprocessor, the digital value is converted back into an analog value.

Although theoretically these two types of processing have to be dealt with very differently, there is no difference in the practical application of classical controllers.