What is a pulse generator

2 pulse generators

Started Rotary encoders and linear encoders are used to record the direction of movement, speed, path and position of moving objects.
Precision, dynamics, handling and economic efficiency are the parameters which, when weighed up, have resulted in a large number of types and designs.
SummaryRotary and linear encoders are used for recording direction, speed, travel and position of moving objects.
Trade-off among parameters like precision, dynamics, usability and cost-effectiveness have produced a variety of types.

2.1 definition

Pulse encoders can be divided into two categories: Rotary encoders (angle encoders, angular incremental encoders) detect the angular position of a rotary axis, linear encoders the position for a linear displacement path. Both provide the quantized information on 1 to n binary channels for further processing.

2.2 Use

Depending on the application, a downstream information processing system determines from this
  • Angular path, angular speed, direction of rotation and angular position or

  • Linear path, linear speed, direction of movement and linear position.

A large number of specialized designs have been developed for the various areas of application. The following description is restricted to rotary encoders. Most of the principles can, however, also be applied to linear encoders.

2.3 Technology

The historical development and the different requirements of various applications led to a large number of technical embodiments:

2.3.1 Mechanical encoders

If you connect the shaft of a mechanical rotary switch with the shaft to be measured, you get the simplest version of a rotary encoder with mechanical sliding contacts. Even if further developed designs were trimmed for higher speeds and longer service life, this type no longer meets today's requirements and is therefore becoming increasingly less important.

2.3.2 Magnetic rotary encoders

Magnetic designs mostly work with Hall sensors and an arrangement of permanent magnets. They are used to advantage in manual operation.

2.3.3 Inductive encoders

They are used primarily in unfavorable environmental conditions such as dirt or vibrations, e.g. in a car. Metallic encoders influence the quality of coils. Sinus-modulated high-frequency oscillations are obtained as output signals, which can still be evaluated even when the machine is at a standstill.

2.3.4 Electromagnetic encoders

"Resolvers" use transformer coupling between the stator and rotor windings to generate sinusoidal signals whose frequency is proportional to the speed of the rotor. Since very slow movements can hardly be evaluated, these types were and are mainly used for speed control. However, they are becoming increasingly less important. Types with several phase-shifted output signals are also referred to as "resolvers".

2.3.5 Light barriers

Optoelectric encoders are the most common. They contain one or more fixed light barriers. In the transmitted light version, the light transmitter and receiver face each other in the form of a fork light barrier. Between the two there is a round code disk with translucent (light) and opaque (dark) sectors (Fig. 2-1). This code disk is directly or indirectly connected to the shaft whose angular position is to be determined. When the shaft rotates, the light beams are transmitted or interrupted by the sectors of the code disk. The photo receiver converts the modulated light received into current pulses. The reflected-light version uses retro-reflective photoelectric sensors in which the transmitter and receiver are on the same side of the code disk, which in this case contains reflective and non-reflective sectors. The following descriptions all relate to the transmitted light version, although most functions can also be implemented with reflected light.
If the width of a sector is large compared to the beam diameter of the light barrier, it is trapezoidal; if it is small compared to the beam diameter, sinusoidal output signals are obtained (Fig. 2-2). For standard applications, both are given a rectangular shape by the following comparators.

2.4 Type of drive

Pulse generators for manual operation are available with resolutions from 10 to around 100 steps per revolution. For a relatively rough selection of table positions, a pointer is moved over a screen menu. Mechanical or magnetic designs are preferred for this because they are available with a pronounced detent and thus prevent letting go at an intermediate position. For more precise position determination on the screen, optoelectronic versions are mostly used in pairs in mice (Section 4.4.3: Mouse) and in trackballs (Section 4.4.4: Trackball).
In the case of controlled drives (speed or position control), motor-driven pulse generators are connected to the drive axis and used to record the actual status. For this application, optoelectronic designs are preferred, which enable significantly higher resolutions with smaller friction losses.

2.5 Basic function

2.5.1 Distance and speed measurement

The simplest application is the determination of (relative) paths. It is sufficient to count the pulses (light-dark pairs) of a channel of an incremental pulse generator (Fig. 2-2):

Distance / angle and angle / impulse are due to the design and are therefore known.
In connection with a time measurement, you get a statement about the speed:

With fast movement you count the pulses per unit of time, with slow movement you measure the time between successive pulse edges.
If you also want to recognize the direction of rotation (direction of movement) (e.g. right / left, forwards / backwards), you need two channels of an incremental pulse generator with electrically phase-shifted output signals (preferably 90 °). When the pulse disc is turned clockwise, the light-dark edge of channel 1 meets the light state of channel 2; when turned to the left, it meets the dark state.
An evaluation circuit consists of a direction detection (e.g. a latch) and a counter that counts up when the forward direction is detected and down when the backward direction is detected. In the simplest version, for example, one counts the positive impulses of a channel. With a little electrical effort (exclusive-OR link between the two channels), with the same geometry, the total resolution (count values ​​/ angle) can be doubled or quadrupled if all edges of the two channels are evaluated.

2.5.2 Position measurement

The evaluation counter of an incremental pulse generator can assume any value when the supply voltage is switched on. In order to be able to use it for the measurement of absolute positions, a certain zero or reference position must first be approached, the reaching of which must be determined by other means. At this position the counter is set to a defined starting value. Some pulse generators already deliver a reference pulse via a third channel. For this purpose, they contain another ring on their pulse disc with only one light-dark and one dark-light transition. During assembly, the pulse generator must be adjusted so that the reference pulse occurs at the defined reference position. If the measuring range covers more than one revolution of the pulse disc, the reference pulse becomes ambiguous. It must then be made clear by other means, e.g. by a second pulse generator that moves one step with the help of a gear while the first makes a full revolution.
If this procedure seems too uncertain or if the absolute position has to be known immediately after switching on or after a power failure, a (much more expensive) absolute pulse encoder is required. In the basic type, its pulse disc contains n sector rings that are scanned by n light barriers. The measured value for the absolute position is transmitted with n bits in parallel to the evaluation circuit. The resolution that can be achieved without pull-ups is 2048 steps per revolution. The use of a Gray code, in which two channels never change their state at the same time, reduces the required mechanical precision and thus the costs (Fig. 2-3).

2.6 Special types

2.6.1 Double step encoder

They are also called double track encoders and each contain a sector ring with a coarse and one with a fine division (on 1 or 2 pulse disks). The coarse ring is evaluated for fast movement and the fine ring for slow movement.

2.6.2 Difference track encoder

They contain two complementary sector rings that are irradiated by a common light transmitter, but contain one receiver per ring. If the two output signals are routed to the inputs of a comparator, a symmetrical square-wave signal is obtained without additional adjustment effort. Aging effects of the light source are also compensated for.

2.6.3 Double reference pulse generator

If the light sector of the reference channel is dimensioned in such a way that it corresponds exactly to the required measuring range, the beginning and end of the measuring range can be used as reference points.

2.6.4 High-resolution incremental encoders

To increase the resolution, the line width of the coding disk can be reduced. Since the opto-receivers cannot be reduced to the same extent as desired, you will eventually get to areas where the line widths are small compared to the active area of ​​the receiver. There are therefore always several light and dark sectors in the beam path. With the principle shown in (Fig. 2-2), the output amplitude would be very small. A fixed "phase disk" is therefore introduced with the same line pattern as on the coding disk. When rotating, all sectors in the beam path are faded in or out at the same time.

2.6.5 High-resolution absolute encoders

With conventional absolute encoders, one opto-receiver and one track on the coding disk are required for each bit. With resolutions of 4096 steps (12 bit) / revolution and higher, the coding disks can therefore become unwieldy. This can be avoided by mapping a cyclic pseudo-random code (PRC) onto a track on the disk. The transmitter-receiver pairs (as in the standard version n pieces for n bits) are now arranged along this track (principle see Fig. 2-1, last line) at a mutual distance of one sector width. The Opto-Asics required for this are state of the art for absolute encoders anyway. However, the dimensions of the Asics used determine the diameter of the coding disk. The advantage is bought by giving up the Gray code. Tolerance problems could now arise again with simultaneous light-dark transitions. For this, however, an additional incremental track is introduced, with which one can not only determine the optimal sampling time for the PRC track, but also obtain further resolution bits via AD converters. A practical implementation of this principle works with two antivalent PRC tracks (differential track transducers) with a resolution of 12 bit / revolution and an additional incremental track that provides a further 6 bit resolution. The result is an 18-bit absolute encoder in a fairly compact design.