Every closed loop control system needs at least one sensor to function properly. The role of the sensor is to feedback to the control algorithm the variable that needs to be controlled. Example, if we need to control the speed of a machine, we need a speed sensor to read the actual speed (in real-time) and feed it back to the control system (controller).
Generally speaking a sensor can be defined as a device that receives a stimulus and responds to it with a signal. This definition is so broad that it can be applied to almost anything in nature.
Imagine a car driving on a road with a 50 kph speed limit. The driver will adjust the accelerator pedal and brake pedal positions in order to maintain the vehicle speed around 50 kph. The actual vehicle speed will be read by the driver by looking at the speedometer.
If we make an analogy of the vehicle being driven on the road with the above block diagram of a closed loop control system, we’ll get:
|Closed loop control system (block diagram) components||Vehicle being driven on a road|
|Reference||Road speed limit (e.g. 50 kph)|
|Measured error||The difference between reference speed and actual vehicle speed|
|System input||Accelerator and/or brake pedal position|
|System||Vehicle being driven|
|System output||Vehicle speed|
|Sensor||The actual vehicle speed sensor + the speedometer + the driver’s eyes|
|Measured output||Actual vehicle speed|
For this particular example, the “sensor” is not an exact component but a complex of several components, which have the purpose to inform the driver about the actual vehicle speed. Without any of these components, the driver could not read the actual vehicle speed.
In engineering we are interested about information being read from artificial, man made systems. Since most closed loop control system have electronic controllers, sensors must send the information as an electric signal. For this type of systems the definition of a sensor narrows a bit:
A sensor is a device that receive a stimulus and sends an electrical signal.
If we regard the sensor as an input-output system, the stimulus (input) is a measured variable, a physical or chemical quantity (in most of the cases non-electrical). In most of the cases the signal (output) is an electrical quantity (voltage).
The purpose of a sensor is to respond with an electrical signal to an input physical property. The electric signal send by the sensor must be compatible with the receiving electronic circuit (control module).
Sensors can be classified by different criteria. A commons criteria is the type of characteristic curve, which represents the value of the output signal function of the input (measured variable). From this point of view, a sensor can have:
- linear characteristic
- non-linear characteristic
Linear (continuous) characteristic curves are widely used in control applications. Linear sensors have the advantage that they are easy to check for consistency, and are easy calibrate for particular applications. Nonlinear (continuous) sensors are used where the control loop is focused on a very narrow working range.
The sensor characteristic is also regarded as a transfer function. Except the characteristic (graph), the relationship between the input and output of a sensor can also be expressed as a table of values or as a mathematical expression. If the input-output relationship of a sensor does not vary in time, we can refer to it as a transfer function.
Another criteria for sensor classification, is the type of output signal. Sensors with analog output signal can send the information as:
- a voltage with varying amplitude
- a wave with fixed amplitude but variable frequency
A discrete sensor output can have two stages: high or low (e.g. +5V or 0V) or can be a rectangular shape wave with variable frequency (duty cycle).
Sensors can not operate independently, they need to be connected to an electronic device. In the case of a closed loop control system, the output of the sensor must be connected to the controller. In most of the cases, the controller is an Electronic Control Module (ECM).
The electric signal generated by a sensor can not be directly used by the ECM. The signal is either too noisy, or too weak, or, generally speaking, doesn’t have the electrical properties expected by the receiver (ECM). To match the output of the sensor with the expectations of the ECM we need a signal conditioning circuit.
The signal conditioning module (circuit) is an interface which has the purpose of aligning the output of the sensor with the requirements of the load device (ECM). The signal conditioning circuit is called signal conditioner and is specifically designed to provide signal scaling, amplification, linearization, filtering, attenuation and other signal processing functions.
A very common signal conditioning function is amplification. Signal amplification is needed in order to provide to the analog-to-digital (A/D) converter a much stronger signal, which results in a higher digital signal resolution.
An analog-to-digital converter (ADC) reads the analog signal from the sensor and produces digital output signals which can be interpreted by the electronic control module. The main characteristics of the A/D converters include resolution, conversion speed, accuracy, linearity, and stability.
A common term used for sensor is the transducer, which is not quite accurate. A transducer is a converter between different types of energies (e.g. chemical to thermal), while a sensor is a converter of any type of energy into electrical energy. A transducer can be part of a more complex sensor, as an intermediate step of energy conversion.
Sensors are widely used in all areas of engineering, especially in those which require closed loop control system. From this point of view, having good knowledge of sensors and signal processing, should be a requirement for every engineer.
For any questions or observations regarding this tutorial please use the comment form below.
Don’t forget to Like, Share and Subscribe!