By Anders Guldahl
Sensing is all about the ability to detect or measure changes in physical properties. In the context of an electronic control system, the requirement is to translate a parameter such as temperature, pressure or movement into an electrical signal. While some sensors directly produce a voltage output, which provides the ideal input for a microcontroller-based system, the majority of sensors depend on resistive, inductive or capacitive circuit elements whose behavior varies according to a known characteristic. These sensors typically require an external circuit to convert their output into a measurable signal for capture by a microcontroller (MCU).
Understanding the different sensing technologies is important in determining the most appropriate type of sensor for a given application, especially when more than one type is available for measuring a particular parameter. For example, the previous two articles in this series looked at capacitive and resistive sensors and both technologies provide solutions for detecting position or movement. Inductive sensors, the focus of this article, can equally be used for position/movement detection so clearly there must be good reasons for choosing one type of sensor in preference to another. In most instances this will depend on the more specific nature of what is being sensed or measured and then finding a sensor that matches that performance requirement. Other considerations maybe those related to accuracy, reliability and environment conditions as well as ease of implementation or quite simply cost.
This article considers a number of inductive sensing applications to give an appreciation for the technology and how these systems work. It concludes by highlighting the importance of energy efficient solutions using microcontrollers that have been optimized for ultra low power sensing applications, such as the Energy Micro EFM32 Gecko series MCUs that can monitor sensor inputs autonomously without needing to continually wake the processor from a low-energy sleep mode.
Inductive sensing applications
In order to appreciate inductive sensors, it is perhaps necessary to remind ourselves about the basic principles of inductance and hence what makes an inductor and affects how it operates. Essentially all conductors possess inductance — it is the property whereby a change in current through the conductor “induces” a voltage in the conductor (see equation below). This is because in normal steady-state conditions a current flowing through a conductor results in a magnetic field being formed around the conductor. But when the current is changing (increasing or decreasing) the resulting increase or decrease in magnetic flux opposes the change in current. Hence an inductor is characterized by this resistance to change in current but also by the way it stores energy as a magnetic field, much as a capacitor stores energy as a electrical charge.
Where V is the induced voltage, L is the inductance and is the rate of change of current
As we shall see, most inductive sensing applications employ the inductive element as part of a resonant circuit, either detecting the change in frequency as the inductance varies or the detuning effect resulting from the presence of metallic objects:
Many industrial control applications rely on inductive sensors for the accurate proximity detection of metallic targets. Figure 1 shows a typical sensor in a cylindrical housing but other shapes and sizes exist to serve different applications. These devices use an inductive coil as part of an L-C oscillator circuit to generate a high frequency magnetic field in front of the sensor face. When the target enters this field it absorbs some of its energy, and attenuates the oscillator output. Proximity sensors of this type are mainly designed for close-range, non-contact, go/no-go operation and hence they incorporate a detector circuit, triggered at a predefined threshold, to drive an output switch. However some designs output an analog signal that is optimized for measurements over a slightly greater range (from a few millimeters to a few centimeters), particularly when used with ferromagnetic targets that absorb more energy than other target metals and therefore offer increased operating distance.
Compared to other technologies, the key advantage of inductive proximity sensors is their accuracy, both in terms of absolute accuracy and, often more importantly, their repeat accuracy. Other benefits of inductive proximity sensors are their high switching rate (see the next section for more detail) and their suitability for use in harsh environmental conditions. The latter is catered for by commercially available devices operating over extended operating temperatures (from -40°C to +250°C), offered in pressure resistant housings (up to 500 bar) and sealed options meeting IP68 (full ingress protection against dust and immersion in water) or even IP69k (high pressure water jet resistant). Sensors approved for use in explosive gas or dust zones are also available as are models with stainless steel sensing faces, suitable for control applications in the food processing and pharmaceutical industries.
On the downside, apart from their limited operating range, is the fact that inductive sensors only work with metallic i.e. conducting targets. Also, to achieve their specified accuracy, these targets often need to be of defined size and material although some manufacturers do specify sensitivity de-rating factors for non-ferrous metals e.g. 0.9 for stainless steel and 0.4 for aluminum or copper.
Figure 1. Generic analog output inductive proximity sensor
Rotational speed sensors
The type of inductive sensor described above can of course be used to detect proximity based on axial movement i.e. towards or away from the face of the sensor or, in the case of specifically developed ring format sensors, can be used for counting small metallic parts passing through the inductive loop. The basic format can also detect a target passing laterally in front of the sensor and this principle is readily extended to tangential movement i.e. detecting the rotation of a toothed wheel as shown in figure 2.
A key difference for rotation speed sensing is that, for most application requirements, it is not necessary to use the sensor’s coil as part of an L-C oscillator to generate an alternating magnetic field. Instead, as shown, the sensor incorporates a permanent magnet whose field is modulated by the teeth passing of the sensor face, inducing a sinusoidal output voltage in the coil.
The automotive industry is a major market for rotational sensors where they are used to provide input to engine control units (ECUs) by measuring the speed and position of the crankshaft (position, for dynamic engine timing purposes, can be detected by having a gap in toothed wheel). They are also used in automatic braking systems (ABS) where individual sensors are located on each wheel hub along with a toothed wheel attached to the CV joint.
Figure 2. Inductive sensor used for rotational speed measurement
Traffic light sensors
Briefly going back to basics again, the capacity of an inductor is determined by various factors; the number of coils, the cross-sectional area of the coil, the length of the coil (short or overlapping coils produce more inductance) and the material at the core of the coil. All these parameters allow inductive proximity sensing to be scaled up to provide a means for controlling traffic lights, by detecting the presence of a vehicle waiting at a stop light rather than simply relying on timer control.
These systems work by embedding a car-sized coil of wire into the road surface and connecting it to a system that can measure the difference in inductance depending on whether a vehicle is positioned over the loop or not. Clearly in this application the large cross-sectional area of the coil, the number of loops of wire (typically 5 or 6) and the short, overlapped form of the coil all contribute to a reasonably high initial inductance. Any vehicle, which is effectively a large metallic (often ferrous steel) object, stopped at the lights becomes part of the coil’s core, significantly increasing its inductance and enabling the system to recognize its presence. Such large-scale proximity detection solutions can also be applied to similar applications such as the movement of larger metallic objects along a production line.
Metal detectors are used for a variety of purposes: from security applications such as the scanners used in airports, government buildings, prisons, schools, etc. or the detectors used to locate land mines, through to more industrial, commercial or even consumer applications; such as mineral or archaeological prospecting, the detection of electrical cables or steel reinforcement bars in buildings and the hobbyist uses for beach-combing or treasure hunting. An interesting commercial application is the detection of metal contaminents in food or garment production.
Although there have been many refinements to the technologies applied to metal detecting, all share similar operating principles to the inductive sensors we’ve already considered. The simplest form uses a transmitter coil as part of a very low frequency (VLF) oscillator to generate an alternating magnetic field. This will induce eddy currents in any nearby electrically conductive metal which, in turn, produces its own magnetic field that can be picked up be a second (receiver) coil in the detector. The sensitivity of such metal detectors depends on shielding the receiver coil from the magnetic field generated by the transmitter coil. Analysis of the received signal can not only indicate the distance from the detector to the metal object, but can also distinguish between different metals by measuring the phase shift that results from the different resistive and inductive properties of these metals. Pulse induction (PI) is another technique used for metal detection which, as its name suggests, transmits the alternating magnetic field in brief pulses. PI detectors can provide increased detection range but are not so good at discriminating different metals.
Energy efficient inductive sensing
Capacitive, resistive and inductive sensing techniques address a variety of applications and the circuits required to interface a given sensor to a microcontroller (MCU) will vary depending on what is being detected or measured. What doesn’t change is the need to minimize the involvement of the processor core in monitoring these sensors to keep energy consumption to the absolute minimum, without compromising system performance. Energy Micro’s EFM32 Gecko series MCUs achieve this with a Low Energy Sensor Interface (LESENSE) and Peripheral Reflex System (PRS) that allows its low-energy peripherals to be configured using sequencer and decoder circuits to detect and evaluate a combination of sensor states and event patterns before waking the MCU, as illustrated in figure 3.
Figure 3. Energy Micro’s LESENSE interface enables conditional MCU wake-up
Anders Guldahl is an Application Engineer
P.O. Box 4633 Nydalen, N-0405
Telephone: +47 23 00 98 00
Fax: +47 23 00 98 01