Displacement Sensor

What is a Displacement Sensor

A displacement sensor is a measurement element that detects how far a target object has moved relative to a reference point and outputs a proportional electrical signal. Displacement is, strictly, a change in position. A device that reports the change from a null or zero point is a displacement sensor, while a device that reports an absolute location within its range is more precisely a position sensor. In practice the two terms blur together, and the engineering distinction that matters is whether the reading is incremental (counted from a reference and lost on power-down) or absolute (recovered directly after a power cycle without re-homing).

Structurally, most displacement sensors share three functional blocks. First, a sensing transducer that converts mechanical travel into a primary physical change, for example the coupling between an LVDT core and its windings, the capacitance across an air gap, or the position of a reflected laser spot on a detector. Second, an excitation and front-end stage, such as the AC carrier that drives an LVDT primary at a constant 1 to 10 kHz, or the oscillator behind a capacitive probe. Third, a signal-conditioning stage that demodulates, linearises, and scales the raw signal into a standard output, typically 0 to 10 V, 4 to 20 mA, or a digital interface such as SSI, IO-Link, or an absolute encoder protocol.

The history of displacement measurement runs from the mechanical dial indicator, refined by the American Brown & Sharpe and later Mitutoyo in the early twentieth century, to the inductive LVDT, whose differential transformer principle was patented in the 1930s and matured for aerospace use during the 1940s. Optical linear scales arrived with precision machine tools in the 1950s and 1960s, magnetostrictive position sensing was commercialised by MTS Sensors in the 1970s for hydraulic feedback, and capacitive and eddy current nanometre-class gauges became laboratory and semiconductor staples by the 1980s. Laser triangulation displacement sensors, now dominant in inline inspection, became practical with affordable position-sensitive detectors and CMOS arrays.

The scale of application is what makes selection non-trivial. At the fine end, capacitive and eddy current gauges resolve into the single-digit nanometre region for spindle error analysis, disk-drive head positioning, and wafer-chuck flatness. At the coarse end, draw-wire and magnetostrictive sensors track several metres of travel on cranes, presses, and mobile hydraulics. Each technology has an optimal window of range, resolution, target material, and bandwidth, and trying to push a principle outside its window is the most common root cause of a measurement that looks fine on the bench but fails in production.

Four engineering metrics dominate quality across all technologies: linearity (how straight the output-versus-travel curve is), resolution (the smallest detectable change), repeatability (the scatter of repeated readings at the same point), and the environmental envelope (temperature, vibration, ingress, and target compatibility). Together these define whether a sensor is suitable for closed-loop machine feedback, laboratory metrology, or rough field monitoring, and they drive the total cost of ownership far more than the purchase price alone.

Contact and Non-Contact Types

The first branch in any displacement sensor decision is contact versus non-contact, because it determines wear life, achievable speed, and tolerance to dirty targets. A contact sensor has a moving member, a probe plunger, an LVDT core, or a draw-wire cable, that physically follows the target. A non-contact sensor reads across an air gap using a field or a beam and has no mechanical wear. The table below summarises the principal type families across both groups.

Contact sensors are simple to mount, read any target material, and tolerate surface contamination because the probe rides through it. The LVDT is the workhorse of this group: a spring-loaded or free-floating ferromagnetic core slides inside a transformer, and the differential of two secondary windings yields a clean, frictionless, near-infinite-resolution signal. Potentiometric sensors use a sliding wiper on a resistive track, which is inexpensive but wears and is the lowest accuracy class. Draw-wire sensors wind a thin stainless cable on a spring-loaded drum coupled to a rotary encoder or potentiometer, trading some accuracy for very long, compact stroke.

Non-contact sensors remove mechanical wear and unlock high bandwidth, which is why they dominate rotating-machinery, high-speed inspection, and precision metrology. Eddy current probes induce currents in a conductive target and read the resulting impedance change; capacitive probes measure the changing capacitance of the air gap to a conductive or, with care, non-conductive target; laser triangulation reads the geometry of a reflected spot. The cost is sensitivity to target properties: eddy current readings depend on target conductivity and permeability, capacitive readings on the dielectric and flatness of the gap, and laser readings on surface colour, gloss, and tilt. These dependencies mean non-contact sensors are usually calibrated to a specific target.

Magnetostrictive sensors sit in a useful middle ground. A permanent-magnet ring rides on the moving member with no electrical contact, while a fixed waveguide rod measures the magnet position by timing an ultrasonic torsional pulse. The result is an absolute, non-contact, wear-free reading over strokes from tens of millimetres to several metres, which is why this principle owns hydraulic-cylinder feedback. The practical limitation is that the magnet must be guided to keep a consistent radial gap to the rod, and the longest strokes carry higher unit cost than draw-wire alternatives.

Sensing Principles Compared

Beneath the type families sit a handful of physical principles, each with a characteristic envelope of accuracy, range, bandwidth, and cost. Selecting a principle is the highest-leverage decision because it sets the ceiling on what any product in that family can achieve. The table below compares the dominant principles using values cross-checked against manufacturer datasheets; treat them as representative class figures, not guarantees for a specific model.

LVDT and inductive principles drive an AC carrier through a primary winding and read the differential voltage induced in two opposed secondaries as the core moves. Because the coupling is purely magnetic and frictionless, resolution is limited only by the electronics noise floor and is effectively infinite. A representative industrial unit, the TE Connectivity Macro Sensors PR series, spans full-scale ranges from about ±1.25 mm (±0.05 inch) to ±250 mm (±10 inch) with maximum linearity error of ±0.25 percent of full range output by a least-squares best-fit line, and hermetic versions reach IP68 sealing for hydraulic and subsea service.

Eddy current probes drive a coil with a high-frequency field that induces circulating currents in a conductive target. The target currents change the coil impedance in proportion to the gap, per Faraday and Lenz behaviour, so the principle works only on conductive targets but is immune to non-conductive contamination such as oil, dust, and water in the gap. This robustness plus extreme resolution makes it the standard for spindle error analysis, turbine shaft vibration, and gap monitoring. Micro-Epsilon eddyNCDT 3700 series systems have demonstrated resolution down to about 0.09 nanometres over short ranges.

Capacitive probes form one plate of a capacitor with the target as the other, so the measured capacitance varies inversely with the air gap. They deliver the finest linearity and sub-nanometre resolution of any class, which is why they dominate semiconductor metrology, optics, and precision flatness work. A Micro-Epsilon capaNCDT 6228 system offers ranges from 1 mm up to 20 mm, while a fast probe such as the CSH05 reaches about 10 nm resolution at up to 8.5 kHz. The constraint is that capacitive sensing requires a clean, electrically stable gap and works best on conductive, flat targets at close standoff.

Laser triangulation images a reflected spot on a CMOS or PSD array, computing standoff from the spot geometry. It is non-contact, material-flexible, and very fast: the Keyence LK-G5000 series reaches about 0.02 percent of full scale linearity, repeatability near 0.005 micrometres, and sampling up to 392 kHz. Triangulation handles long standoff and large range but is sensitive to surface colour, gloss, transparency, and tilt angle, so dark, mirror-like, or translucent targets degrade the signal and may require confocal or time-of-flight alternatives. Magnetostrictive sensors trade fine resolution for absolute long-stroke position: a Temposonics R series device offers strokes from 50 mm to 2,500 mm with non-linearity below ±0.01 percent FS and repeatability below ±0.005 percent FS, making it the default for closed-loop hydraulic actuators.

Standards and Target Materials

There is no single global standard that governs every displacement sensor, so the relevant requirements come from the application domain rather than the device category. A purchasing engineer should ask which standard the receiving inspection, the safety case, and the calibration certificate must satisfy, then specify accordingly. The table below maps the most common standards to what they actually constrain.

Ingress protection is the most frequently misread spec. Under IEC 60529, IP67 means dust-tight plus temporary immersion to 1 metre for 30 minutes, while IP68 means dust-tight plus continuous immersion deeper than 1 metre at a depth and duration agreed between maker and user. Hermetically sealed LVDTs are routinely built to IP68 so they survive inside hydraulic cylinders and underwater. The trap is the connector: an IP68 housing fitted with an IP65 connector is only IP65 overall, so always confirm that the cable entry and mating connector carry the same rating as the body.

Extensometer standards matter whenever the displacement sensor measures strain on a materials testing machine. ISO 9513 assigns a numeric class (0.2, 0.5, 1, 2) and ASTM E83 assigns a letter class (A, B1, B2, C), each defining both a relative error in percent and a fixed error floor in micrometres or strain. For example an ISO 9513 class 0.5 instrument permits a relative error of 0.5 percent of reading plus a fixed error around ±1.5 micrometres. Choosing the class drives the price of the extensometer and the frequency of recalibration.

For non-contact sensors the target is part of the measurement system, so material selection is a calibration question rather than a corrosion question. Eddy current readings shift with target conductivity and magnetic permeability, so a probe calibrated on aluminium reads incorrectly on steel and must be re-mastered for each material. Capacitive readings assume a conductive, flat, grounded target; measuring a non-conductive target is possible but needs a dedicated mode and reduced accuracy. Laser triangulation is the most material-flexible but still degrades on very dark, glossy, or translucent surfaces.

For contact sensors the wetted and exposed materials follow the same logic as other field instruments. Stainless steel 304 and 316L housings cover most industrial and washdown duty; LVDT cores use high-permeability nickel-iron alloy; draw-wire cables use stainless steel for corrosion resistance; and seals use FKM, EPDM, or FFKM depending on the surrounding chemistry. In food, dairy, and pharmaceutical lines, specify electropolished 316L and hygienic IP69K sealing so the sensor survives caustic clean-in-place cycles and high-pressure washdown.

Key Specification Parameters

Reading a displacement sensor datasheet is a core skill, because a single product may list 15 to 30 parameters while only a handful drive the decision. The terms below recur across LVDT, capacitive, eddy current, laser, and magnetostrictive datasheets, and confusing them is the most common cause of an over-specified or under-specified order.

Measuring range and standoff. Range is the span of travel the sensor reads. For non-contact probes there is also a standoff or working distance, the gap at the centre of range, which is not zero: a laser sensor may have a 50 mm measuring range centred at a 100 mm standoff. Mounting the target outside the standoff window gives no signal even though the range looks adequate, so always specify both numbers.

Linearity. Linearity, often the headline accuracy figure, is the maximum deviation of the real output from a reference straight line, quoted as a percentage of full scale (FS or FSO) or full range output (FRO). The reference line is usually a least-squares best fit, so two makers quoting the same percent may use different fit methods. On a ±25 mm LVDT at ±0.25 percent FRO the worst-case error is about ±125 micrometres; on a laser sensor at 0.02 percent FS over 50 mm it is about ±10 micrometres.

Resolution and repeatability. Resolution is the smallest detectable change, set by the electronics and bandwidth, and for frictionless principles such as LVDT and capacitive it can be far finer than the linearity. Repeatability is the scatter of repeated readings at the same point and is the truest indicator of closed-loop control quality. A sensor with excellent repeatability but modest linearity still controls a process well, because the controller compensates the systematic offset but cannot remove random scatter.

Bandwidth and sampling rate. Bandwidth, the frequency to which the sensor tracks moving targets, separates static gauging from dynamic measurement. A dial-gauge-replacement LVDT may resolve only a few hundred hertz, while a laser sensor reaches hundreds of kilohertz for inline profiling at line speed. Under-specifying bandwidth aliases fast motion into a slow, wrong reading; over-specifying it adds noise and cost.

Temperature coefficient. Thermal drift, expressed as a percentage of FS per kelvin for both zero and span, dominates field accuracy more than the bench linearity figure. A laboratory sensor specified at 0.02 percent FS linearity is meaningless if it drifts 0.05 percent FS per kelvin across a 40 kelvin shop-floor swing. For precision work, request the measured thermal drift curve over the full operating temperature range, not just the catalogue number at 20 degrees Celsius.

Output and interface. The output is the interface to the control system, and the choices mirror other field instruments:

• 0 to 10 V / 0.5 to 4.5 V ratiometric: simple analog input to a PLC card, but susceptible to long-cable voltage drop and ground-loop error.
• 4 to 20 mA: two-wire current loop, immune to cable resistance over long runs, the default for process-area transmitters per IEC 60770.
• SSI / BiSS: synchronous serial digital, the standard absolute interface for magnetostrictive and encoder position feedback into a drive or motion controller.
• IO-Link / Modbus RTU: smart, parameterisable digital output for setup, diagnostics, and Industry 4.0 data collection.
• CANopen / PROFINET / EtherCAT: fieldbus and industrial Ethernet for multi-axis machines and large distributed systems.

Ingress and environment. Beyond the IP code, check the operating and storage temperature window, vibration and shock rating per IEC 60068, EMC immunity, and for safety-rated loops the SIL level and any Ex certification. A sensor that meets every accuracy number on the bench but lacks the required IP, vibration, or Ex rating cannot be installed, so these line items are pass-or-fail filters before accuracy is even compared.

Selection Decision Factors

To turn the preceding five chapters into a specific model, work the decision sequence below in order. Most selection mistakes come not from a single wrong value but from deciding low-level details, such as the output protocol, before the high-level principle is locked. These eight steps double as a reusable RFQ template.

1. Measured quantity and motion: First confirm linear versus rotary, then absolute versus incremental. If the machine must recover position after a power loss without re-homing, an absolute principle such as magnetostrictive or an absolute encoder is mandatory.
2. Range and standoff: Choose the range so the working travel sits within the central 30 to 80 percent of full scale, and for non-contact probes specify the standoff or working distance explicitly, not just the range.
3. Accuracy class: Separate machine feedback (0.1 to 0.5 percent FS is usually ample), precision gauging (sub-micrometre, capacitive or LVDT), and laboratory metrology (nanometre, capacitive or eddy current). Each tier roughly doubles cost, so do not buy metrology accuracy for a feedback loop.
4. Contact or non-contact and target material: Decide whether mechanical wear and target contact are acceptable. For non-contact, confirm target conductivity, permeability, flatness, colour, and gloss against the chosen principle and plan to calibrate to the actual target.
5. Bandwidth and dynamics: Match sampling rate to the fastest motion you must capture. Static gauging needs only hundreds of hertz; vibration, profiling, and high-speed inspection need kilohertz to hundreds of kilohertz.
6. Environment and ingress: Set the IP code per IEC 60529, the vibration and shock rating per IEC 60068, the operating temperature window, and any washdown or hygienic requirement. Verify the connector rating matches the housing rating.
7. Output and interface: 4 to 20 mA or 0 to 10 V for analog PLC input; SSI or BiSS for absolute motion feedback; IO-Link, CANopen, PROFINET, or EtherCAT for smart and multi-axis systems. Confirm the controller already supports the chosen protocol.
8. Certifications and total cost of ownership: Add SIL2 or SIL3 functional safety and ATEX, IECEx, or NEPSI explosion protection where the area demands it, then weigh purchase price against calibration interval, drift, and spare-part availability over the service life.

One commonly overlooked dimension is manufacturer serviceability: local calibration laboratory access, spare-probe and controller lead times, firmware and parameter-set portability, and the availability of a matched master target for non-contact recalibration. These seem secondary at purchase but determine downtime years later. TE Connectivity (Macro Sensors), Micro-Epsilon, Keyence, Temposonics, Honeywell, Balluff, SICK, and ASM all maintain regional support and calibration capability, which makes them dependable anchors for long-life machine and process installations, while lower-cost regional suppliers can be appropriate for non-critical or short-life loops.