Sensors - November 2000 - Incremental Optical Encoder

An interferometric technique allows an incremental optical encoder to read light signals transmitted through slots as narrow as 1 micron on the code disc, greatly increasing precision of the device.

The incremental optical encoder has changed little over the past 50 years. Conventional incremental encoders (see Figure 1) basically consist of a light source; a stationary mask with a spoked pattern of clear slots that shutters the light through an identical pattern on a rotating disc; a photoelectric diode on the side of the disc opposite the light source; and a signal processor.

Figure 1. In the conventional incremental encoder, light passes through an optical mask before striking the patterned encoder disc. As the disc rotates, the light is interrupted and the photodetector receives half to zero light for every slot the disc moves.

As the disc rotates, light either passes through one of the slots or it falls between two slots and is blocked. The flashes of light passing through are detected by the photodetector and interpreted by the signal processor as ones; absence of light is read as zeros. The signal thus produced is a square wave pattern corresponding to a series of voltages proportional to the angle of rotation of the shaft on which the disc is mounted. Because the processor must know the shaft’s previous position in order to calculate its new position, a single reference slot is usually placed on the outer edge of the code disc for use in re-zeroing after a power glitch.

There have been improvements in the techniques used to manufacture the encoder discs, and modern components such as LEDs and differential-reading photo elements have made the devices more reliable. Nevertheless, a standard encoder’s threshold for precision is still determined by its ability to accurately detect light traveling through the 5000–10,000 slots, each with a width of ~10 microns, in the code disc. The physics of refraction dictates that light cannot pass straight through narrower slots. The simple geometrical model of energy propagation does not work with small and narrow slots. There are deviations in the immediate proximity of the shadows, manifested in the dark and bright bands known as the diffraction fringes. Efforts to electronically multiply the signals from these lines are encouraging but still lack precision.

An Interferometric Technique

Figure 2. Two waves are created on two side-by-side slots. Interference occurs when the waves' peaks meet through the slots The meeting points of the wave peaks are the interference points.

The physics of interference underlying a new encoder technology can be illustrated by the behavior of water in a bucket (see Figure 2). Two sources of disturbance create two sets of concentric waves that form an interference pattern at points where the waves overlap.

In optics, interference requires an arrangement by which light from a source is divided by an apparatus into two beams that are then superposed. The intensity in the region of superposition is found to vary from point to point between maxima, which are greater than the sum of the beams’ intensities, and minima, which may be zero.

To apply the phenomenon of interference encoder technology, three problems must be solved:

One-micron slots won’t work with a standard encoder mask because of diffraction, a physical phenomenon that causes light to go around the slots. Actually, diffraction always exists. But in the case of small sizes it becomes quite significant and destroys the shadow pattern beyond the mask.

To form a straight-line interference pattern, two flat waves are required. Even a very small departure from absolute flatness in the wavefront causes the lines of the pattern not to be straight and parallel. Precision is therefore a must for the mirrors and the prism, or beam splitter.

D	= period of the interference pattern distance between lines or their width
	= wavelength of the light
	= angle at which the two waves cross each other

For an encoder with a different resolution, the pattern’s period, D, must be adjustable. To make the line width adjustable, it is necessary to change

. The solution here is to use two mirrors.

Figure 3. To ensure stability, all optical components-waveplates, beam splitter, mirrors, and light source-are mounted on one sturdy torsion-free aluminum body.

The beam splitter was engineered to eliminate speckles and energy losses and the make the whole encoder design immune both to dust and to imperfections in the various components of the final project.

In the new encoder, light from a monochromatic point source passes first through a collimator that sends a coherent beam of light through an aperture large enough to cover the sensitive areas of the four photodetectors. The light then travels through a beam splitter that divides it into two beams of equal intensity and sets them at the correct angle to form an interference pattern. The width of the beams and the separation between any two neighboring beams match the pattern on the code disc placed in the area of interference.

Between the collimator and the beam splitter are two mirrors that can be used to adjust the period of the interference pattern, which is the distance at which the pattern maximas are located. A polarizer placed after the beam splitter serves to increase the contrast of the interference pattern by equalizing the polarization state of the two beams. The entire assembly is shown in Figure 3.

Figure 4. The direction of the lightwaves illustrates the way the light splits and comes back together, creating an interference pattern.

Figure 4 illustrates an incremental encoder using a code disc with four rows of light-absorbing lines shifted ¼ period to one another. An array consisting of four photodetectors senses the intensities of light passing though these four areas of the encoding disc. Therefore, light signals coming onto the photodetectors are also phase-shifted ¼ period. Two groups of two ½ period signals of the photodetectors are used in complementary amplifiers to produce ¼ period-shifted outputs, as is most commonly required of incremental encoders.

The interferometry-based technique produces adjustable parallel lines of light as small as 1 micron (see Figure 5), allowing 32,768 lines to be placed on a single encoder disc.

Figure 5. The interferometric technique creates fine parallel light stripes in very short distances. Contrast this pattern to that in Figure 1.

Figure 6. The two signals arriving from the photodetector are phased so that a clockwise or counterclockwise direction can be detected by an electronic controller.

When the disc is moving, the light intensity at the photodetectors changes accordingly, and counting the signals in both directions allows bidirectional motion detection (see Figure 6).