What is an absolute encoder?

An absolute encoder assigns a distinct position value to each position within its measuring range. A rotary encoder reports shaft angle; a linear model reports the location of a moving part. Its defining feature is the ability to read where it is within that range, rather than only tracking how far it has moved. The reading initially refers to the encoder’s own reference; its relationship to the machine’s zero position is established during commissioning.

How does it generate position data?

A traditional optical absolute encoder uses a coded disc, a light source and detectors. Each disc region has a different pattern, which the electronics convert into a position value. A magnetic absolute encoder can instead detect magnetic field direction. Optical versus magnetic describes the sensing method; absolute versus incremental describes how position information is provided.

Other sensing methods include Hall effect and capacitive encoder technologies. Consequently, not every absolute encoder contains a disc with several coded rings. The same absolute measurement concept can be applied to positions along a scale in a linear encoder.

Animation: the same position gives the same code

This 4-bit teaching model divides one revolution into 16 distinct positions. The red line is the stationary reader; the blue pointer follows the shaft. Black represents 1 and white represents 0, with the rings read from the most significant bit outside to the least significant bit inside. Each sector spans 22.5°; the reported angle is the start of that sector.

Select Switch power off: the disc continues moving, but the position output is unavailable. Restore power and the reader reads the code currently beneath it. This does not mean the encoder transmits data without power; it illustrates a single-turn absolute measurement being available again after power returns. Actual resolution, coding and data transmission depend on the device.

Why use Gray code?

In Gray code, only one bit changes between adjacent positions. In this model, position 7 has code 0100 and position 8 has code 1100. This reduces ambiguity caused by several bits changing at slightly different times at a code boundary. Gray code is a representation of a number: interpreting 1100 directly as binary 12 would be incorrect. The controller must apply the appropriate conversion.

Single-turn versus multi-turn

Single-turn: Reports position within one revolution. A shaft angle of 40° produces the same single-turn value as 400°, after one additional full revolution. That may be sufficient to identify the orientation of a rotary table, but it does not tell you how many complete turns have occurred.

Multi-turn: Reports both the angle within a revolution and the revolution information. This matters for lead screws and other mechanisms with travel extending across several turns. Tracking rotations during a power outage depends on the design, which may use gearing, battery-backed electronics or Wiegand energy harvesting. Not every multi-turn encoder is battery-free; check its power-off behavior and measuring range in the technical documentation.

Incremental versus absolute: how position is obtained

An incremental encoder generates pulses as it moves. A and B channels can indicate direction, while the controller counts edges to calculate relative position. Establishing machine position generally requires a homing procedure. A Z index pulse marks a reference within a revolution; it does not, by itself, identify which complete revolution the machine is in.

An absolute encoder supplies a current position value. In a suitable system this can reduce or eliminate the need for a homing move at every startup. However, absolute does not automatically mean more accurate, nor does incremental automatically mean less accurate. Compare the actual accuracy specification, resolution, mounting and mechanical transmission.

Bits, resolution and accuracy are different

n bits can represent 2ⁿ distinct positions per revolution. The theoretical angular step is 360° / 2ⁿ. For example, 12 bits = 4,096 positions ≈ 0.0879° per step, while 16 bits = 65,536 positions ≈ 0.00549° per step. These are resolution figures, not a guarantee that the measured angle falls within those error limits.

Accuracy describes the difference between reported and true position; repeatability describes the consistency of results when returning to the same position under the same conditions. More bits cannot remove backlash or a loose mechanical connection. A multi-turn encoder’s total bit count may also include both single-turn position bits and revolution-count bits; they do not all increase angular resolution.

Connecting to a PLC or servo drive

Match the encoder output type to an interface supported by the PLC or drive. An encoder with SSI or BiSS output cannot be read through a standard A/B pulse input merely by connecting the wires. The system needs a compatible interface among its PLC input and output modules, or an appropriate converter. A matching connector does not imply a matching protocol.

Depending on the model, network interfaces can include CANopen, PROFINET or EtherCAT. Check supply voltage, pin assignment, bit count, Gray/binary format, data order, update time and the required device description file. The same compatibility checks apply to encoder readout devices; the term “encoder reader” alone does not establish compatibility.

How does it differ from a resolver?

A resolver is an electromagnetic position sensor excited by an AC signal. It produces sine and cosine signals related to shaft angle, which require a suitable resolver interface. A single-speed resolver can provide absolute angular information within one revolution; this capability is not exclusive to digital absolute encoders.

A resolver versus encoder comparison should not assign universal accuracy ranges such as “0.01°–0.1°” and “0.5°–1°” to entire device categories. Resolver accuracy, converter resolution and system errors are separate characteristics. A resolver is not inherently a 10–12-bit device. Choose according to temperature, vibration, interface compatibility and the required measurement performance.

Applications and selection priorities

Robot joints, rotary indexing tables, packaging axes and machines that need position information at startup are typical applications. Feedback in servo systems involves more than reading an angle: the value must be scaled correctly and reach the control loop on time. An encoder mounted on the motor cannot directly measure backlash farther along the transmission.

Consider single-turn or multi-turn operation, accuracy, resolution, maximum speed, interface and environmental conditions together. Encoder couplings and accessories and the mounting bracket are also part of the measuring chain. Shaft misalignment, excessive bearing loads or a slipping connection can make a correctly functioning encoder misrepresent machine position.

Two common questions

Does an absolute encoder never need homing? It can read its own position, but machine zero, direction and scaling may still need to be established during installation. Recheck that relationship after replacing the encoder or changing its mechanical connection.

Will a failed absolute encoder report zero? A fault can cause missing data, inconsistent values or a diagnostic/status indication. A zero reading alone does not prove a fault. The controller should evaluate communication and data validity alongside the position value.