Testing and Encoder Background Information

The Theory of Operation section will provide technical details concerning system hardware and software. The information in this section is designed to provide detail beyond operating instructions to help the user obtain the maximum benefit from the system. 

Terminology

This section will attempt to explain various terminologies such as serial, absolute, incremental and so forth. Unfortunately, often the same terms will be used in a different context by different manufacturers, but this section will at least describe what is meant by these terms in reading this manual. 

INCREMENTAL VERSUS ABSOLUTE 

There is considerable confusion over the use of the incremental and absolute terminology. This is understandable considering the evolution of encoder technology. Hopefully this section will clarify rather than further muddy the waters. At least it will provide a basis for understanding the use of the terms incremental and absolute in this manual.

TRADITIONAL ABSOLUTE 

A traditional absolute encoder will have numerous code rings with different binary weightings which will provide a digital data word representing the absolute position of the encoder within one revolution down to the precision of the least significant bit. This is the type of encoder that we call “Parallel Absolute” in the encoder selection. A simple example would be 3 code rings in which the most significant bit is HI during 1/2 the turn and LO during the other 1/2 turn. The next code ring goes from LO to HI two times during a turn while the least significant bit switches from LO to HI 4 times as shown in the following table:
 

There are 8 unique data words to represent the 8 positions which can be reported by this encoder. Adding more bits will divide one revolution into smaller segments, and the encoder will therefore provide higher resolution. It will resolve one rotation into finer segments, but at the expense of adding more output lines to represent the additional bits. Typically this type of absolute encoder will be offered with a gray code output rather than the straight binary code shown above. Gray codes are designed to give the same information, but only one bit changes at a time when moved from one segment to the next. In the above example, all bits change state when moving from segment 7 to segment 0. Due to electronic delays, all bits do not change at the same instant in time, so there may be a short time in which the bits do no correctly represent the position. The gray code eliminates this problem.

The distinguishing feature of the absolute encoder is that the encoder reports the absolute position of the encoder to the electronics. It can do this immediately upon power-up with no need for indexing.

TRADITIONAL INCREMENTAL

A traditional incremental encoder works differently. It provides an A and a B pulse output as shown in the table above. These pulse outputs really provide no usable count information in their own right. Technically they provide a 2 bit gray code which can count from 0 to 3, but this is not how it is used. With the incremental encoder, the counting is done in the external electronics. The beauty of the A/B quadrature pulse pattern shown above is that the change in state can be counted by the electronics, and the direction of the count can be determined by which changes first: A or B. One of the advantages to incremental encoders is that there are always only two lines, A and B (and sometimes their complement lines A* and B*) for any number of counts per turn resolution. The disadvantage is that the point where the counting begins depends on the counter in the external electronics and not on the position of the encoder. Usually the counter is zeroed on power up, and it begins counting from there as the encoder is turned. The encoder could be at any position when power-up occurs. The motion of the encoder simply causes the external counter to increment (or decrement) the count up or down as it moves, and the count is not referenced to its absolute position.

In most cases, the incremental encoder must be referenced to the absolute position to be useful. One way this is done is to include an index pulse on an additional line, Z, which will go HI at one point during the encoder revolution. On power-up the encoder can be turned during a "homing" sequence until the index pulse is found by the external electronics. If the external electronics zeros the counter (or sets it to some other known count), the encoder now becomes referenced to the encoder's absolute position. Homing can also be accomplished with reference to proximity switches and other signals that indicate a reference position.

The distinguishing feature of the incremental encoder is that the encoder reports an incremental change in position of the encoder to the electronics. The electronics must

INCREMENTAL/ABSOLUTE GRAY AREAS

As noted above, the absolute encoder presents the absolute position to the electronics, while the incremental reports to the electronics that it has moved through some increment. The downside to the incremental is the requirement for initial homing to provide absolute position information. What if we worked on our incremental encoder system so that we could remember the homing information through a power cycling, and homing during power-up was not required? Could we be justified in calling our incremental encoder an absolute encoder? Fanuc apparently takes this point of view because they use “absolute” as their terminology for A,B, and Z pulse encoders with battery backup. If the encoder and external counter is powered by a battery so that it truly can provide absolute position information on power up, it is functionally very similar to an absolute encoder. Of course if you want to make connections to this type of encoder and test or otherwise evaluate it, you find that for all practical purposes you have to consider it an incremental encoder because it looks nothing like the traditional absolute encoder described above.

SERIAL VERSUS NON-SERIAL

Serial encoders are encoders that transfer their information to the electronics in a serial data stream. Any of the counting schemes discussed above could transfer their count to the electronics using serial data. The traditional absolute encoder described above would have 16 data lines if it provided 16 bit resolution (counts up to 65535). If each data bit used a true and complement line for noise immunity, this would be 32 lines. The term normally used for this kind of data is "parallel". The Centronix printer printer interface (at one time used on most PC computers) transfers data in a parallel manner, 8 bits at a time. It is fast at the expense of a lot of wires. Serial data provides for the transmission of the data one bit after the other rather than all at once. All 16 bits mentioned above could be sent on one data line, but it could take 16 times as long (all other things equal). Many serial encoders have a request line to interrogate the encoder for data, and a data line for receiving the data from the encoder. If true and complement lines are used in each case, only 4 wires plus power and ground are needed for any resolution.

From the standpoint of trouble-shooting, serial encoders present several problems. Often the encoder does not send any data unless it is properly interrogated by the request line. The bit pattern that appears on the data line after interrogation is not simple to interpret and is constantly changing as the encoder moves. It is a complex problem to correctly read and evaluate this information, and it is beyond the capabilities of standard universal test equipment. The TI-5000JX provides this information in an easily interpreted format which makes evaluating serial encoders practical.

The term serial encoder is, of course, combined with incremental, absolute, etc. depending upon other details of its design and construction. Just remember that the serial terminology refers to the fact that the information is sent from the encoder to the electronics in a serial fashion, and that is independent of other terminologies applied to it. For instance, the Fanuc Alpha encoders are serial. The Alpha I model is called incremental because it has no battery, but the Alpha A model is called absolute because it has battery backup capability.