Sensors – How to select the right sensor for the projects?

The sensor selection is a crucial activity in the systems design process, as it will make a great impact on the production of the measurement instrument and the performance during its entire lifetime. It is an act of engineering, in which the design is supported by advanced tools for simulating system behavior based on scientific knowledge.

Sensor selection means meeting requirements. The first task of the designer is to get as much information as possible about the future applications of the measurement instrument, all possible conditions of operation, the environmental factors, and the specifications concerning quality, physical dimensions, and costs.

Let’s understand the steps to select the right sensor by one illustration. We will take a position sensor because it is most widely used across many industries.

Position sensors get called a whole range of things and this can confuse. They get referred to as encoders, transducers, resolvers, transmitters. They can also include a description of their geometry such as a rotary transducer, angle encoder, linear transducer, etc. There are some subtle differences but to all intent and purposes, they all refer to the same thing – a device whose electrical output is proportional to the position or change in position.

Incremental and absolute sensors

There are two basic types of position sensors – incremental and absolute. The output from an incremental sensor is proportional to a change in position. The output from an absolute sensor is proportional to the position. The two types can be categorized by what happens at power-up. An incremental sensor will only provide an output when position changes whereas an absolute sensor will output a position signal without any need to move.

There are more incremental position sensors than absolute sensors. This is because traditionally absolute position sensors were more expensive but the price difference diminishing and absolute position sensors are usually preferred.

What is important for good measurement performance?

This is one area that more design engineers get wrong than right. Measurement performance for position sensors is usually specified by the following:-

Accuracy: the maximum between the sensors output and the actual position

Resolution: the smallest change in position that the sensor can measure

Repeatability: a measure of the sensor reproducibility

Linearity: a measurement of the deviation between the sensor's output to a straight line formed by the output of a perfect sensor and actual position.

Think about what is most important in your overall design. Many engineers either over-specify and pay over the odds for the sensor, or they will under-specify and couldn’t achieve the required performance.  

Match the kind of sensor to the application

Position sensors are used in many applications from ultra–high spec spacecraft to low-cost automotive and consumer appliances. There is numerous of different positions sensor manufacturers, sizes, shapes, and technologies. Of course, you will need to match the cost of the sensor with the allowable budget but it is simply not true to think that the more expensive the sensor the better it will be for your project. The key to success is matching the right kind of sensors to your project. Each of the different techniques used to measure position has its pros and cons. You should aim to match the characteristics of your application with the pros of the position sensing technology.

An important point is a difference between contacting and non-contacting sensors. Although there is a trend towards non-contact position sensors, potentiometer remains the most common position sensors. They measure a voltage drop as a contact slide along a resistive track. Potentiometer operates well in applications with a modest duty cycle, benign environment, and relaxed performance. Unfortunately, pots are susceptible to wear and foreign particles such as dust. They are typically only suited to low-end, low cots applications and they have a poor reputation.

Position sensing types:

	Rotary or Linear type	High measurement performance	Low thermal drift	Suitable for wet environment	Compact and lightweight	Suitable for dirty environments	Suitable for high shock and vibration	Easy installation
capacitive
Inductive proximity
optical
potentiometer
Reed switch
Hall effect

The table should be taken as a general guide and undoubtedly there will be exceptions. It is meant to help you seek out the kind of position sensor that might be right for your project. The column is the features against which you can filter out the sensors and select the right for your project. Get this selection wrong and you will pay the price later.

Understand the full cost of a sensor

The true cost of a sensor is not just the purchase cost. To pick the minimum cost solution be sure to consider the cost of failure. In the best-case scenarios, sensor failure is spotted at the design stage when an alternate sensor can be found. The impact is usually not great.

In the slightly worst case, failure is found at the test or qualification stage. Swapping out one type of sensor may have some knock-on consequences –but it's only rarely a disaster.

The trouble starts is when a position sensor fails in the field. In such scenarios, the few cents or dollars spent to fix different sensors become less significant. Often, the wrong kind of sensor was chosen for the environment. The potential scenarios are higher than anticipated repair/replacement costs; product recall for fitment of an alternate sensor; widespread field failure causing loss of reputation in the market; and field failure causing catastrophic damage.

The best thing a design engineer can do in such an instance is to avoid it ever happening. This is best done by choosing the right position sensor for the job – not necessarily the sensor with the lowest purchase price.