Want the entire guide sent to you ? Register here and we’ll send it to you straight away !
Sensor Selection Guide
Selecting the right sensor for your application can mean the difference between collecting meaningful data that enables you to make informed decisions, or not. With many parameters to consider and numerous sensor options available, selecting the right sensor can be a minefield.
The team at Emolice have combined years of sensor experience to produce this 3 part guide, designed to demystify the key areas that you need to consider in the sensor selection process.
Part 1 of the guide covers the Environmental factors that significantly impact the selection of your sensor.
Part 2 of the guide looks at Application considerations including technologies, communications and constraints
Part 3 of the guide concludes with a review of how key sensor specifications such as frequency response, drift, stability and impedance matching impact Data accuracy.
Part 1 : Environmental
Whatever you are measuring, one variable is always present; the environment the sensor has to operate in. Variables related to the environment are the key to sensor selection and are the first factors that need to be considered.
Indoor / Outdoor Installation
These factors influence the ingress protection (IP) rating that the sensor needs to meet in order to perform accurately and without premature failure in the chosen application environment. It is important that you select a sensor with a suitable IP or NEMA equivalent rating. To learn more about IP ratings download our Short Guide to IP Ratings from the download section of our website.
The temperature that the sensor will experience in situ can have a significant impact on sensor accuracy.
The effects of temperature are easily managed if the sensor is installed in a controlled environment such as a laboratory or in a test environment where temperature fluctuations during operation are measured in single degrees that fall within the operating range of the sensor.
However, greater consideration needs to be given to the operating temperature of the sensor if it is to be used in an outdoor environment that can experience wide fluctuations in temperature (of note is the influence of direct sunlight, which can add significantly to the operating temperature), or in an application where the sensor is subject to other sources of heat generated by equipment or processes.
Most sensors will quote an operating temperature range within which the accuracy of the sensor can be achieved. However, at hot or cold temperature extremes the accuracy can decrease non-linearly. To account for this change in linearity, sensors can be temperature compensated (either actively or passively) to reduce the effect of temperature within the operating range.
The main factor that limits a sensors’ smallest possible measurement is electrical noise. Some environments such as factory floors are inherently noisy and the contribution of constant noise from 50 or 60 Hz AC power circuits can result in the desired output signal from the sensor being lost in the noise floor.
Reducing the entry of noise into cables connecting sensors to instrumentation through correct routing and use of shielded, twisted pair cabling can reduce erroneous readings and reduce the resolution of measurement.
Additionally, electrical noise is generally broadband (meaning it contains a wide spectrum of frequencies) so selection of a sensor with an integral low-pass filter will reduce or eliminate high frequency noise, but at the expense of usable bandwidth.
Finally, whilst they are typically less expensive, the signal output of devices using voltage (such as 0-10V outputs) are more easily influenced by environmental noise than current devices (such as 4-20mA outputs) which are largely unaffected.
Part 2 : Application
The most fundamental decision to take when considering an application is to determine which physical parameter to best measure in order to give you the required insight into your process. Is the parameter easily measurable or can another more simply or accurately measured parameter be used?
A basic example is an application that requires the determination of the amount of solid product in a vessel.
A solid product can be measured through level, volume or weight.
Level measurement can be inexpensive, but a solid product will not settle completely in the vessel causing peaks and valleys on the surface. This causes problems for accurate level measurement as the level at a peak may be very different from the level at a valley.
Volume measurement overcomes the limitations of level measurement by measuring levels at multiple points on the surface and creating a map of the surface from wall-to-wall within the vessel. With sufficient detail, this map can provide accurate volume calculation regardless of peaks and valleys. However, installing such capability requires significant capital investment.
Measuring the weight of a solid product overcomes the problems of product settling but introduces other challenges. For example, when retrofitting weighing solutions to existing processes where vessels need to be emptied and raised for installation and commissioning.
Some parameters can be measured using a variety of technologies. Using the example above, a solid product can be weighed statically using strain gauge technology load cells underneath a vessel or hopper, or dynamically using capacitive technology load cells on a moving belt system. The decision of which solution to choose is normally a trade-off between price, accuracy and minimisation of disruption to existing processes during installation and commissioning.
The space in which the sensor must fit can often be an overriding consideration when selecting a sensor. When designing a new installation space is a compromise that can often be designed out, however when faced with a retrofit situation it can be the dominant consideration. Typically the physical size of the sensor is heavily influenced by its’ durability, number of interfaces and its’ accuracy or precision.
The distance the sensor is located from the instrumentation is a key consideration when deciding which output signal to select. If the distances are large, a current output device (4-20 mA) is usually a better choice than a voltage output device, as current output devices are not susceptible to losses due to the resistance in the communication cable. Another distinct advantage to a 4 -20mA system is that when the sensor is subjected to no inputs it will give a 4mA output. If the output drops to zero it is a clear indication that the communication cable has been damaged or disconnected.
Wireless sensor systems can transmit signals up to ½ a mile in a clear line of sight, noise free environment but this can be significantly reduced if obstructions or interference is present.
Specifying the most suitable range for a sensor involves a number of considerations. Often a sensor is over specified in the hope of avoiding any possibility of overloading it, resulting in damage to the sensor, costly downtime and delays. However, often as the useable range of a sensor increases, the measurement resolution decreases.
Part 3 : Data
Part 3 of the guide concludes with a review of how key sensor specifications such as frequency response, drift, stability and impedance matching impact Data accuracy and will be published shortly. Register here if you’d like us to send the entire guide to you now.