Quality 101: Encoders, Decoded
Understand absolute position encoder feedback.
Position encoders are ubiquitous in myriad types of equipment. Encoders are found in applications as diverse as high tech medical equipment (surgical robots, cancer treatment machines), military and aerospace (pilotless drones), semiconductor and electronics manufacturing machinery, and computer numerically controlled (CNC) machine tools, to name just a few. Indeed, many types of modern automated machinery would not be possible without encoder feedback. Dimensional metrology equipment, such as CMMs, length gages and roundness gages, all incorporate precision encoder feedback.
Position encoders have been around for decades. The styles and types of encoders have evolved to meet the various application demands. There are many competent encoder manufacturers serving the market, providing a wide array of choices. However, one aspect has stayed fairly constant over time: the output signal format. The advent of a new type of position encoder is changing even this aspect.
The Incremental Encoder
Those who are familiar with encoders probably recognize the common terminology long associated with the encoder output signal format: A quad B, TTL encoder, square wave encoder, digital encoder. Despite the different terms, they all refer to essentially the same thing, namely an incremental quadrature output encoder. Engineers specifying encoders could be confident that the various types of electronics that an encoder is interfaced to would be able to receive and correctly interpret the signal (i.e. count the pulses coming from the encoder), regardless of the style of encoder or application. This is because the encoder output signal formats from nearly all manufacturers were essentially the same. Whether it is a digital readout, an embedded motion control processor card, or a CNC, they pretty much all support an incremental quadrature encoder input.
The Move to Absolute Feedback
With the encoder industry settled on a de facto output signal standard, one may wonder why anyone would want to change this. It turns out there are very good reasons. But first it is important to understand that an absolute encoder provides an output signal in a completely different format from an incremental encoder, meaning that the receiving electronics needs to be able to accept this new type of signal. Moreover, there is not yet a widely adopted de facto standard for the signal characteristics. So absolute encoders from different manufacturers are not electrically interchangeable. Engineers need to be careful to assure that their receiving electronics support the encoder they have chosen.
To complicate matters further, some encoder manufacturers have chosen to adopt their own proprietary output signal format (referred to as a proprietary protocol), assuring incompatibility with encoders from other manufacturers. Others have recognized that a proliferation of proprietary protocols can effectively limit consumer choice and create difficulties for developers of receiving electronics, who have to support multiple incompatible protocols and negotiate separate licensing agreements with each protocol owner. These encoder manufacturers have chosen to adopt open source protocols that allow anyone to develop compatible devices supporting the open standard, free from onerous licensing restrictions. The goal is that the encoder industry will eventually coalesce around a new de facto standard for absolute encoders that everyone can use, leading once again to interchangeability and widespread industry support. The most promising of these open source protocols, and the one that appears to be gaining the most traction in the market, is the BiSS protocol. Several leading encoder manufacturers have adopted the BiSS protocol, as well as dozens of manufacturers of interface electronics.
Absolute encoders offer several distinct advantages over incremental encoders. Moreover, the prices of absolute encoders have finally in many cases achieved parity with comparable incremental encoders. This is leading to the rapid adoption of absolute feedback, benefiting machine developers and users alike.
Absolute encoders provide an output that represents the exact position of the machine at any location in the travel. The output is a (binary) number corresponding to the actual position value encoded on the scale at that precise location. This frees the receiving electronics from having to count and keep track of incremental encoder pulses. The receiver simply requests the current position from the encoder (when it needs to know it), and the encoder responds with the position value (i.e. a data word). The data word coming from the encoder is in real time, meaning it is perfectly suitable to high performance servo control systems.
An absolute encoder always knows where it is. There is no need for a reference mark, so absolute encoders do not have one. An incremental encoder requires a home position or reference mark to establish a datum location for subsequent pulse counting. The receiving electronics count these incremental pulses to determine the location relative to the home position.
An absolute encoder knows position immediately upon power up. If you remove power or lose an incremental encoder signal for any reason, you must re-home an incremental system. This can affect cycle time, and maybe result in costly scrap material.
Note that a true absolute encoder does not require any battery back up for storage of position data. The absolute information is encoded directly onto the scale. This is different from a quasi-absolute encoder that can be merely an incremental encoder with an internal counting circuit and a battery to store the position when external power is removed.
Absolute encoders cannot miscount. For this reason, they are more robust than incremental encoders. An incremental encoder feedback system can be susceptible to “miscounting,” which refers to the loss of (or erroneous accumulation of) position pulses. There are many potential causes, including electrical noise, contamination on the encoder scale and electronic interface problems. Regardless, the result is a position error with potentially costly or even dangerous consequences.
Absolute encoders are capable of both very high speeds and extremely fine resolution. Resolution as fine as 1 nanometer and speeds above 10 meters per second are simultaneously achievable in linear systems, and 0.0003 arc-second and above 10,000 rpm in angular systems. The speed vs. resolution tradeoff present in incremental systems does not exist with absolute encoders-a critical breakthrough in performance.
Absolute encoders have easy alignment and setup tolerances, comparable to incremental encoders. New technology “single track” absolute encoders are as easy to setup and align as their incremental predecessors. Some older style, “dual track” absolute encoders do have demanding alignment tolerances limiting the range of suitable applications, but newer generation offerings do not have this limitation.
The selection of absolute encoder packages and styles is comparable to the availability of incremental encoders. There is now an absolute encoder for nearly every application. In the early days selection was limited, prices were high, and only a few manufacturers offered absolute encoders. Today, the situation is completely different.
Absolute encoders have achieved price parity with comparable incremental encoders. More selection, wider market adoption and more suppliers means prices are dropping. There is no longer a big cost penalty to go absolute.
Because new generation absolute encoders use advanced optics and digital signal processing, they are more resilient to contamination compared to similar incremental encoders. And status bits aid reliability. The absolute position data word can include additional information such as status and error bits, alerting the control system or operator to conditions requiring attention. An incremental quadrature signal contains only the pulse train.
Absolute encoders can be more suitable to intermittent usage, battery powered applications. Since position is available immediately upon power up, they can be shut off when not in use, then momentarily powered up to take a reading. An incremental encoder needs to remain powered, or be re-initialized (homed) when powering up.
As the encoder market continues the transition to absolute systems, machine builders will employ this exciting technology in more applications. Both well-established incremental encoders and newer technology absolute encoders will coexist in the market for the foreseeable future, but with absolute feedback becoming increasingly prevalent. It is important for engineers to understand the relative merits of the different encoder sensing principles to be able to make the best choice.