Gnuman requires 3 position encoders and 3 velocity encoders. The velocity encoders are already on the robot and use incremental encoders. As for the position encoders that require an absolute position (independent of power cycles), Incremental encoders with a home/zero position make the most sense. While Synchros, Resolvers, and Absolute encoders would suffice, the cost can be as much as 7-8 times more expensive. The benefits of having a Resolver/Synchro able to operate in extreme conditions, be more durable, and more accurate are offset by the price. Likewise, while Absolute encoders provide digital position data which can be processed more easily, the cost makes it impractical.


Position control for motors (Rotary Encoders)


A rotary encoder is an electromechanical switch used to measure angular position and angular motion. Its output is generally composed of a digital encoding of relative or absolute position. Specialized rotary encoders called synchros actually output into sinusoidal waves where the phase difference represents the angular position.






A Synchro is a generic term for a family of transducing elements. These transducing elements can be arranged in various ways to measure angular position. The synchro works on the rotating transformer principle to generate secondary currents from a primary sinusoidal current.


The rotor is an extension of the shaft. Three windings are attached 120 degrees apart and are connected to the terminals S1, S2, and S3. This type of connection is known as a wye connection. The windings from the rotor are connected by slip rings and brushes to the terminals R1 and R2.


An AC voltage, known as the reference voltage (usually either 60Hz or 400Hz), is sent through the rotor winding and subsequent voltages are induced in the stator windings proportional to the cos angle (f) between the rotor coil and stator coil. .The voltage induced across any pair of stator terminals will be the sum or difference, depending on the phase, of the voltages across the two coils concerned..


Reference voltage: A Sin wt

The voltages which will appear across the stator terminals will be:

S1 to S3 = A sin wt Sin f

S3 to S2 = A sin wt Sin (f + 1200)

S2 to S1 = A sin wt Sin (f + 2400)

where f is the synchro shaft angle.





The Resolver is a type of synchro, sometimes called Synchro Resolvers, in which the windings on the stator and rotor are at 90 degrees to each other instead of 120 degrees as in the case of the general synchro. This allows the Resolver to take advantage of the sinusoidal relationship between output voltage and shaft angle. Resolvers come in many forms with a wide variety of winding configurations and transformation ratios. The simplest Resolver would have a rotor with a single winding and a stator with 2 windings at 90 degrees to each other.




AC reference voltage: A Sin wt


Then the voltages appearing on the stator terminals will be:

S1 to S3 = V Sin wt Sin f

and S4 to S2 = V Sin wt Cos f

where f is the Resolver shaft angle.



Optical Encoders


Incremental encoders


Incremental encoders consist of a disc, light source, and light detector. The disc is made up of evenly divided alternating light and dark sectors (ranging from opaque to transparent) and is attached to the shaft. The light source is placed on one side of the disc while the light detector is on the other side. As the disc turns (upon given rotation from a motor), the detector will detect on or off depending on which part of the disc is in between the light source and detector. From this, a square wave pulse stream is produced. The sum of the pulses is used to indicate angular position of the shaft and velocity information can also be extracted based on the resolution. The resolution of the encoder is related to the number of sectors the disc is divided into. .The resolution of the encoder is governed by the number of opaque and transparent sectors and usually falls between 100 and 6000 counts for one complete revolution of the input shaft.. Almost all incremental encoders have a second light source and detector which is phased correctly so that the direction of the rotation can be obtained.


Also, many encoders even included a third light source and detector which output only once per revolution making velocity measurements attainable.



.Incremental encoders come in sizes ranging from 1 inch diameter (25.4 mm) to 3 inches

diameter (88.9 mm). They are also available in all types of external construction ranging from plastic, which is suitable for low cost commercial applications, to stainless steel where the required specification is more rigorous. While this type of encoder may be useful in some applications it has the disadvantage of having the angular information stored in an external counter. If the information in this counter is lost (for example if the power supply was temporarily interrupted) there is no way of knowing the shaft angle. Also at initial switch on, there is no way of determining the shaft angle until it has been rotated through the revolution





Absolute encoders


Absolute encoders are very similar to incremental encoders. Here, the disc is divided into N alternating dark and light sectors pertaining to a count (based on the coding scheme). Each count relates to the bit representation of the absolute position. For example, for a 6 bit resolution, there will be 6 counts, one for each bit. Each bit/count has its own light source and detector. A set of light sources are arranged radially on one side of the disc and corresponding detectors are positioned on the other side such that a parallel word representing the input shaft angle can be obtained at any one of the angular positions.



Therefore, the absolute position on the disc, corresponding to the absolute position of the shaft, is always known. Even powering down does not affect the data. .Absolute optical encoders come with resolutions of 6 to 16 bits in Gray code, binary or BCD and their sizes vary from 2 inches (50.8 mm) to about 7 inches (177.8 mm) in diameter..