Incremental encoder working principle

The incremental encoder is a device that can measure a shaft’s angular position. It can detect only the changes in position on the shaft.

The laser light reflects from an optical sensor’s edge or hole in an optical sensor and send to a photoelectric detector. The incremental encoders are available with a different number of encoding bits. It generates pulses with fixed time intervals between them to identify the sensor position, also called pulse spacing (usually about ten µs).

There are two modes for reading out values: absolute and relative way. Moderately, all pulses are stored until there is one less pulse than expected from counting during calibration in receiver mode. In this case, the absolute position is read out. In the other method, the sensor outputs one pulse per frame, counting the work via structures.

In absolute mode, a stop bit (usually 0) is added to each frame to increase accuracy in certain applications.

The incremental encoder contains at least one disk and one or more encoding bars or teeth, producing a stepped pattern that advances as the disk rotates. (An alternative scheme involves using two disks.) One example of this design, quadrature encoding, moves all of these elements except one bar counterclockwise by 1/4 turn.

A disk has a pattern of drops of liquid metal, or other conductive material, resulting in a pie-shaped area of conductivity on each disk. This pattern is repeated several times around the disk.

The bars are held by a carrier, moving them relative to the disk as it turns. The bars have teeth that fit into slots on the page. As the bars turn and move close to the places on their carrier, they create pulses in the electrical signals produced by material changes on that pie-shaped conductivity area.

The electrical signals are amplified and processed using electronics that count these pulses and output an analog voltage proportional to how many steps have been taken (as an indication of position).

How Does An Incremental Encoder Work?

The first graphic shows a single bar rotating 1°. Each bar is divided into 16 segments. (In this example, the individual divisions are not labeled.) The disk is divided into four quadrants, each further into eight steps. The bar shown can be considered in its home position, just before it begins to move.

To detect movement, the laser beam is projected onto one of these segments and reflected by a photoelectric device that detects the beam and generates a pulse every time it is reflected. As the segment shifts, it produces an interrupt in the pulse stream that indicates angular displacement.

If the bar is rotating at a constant velocity, this interrupt sequence will settle into a pattern of pulses, as shown in the following figure. Each pulse height is proportional to the angular displacement ( e.g., at each pulse height, after 1° of rotation, there are five pulses).

When the steps are read out and processed in order, they produce an output that encodes only changes in position (in this case, one full revolution). Because all steps make equal contributions to the final position reading of any given segment and because only one new feature is activated each time a bar travels its full distance around the disk, it follows that routine steps do not interfere with one another.

In a device designed to detect displacements as small as 0.2 mm, the number of pulses per revolution must be known precisely. Therefore, the incremental encoder must be calibrated by rotating the disk while monitoring the position of a reference step. This is done by measuring and storing the number of steps whose pulses are received in a given period (typically 25 ms) while the disk rotates at some constant speed (e.g., 200 rpm). The resulting value is called “pulse train length.” A similar procedure defines “pulse interval,” or the amount of time between successive pulses in a pulse train.

Incremental Encoder Types

There are different types of incremental encoders. A few common ones include:

The most common type is the quad-concentric encoder, which features a rotating ring with four teeth. As the teeth pass by a stationary sensor, they form a pattern that consists of two complete revolutions, two partial revolutions, four divisions and eight steps per revolution. This design produces 2n – 2 transitions per revolution – one for each tooth plus one for the ring itself. All of these transitions must be recorded to encode a step by the encoder.

Quadrature incremental encoders can be found in some industrial controllers, such as on multi-rotor drones, which need to track the rotational velocity and acceleration of the propellers.

The SN75330B is a series with ten different sensors in an H-pattern. The encoder output is a 5-bit binary.

The TSI 3200/6300 series are based on the TSI 6319 incremental encoder chips with pulse widths as wide as 1 ms between pulses. Depending on the chip version, they can be used with 8 bit or 16 bit resolution.

The TSI 3200 series are available up to an encoder count of 2048, and the TSI 6300 series have a range of up to 4096.

The outputs of these chips pulse at least every 512 steps, and each step is 1/512 of a circle (360 degrees divided by the step angle).

The TC10 series have superseded LSI’s TEC10 series incremental encoders. The TC10 contains a built-in ROM. In addition, they output both direction and rotation information on their Q-channel outputs. The Q-channel outputs are useful for applications where the direct K-channel outputs will not be used (e.g.


Here, Tech Intuitions Meet Practicality. Find tech guides, tricks, tweaks & Mods to all matters mobile phones, computers, software, games. OS & firmware upgrade guides and purchase guides.