Resonant Inductive Position Sensors
Resonant inductive position sensing is a non-contacting technology for sensing linear and rotary displacement. It is an evolution of the industry standard LVDT (linear sensing) and RVDT (rotary sensing).
The key difference is that complex coil windings are replaced with a sensor board built with conventional PCB technology. Printing coils on a PCB dramatically lowers cost and allows sensors to be tightly integrated with other circuitry.
Fig 1 : 50mm linear resonant inductive sensor
Unlike LVDTs and RVDTs, the moving element does not require careful alignment with the sensing coils. Instead, a magnetically coupled resonator marks the position of the target. This has a unique electronic signature akin to the vibration of a tuning fork. This vibration is readily distinguishable from metals and other items near the sensor which would otherwise be detected by the sensor along with the target.
Fig 2 : Target for resonant inductive sensor built by TDK-EPC
The coils used to build the target's inductor may be wound, as illustrated above, or printed on a PCB. Winding yields a much higher quality factor, which means larger signals and hence higher resolution and better immunity to EMI.
Fig 3 : Equivalent circuit of a simple resonant inductive sensor
A resonant inductive position sensor uses coils to energise the resonator inside the target and to detect the signals returned by it. The number of coils and their shape depends on the application and sensing geometry. The digitisers used in Tablet PCs built by Wacom sense the position of a resonator inside the pen relative to a sensor board placed behind a display. This may be 17 inches across and comprise an x/y array of perhaps 50 coils. However for sensing linear and rotary position a much smaller number of coils is required. A practical minimum is 3, as illustrated in the equivalent circuit above.
The processing electronics required for resonant inductive position sensors must sense the coupling factors between the target and sensor coils (kCOS and kSIN in the figure above). Position is calculated from these values. The calculation is typically a ratiometric one so that absolute signal levels , Q-factor and temperature have minimal effect.
Two different methods of detecting the coupling factors are in common use: continuous and pulse echo. For continuous operation, the electronic processor drives current into the sensor's excitation coil(s) and measures the EMFs induced by the resonator in sensor coils continuously. To help separate the resonator's signal from that of nearby metals and direct breakthrough from excitation to sensor coils, the processor uses synchronous detection. The electrical phase of the detection is set 90° away from the unwanted signals so that they are eliminated. However precise control over the phase is difficult due to unit to unit variability and temperature changes, and residual breakthrough causes inaccuracy and drift.
Pulse echo detection is used in higher quality systems, because it cleanly separates the resonator's signal from the unwanted ones. Processing electronics first generates an excitation waveform, which comprises a number of cycles of current at the resonator's frequency. This current causes oscillations to build up in the resonator. The current is then removed. The oscillation in the resonator starts to decay. The electronic processing detects the EMFs induced by the decaying oscillation in the sensor board's sensor coils. These EMFs are proportional to the required coupling factors between resonator and sensor coils.
Fig 4 : Pulse echo interrogation
As noted above, sensor boards come in different sizes and geometries. These details determine the relationship between coupling factor and position, and therefore the calculation required inside the processor. One of the simplest coil arrangements is designed to yield a sinusoidal relationship between coupling factor and position, with COS and SIN coils in (spatial) phase quadrature. This can be achieved with the coil layout shown below.
Fig 5 : Linear sensor with sinusoidal sensor coil patterning
The excitation coil around the perimeter of the sensor "illuminates" the resonator (kEX) independent of position. The SIN coil yields a sinusoidal variation in coupling factor, and the COS coil its quadrature equivalent. In this case, the calculation of position is performed using a "4 quadrant inverse tangent". This is equivalent to measuring the angle (Pr) of the (kCOS, kSIN) vector as illustrated below.
Fig 6 : Position calculation for a sinusoidally patterned sensor
This calculation is ratiometric because it depends only on the relative value of the two coupling factors kCOS and kSIN. If they are both doubled or halved, the angle Pr remains the same. This yields immunity to variations in supply voltage, coil resistances, temperature and the sensitivity of processing electronics.
The processing electronics for a resonant inductive position sensor is ideally implemented on a single chip such as the CAM204 Central Tracking Unit (CTU) chip from CambridgeIC. This may be used in conjunction with a variety of different linear and rotary sensors. One of the attractive features of resonant inductive position sensing is the ability of a single chip to measure multiple axes, as illustrated below. This arrangement yields a particularly cost effective solution, because the same chip can sense multiple axes.
Fig 7 : Multi-axis resonant inductive position sensor system using a single processor chip
The strengths and weaknesses of resonant inductive position sensors are summarised in the table below.
Resonant inductive position sensing Strengths Weaknesses * Operates in harsh environments
* Extremes of temperature have a minimal effect on output
* Can sense at gaps of several millimeters between sensor and target
* Simple to mechanically integrate
* No loss of absolute position at power on
* A single processor chip can sense multiple axes
* Sensors are built from conventional PCBs and can be integrated with other circuitry
* No difficult materials or manufacturing processes
* Cost effective
* Does not sense through metal due to eddy currents
* Measurement takes time (~100µs or more)
* Resolution and accuracy can not match the best optical encoders
* There is a limit to how small sensors and targets can be made
This document is © 2010 Cambridge Integrated Circuits Ltd (CambridgeIC). It may not be reproduced, in whole or part, either in written or electronic form, without the consent of CambridgeIC. This document is subject to change without notice. It, and the products described in it ("Products"), are supplied on an as-is basis, and no warranty as to their suitability for any particular purpose is either made or implied. CambridgeIC will not accept any claim for damages as a result of the failure of the Products. The Products are not intended for use in medical applications, or other applications where their failure might reasonably be expected to result in personal injury. The publication of this document does not imply any license to use patents or other intellectual property rights.
This information has been kindly supplied by:
Cambridge Integrated Circuits Ltd
21 Sedley Taylor Road, Cambridge, CB2 8PW. UK
Tel: +44 (0) 1223 413500
Web : www.cambridgeic.com
Home - Website - Search - Suppliers - Links - New Products - Catalogues - Magazines
Problem Page - Applications - How they work - Tech Tips - Training - Events - Jobs - Register