How it works
Optical encoders consist of a light source, a code disc, and a detector. In the case of an absolute encoder, the code disc is imprinted with a symmetric pattern of alternating transparent and opaque arcs arranged in concentric tracks. Each track corresponds to an individual sensing element within the photo-detector, representing one bit of resolution. For each angular position of the disc, the system produces a unique “whole word” output whose bits indicate the presence or absence of light through each track.
Absolute encoders are generally available in single and multiturn versions. Single- turn encoders are recommended for back-and-forth rotary motion in which the shaft never makes a full revolution.
Multiturn encoders are designed for moves spanning tens or even hundreds of revolutions. They consist essentially of two encoders in one housing. A “fine” encoder, fixed to the input shaft, provides the resolution per turn of the device. A second encoder, called the “turns counter” or “coarse encoder,” connects to the input shaft through a gear set. The gear ratio is usually adjusted to increment the coarse encoder one position for each full rotation of the input shaft. In some cases — as in electronic multiturn encoders — a counter circuit is used in place of the second encoder.
General sizing procedures
Absolute encoders are often specified in terms of resolution, the number of bits or words contained in the complete code. For incremental and tachometer encoders, resolution is defined as counts per turn. For absolute single-turn encoders, it is called positions per turn. Multiturn encoders are specified as positions per turn of the input shaft and the number of turns of the internal gear ratio.
The amount of resolution required is a function of the number of positions that must be measured. If a machine needs to measure the travel of a 25-in. lead screw in increments of 0.001 in., for example, then it will take an absolute encoder with a resolution of 25,000 words.
Resolution is usually defined in terms of bits, however. The term “12-bit encoder,” for example, refers to the binary number 212 which is 4,096 decimal. Thus, a 12-bit encoder has a resolution of 4,096 words.
Absolute encoders are also specified on the basis of accuracy and repeatability. Accuracy, which is traceable to the encoding disc, is the deviation between the actual position and the theoretical position of each bit edge. In other words, it refers to the precision of the output signal in relation to shaft position. A good 12 or 13-bit encoder, for example, will be accurate to within 1/2 count on the least significant bit (LSB).
Repeatability, the ability of the encoder to read the same word each time the shaft is in the same position, is defined as the deviation of the actual encoder position between subsequent identical code readings. Although it has no relation to accuracy, it is usually four to ten times better.
Other selection criteria include shaft rating, shaft and bearing seals, output format, and driver type. Shafts, for example, are typically available in standard and instrument grades, and some are installed with watertight seals. Bearing seals come in several grades as well.
Common applications
Absolute encoders are the preferred transducer for applications requiring a true indication of position at all times. If a system is prone to losing power or can’t be returned to a reference or home position, an absolute encoder is the way to go. It’s also a good fit for machines that may be inactive for long periods of time or ones that move very slowly, such as telescopes, cranes, and motorized platforms.
On the other end of the spectrum, absolute encoders are also well suited for high-speed precision instruments because they produce parallel digital outputs. Parallel data can be read quickly, a necessity for sensing and controlling fast moves.
Other applications that work well with absolute encoders are those that involve very fine movements. The ability of absolute encoders to precisely measure angular motion stems from a combination of multiple code tracks and optical sensing, and is available in a relatively compact package, without need of additional (external) support circuitry.
Continue on page 2
Installation and troubleshooting tips
Over 90% of absolute encoder applications use a flexible coupling to connect the encoder shaft to the system. If the stator is flexible, you may use a rigid coupling, gears, pulleys, or a timing belt.
The center lines of the encoder and system shafts should be coincident, creating a single axis. If the shafts are parallel but offset, it’s going to cause additional loading on the shaft bearings as well as the coupling. (A slight offset will not affect accuracy.) If the shafts are not parallel, the encoder will not work correctly.
Another installation concern is end float, a condition where the connecting shaft moves along the axis. Usually, end float is at its maximum when the rotation of the shaft is reversed. Although end float doesn’t affect accuracy, it loads the coupling and, if excessive, may damage the shaft bearings.
In a servo mechanism, system accuracy, of course, is only as good as the weakest link. Although additional encoder resolution may compensate for inaccuracies elsewhere in the system, positioning accuracy depends on the aggregate accuracy of all the components in the system.
Repeatability can also be negatively affected by other system components. Although encoders themselves have no mechanical hysteresis (backlash), other components, such as couplings and gears, often do. Naturally an encoder can’t measure backlash unless the system is reversed, so if you want to compensate for backlash, you need to do it during the design phase.
As for the electrical interface, most problems stem not from encoder malfunctions, but from installations that do not deliver the required amount of power. These malfunctions are almost always intermittent, so they are difficult to isolate.
Line losses are a common cause of inadequate voltage. Usually, it’s a matter of the cable being too long or the wire gauge too fine to carry the required current. If you suspect line loss, try increasing the gauge of the wire or connecting additional power and ground conductors.
Standards and regulations
Like other electromechanical devices, encoders should comply with UL and CSA standards. You’ll also find that many meet additional standards as well.
Encoders that meet NEMA Class 4 and 13 standards, for example, have the highest level of dust and moisture protection. For global use, encoders must meet either IP65 or IP66 classifications, which are equivalent to IEC standards.
Besides defining a minimum level of ingress, the Canadian Standards Association provides approvals for intrinsically safe operation. Cenelec sets similar standards for encoders used in Europe, and also defines minimum requirements for explosion-proof applications.
Encoders bearing the CE mark (Communaute Europeene) are guaranteed to meet electromagnetic compatibility standards.
Historic notes
BEI’s roots in optical encoders go back to the late 1940s, when the Baldwin Piano Co. was developing electronic organs. Trying to reproduce the thick, resonant sound of pipe organs played in European cathedrals, Baldwin engineers developed an optical device that modulates electronic signals by adding harmonics to the fundamental frequency or note.
In the early 1950s, the Army Signal Corps discovered the technology and incorporated it in a two-way radio capable of sending and receiving encrypted (scrambled) messages. Uncle Sam then commissioned Baldwin to develop a machine to mass-produce the rotary code wheels necessary for optical signal modulation.
Working with the National Bureau of Standards, Baldwin engineers spent two years perfecting their design, and by 1953 had built two machines that could produce code wheels with patterns accurate to within a fraction of an arc-second. The “divided circle” machines, as they were called, were among the first to combine air-bearing, optical, and electronic technology.
Once the Army produced all the code wheels it thought were needed, it sold the machines back to Baldwin Piano. Subsequently, the musical instrument maker opened a plant in Little Rock, Ark., where it began manufacturing rotary optical encoders for position and rate feedback applications. The military, not surprisingly, was the first customer, using the encoders in tracking stations and pointing systems, but commercial uses soon followed in machine tools and reprographics.
In 1973, the Baldwin Piano Co. divested itself of the electronics division, which by then, was known as Baldwin Electronics Inc., and is now known as BEI.
Continue on page 3
Cost saving strategies
Observing the following guidelines during system design can save time as well as money.
• Flex-couple the rotor to the stator.
• Provide for torsional windup.
• Simplify service by providing access to system modules whenever possible.
• Consider inertial loads and torsional vibration resonances in the servo system.
• Design the system so as not to require excessive force to remove or replace components.
• Provide a way to zero the system after the encoder is mounted.
• Install the encoders for best possible alignment. Misalignment can cause errors and shorten bearing life.
Failure modes
Encoders are rugged devices built for years of use. If there is a problem, it will most likely come from gradual bearing wear or the light source reaching the end of its life expectancy.
The best indicator of bearing wear is often decreased accuracy. Of course, improper installation can shorten bearing life dramatically.
Most LEDs have a rated life of 100,000 hours, with many lasting ten years or longer. Failure typically diminishes light output, which will result in an intermittent sensor signal.