Fitting the right type of linear position sensor to an application requires at least a working knowledge of the attributes of this electromechanical device. Starting with the basics, the LVDT (linear variable differential transformer) is a common type of linear position sensor widely used in electromechanical systems today. It consists of two basic elements: a stationary coil assembly and a movable core or armature. While most LVDTs are fundamentally AC-in/AC-out devices, some have electronics built-in to make them DC-in/DC-out devices. This gives rise to the terms “AC-LVDTs” and “DC-LVDTs”.

An LVDT has a natural null point in the magnitude of its AC output because it is typically connected differentially. With no end position stops on the LVDT, the null position, located in the middle of the range of motion of the LVDT’s core, is the “stake in the ground” for determining core position.

With the myriad of linear position sensors available on today’s market, selecting the right LVDT for an application involves two high-level choices based on interfacing to the LVDT, as well as some lower-level choices based on the LVDT’s performance specifications and the application environment.

First, an engineer should be concerned about the mechanical interface, followed by the electrical input/output (I/O). After high-level choices have been made, lower-level choices must be made based on an LVDT’s performance specifications and environmental ratings. Environmental ratings for either an AC-LVDT or a DC-LVDT are typically fairly easy to interpret. However, the performance characteristics of an LVDT often require a more detailed explanation. This is true both when choosing an available LVDT or developing the specifications for one for an OEM application. The following five terms and parameters often cause the most confusion when choosing an LVDT.

Nominal Linear Range

The basic variable in LVDT selection is the maximum range of core motion, which produces an analog output of specific linearity. Full-scale displacement is the distance a core can travel from its null position in this linear region. Since the core can be displaced from null toward either end, the linear operating range is twice the full-scale displacement. When stated as plus or minus (±) fullscale displacement, it is referred to as the nominal linear range. When stated without a polarity, it is called the LVDT’s full range, full stroke, or total stroke.

The nominal linear range of any LVDT varies somewhat with frequency. When the LVDT is used with the correct core for the specified frequency, the actual linear range will always equal or exceed the nominal value. When optimum linearity is not essential in an application, the practical operating range may extend well beyond the specified nominal linear range. Nominal linear range is specified for a high impedance load, typically 50 kOhm to 0.5 MOhm. A low load impedance can have a deleterious effect on linearity and nominal linear range.

Linearity Error

Shown are the components of a typical LVDT linear position sensor. The electromechanical transducer’s internal structure consists of a primary winding centered between a pair of identically wound secondary windings, symmetrically spaced about the primary.

As LVDT output is a nominally linear function of core displacement within its linear range of motion, a plot of output voltage magnitude versus core displacement is essentially a straight line. Beyond the nominal linear range, output begins to deviate from a straight line into a gentle curve. From a statistically best-fit straight line versus core displacement within an LVDT’s nominal linear range, the maximum deviation of LVDT output is defined as the linearity error or the non-linearity of the LVDT.

Linearity error is typically expressed as ± a percentage of full-range output (FRO), or in terms of an error band width that envelopes the straight line and deviations. The statistically best-fit straight line is usually determined by applying the method of least squares to a series of calibration readings. The proper interpretation of the linearity error specification for an LVDT depends on the ultimate application on the LVDT in a measuring system. Some users use non-linearity as a measure of system accuracy as it is often the largest error.

While typical linearity errors of ±0.25% of full range output are common with standard LVDTs, improvements to these specifications are possible with special construction techniques or by the use of onboard signal processing. Linearity errors as low as ±0.05% of full range output can be obtained in this manner. In some cases, improved linearity may also be obtained by using an AC-LVDT at less than its full range, or on only one side of null. Linearity error means the same for AC-LVDTs as well as DC-LVDTs.

Full-Scale Output

For an AC-LVDT, full-scale output is the output of an LVDT with its core positioned at full-scale displacement and with its primary excited at a specified nominal input voltage. In most cases, though, a better way to compare AC-LVDTs of the same linear range is through sensitivity. Sensitivity is usually specified in terms of milliVolt output per thousandths of an inch core displacement per Volt of excitation (mV/mil/Volt). Sensitivity varies with excitation frequency, which must also be specified. Sensitivity mostly affects the gain required of the LVDT’s signal conditioning electronics.

For most DC-LVDTs, the comparable characteristic to sensitivity is scale factor, which is usually expressed as Volts DC output per inch of core displacement. Some legacy DC-LVDTs use a ratiometric configuration, which requires that they also use the same units as sensitivity, or have their scale factor specified for a particular DC input voltage. There are also DC-LVDTs whose output is into a 4-20mA current loop, so their scale factors are expressed as milliamperes per inch (mA/in) or milliamperes per mil (mA/mil).


Resolution is the smallest core position change that can be observed in LVDT output. An LVDT’s resolution is essentially infinite, as it operates on the principle of magnetic coupling. An infinitesimal change in core position will produce an output change. In practice, the limitation on system resolution is the ability of the associated electronic equipment to sense the change in LVDT output, which is called the signal-to-noise ratio of the system. With a properly designed LVDT measuring system, microinch resolution is not uncommon.


The ability of a sensor to reproduce the same output for repeated trials of exactly the same input under constant operating and environmental conditions is the single most important factor for sensor selection. Called repeatability, this parameter is the only irreducible and uncorrectable source of static error in any electromechanical measuring system. Repeatability error is the limiting factor in making any sensor-based measurement.

A well-made LVDT is so repeatable that overall transducer repeatability is affected only by the mechanical factors of the physical members or structures to which the LVDT’s core is attached, and to which the LVDT’s coil is mounted. Both repeatability and resolution contribute to overall measurement error, and are usually expressed as a percentage of full-scale output. These parameters apply equally well to AC-LVDTs and DC-LVDTs.

This article was written by Lee Hudson, Application Engineer at Macro Sensors, Pennsauken, NJ. For more information, Click Here .