Design Feature: September 2, 1996

A wide variety of automatic equipment, from office copiers to semiconductor-wafer separators, depends on photoelectric sensors to detect and count targets and to provide position information. Such sensors often must perform in cramped spaces and in dirty or hazardous environments where there are wide swings in temperature. Many sensors are available to handle such wide-ranging applications and to satisfy even the most stringent demands.

Choosing the best sensor for your application from among the different types requires care and some basic background. Underspecifying sensors risks system reliability and may cost your project dearly if production lines come to an abrupt halt. On the other hand, overspecifying during the design stage squanders costs, which can ultimately lead to a rejection of the system in favor of an alternative. Although many applications involving photoelectric sensors are relatively straightforward, it is generally helpful to chat with a vendor's application engineer to choose the best sensor.

Review basic photoelectric configurations

A photoelectric system basically includes an emitter (transmitter) and a receiver (photodetector). The emitter is usually an invisible infrared LED that directs pulse-modulated bursts of light energy toward the target. When the receiver detects the presence or absence of light energy (depending on the receiver's operating mode), the receiver's output state changes and triggers programmable-logic controllers (PLCs).

You can use one of three basic configurations to detect and count targets: through-beam, retro-reflective, and diffuse-reflective (also called proximity). You can also use a fourth arrangement, optical-triangulation technology, for position detection. Thru-beam sensing, which uses separate housings for the emitter and receiver, is appropriate when the target distance is more than 30 ft from the sensor and the environment is dirty or polluted. Retro-reflective systems are suitable for distances of 30 ft or less. You can use proximity systems when spacing is 6 ft or closer. You can use Table 1 and the flow diagram in Figure 1 to determine which photoelectric configuration is best for a particular application.

Figure 2a depicts a typical through-beam arrangement. A target passing between the emitter and receiver interrupts the beam. The resultant "dark" condition causes the receiver's output stage to change to its on state. The range of operation of through-beam systems extends to 600 ft with opaque and reflective targets. For reliable through-beam operation, the target should be larger than the diameter of the emitted beam, and the optical alignment should be exact, so that the target completely interrupts the beam. A drawback to through-beam arrangements is the need for precision optical alignment, costly wiring runs, and two separate mountings and power supplies. A through-beam arrangement designed for long-distance sensing can perform improperly if you use it for closeup sensing without sensitivity readjustment. Excess sensitivity can pick up light reflected by adjacent surfaces in addition to the intended targets and cause false triggering.

Unlike through-beam systems, retro-reflective systems require only one housing for the emitter, receiver, and single power source (Figure 2b). The emitter directs its beam to a retro-reflector, which bounces the beam back when a target is absent. Similar to thru-beam systems, the receiver changes to its on state when the target interrupts the beam. A retro-reflector, such as the reflector in Figure 2b, reflects back light energy even when the sensor is not absolutely perpendicular to it. You can use a mirror to bounce the transmitted light back to the receiver, but a mirror directs light away at an angle equal to the angle of incidence. Common retro-reflector devices are highway signs, which collect light energy from a variety of sources and redirects the source to the drivers' line of sight to display messages.

Because they are easy to install, retro-reflectors are popular for use on production conveyer lines and for packaging equipment. However, these systems are not recommended for applications that involve grime, haze, or highly reflective targets. Retro-reflective systems are not suitable for small-target detection, because the reflector should be smaller than the objects being sensed. Retro-reflective polarized sensors are available for shiny, highly reflective targets, such as aluminum cans and foil.

Like retro-reflective devices, proximity, or diffuse-reflective, arrangements require only a single housing for the emitter and the receiver. However, unlike the retro-reflective scheme, diffuse-reflective arrangements do not require a reflector (Figure 2c). Energy received from the transmitted light beam results from reflections coming off the target as it moves by. With retro-reflective and through-beam systems, the absence of a target allows light energy to pass from emitter to receiver. However, in this case, the sudden presence of a target interrupts the beam, and the receiver changes to its on state. This form of operation, in which light reflects onto the receiver when the target is present, is called the light-on mode. Because the diffused or scattered light provides a weak signal compared with through-beam or retro-reflective systems, high-gain amplifiers are necessary, and the range is limited to only several feet.

Diffused-reflective systems are well-suited for detection of transparent and translucent objects, such as empty or filled bottles. Target size, texture, color, dust, and humidity all influence the amount of the reflected diffused light. In addition, a reflective surface should not sit behind the target. To improve proximity-reflective-system performance, you can apply convergent-beam background suppression. This technique optically confines the sensing to a precise window in front of the assembly. Convergent-beam background suppression also minimizes the influence of background reflections and the effects of differences in target color and texture.

Use triple beams for position detection

Triple-beam technology, which is relatively immune to dirty or polluted environments, can detect targets ranging from deep black to transparent. An optical-triangulation, also known as area-reflective, configuration (Figure 3) includes both projection and receiver lenses, as well as a position-sensing diode. When the target is in position A, the reflected light produces a spot at point a on the sensor diode. When the beam moves to point B, the beam strikes the diode at b. Unlike a conventional photodiode detector, in which the output is proportional to the light energy impinging on the junction, the position-sensing-detector output is proportional to the distance between the two spots that the reflected light strikes. Today's popular automatic-focus cameras employ this technique.

In triple-beam sensors, light receivers sit symmetrically on either side of the axis of the light projector to create a triple-beam arrangement (Figure 3b). As the target moves along its path, the spot reflected when the beam strikes the edge of the target is different from the light energy reflected when the beam is completely intercepted. Position-detector 1 generates a signal equal to A+[delta]X, and position-detector 2 produces a signal equal to A-[delta]X, where A is the real position and [delta]X is the change in position that occurs as the target enters the beam. Averaging both of these signals results in an accurate range measurement. The system ignores background signals because these signal values are outside the range set by the sensor's potentiometer.

Tight squeezes need fiber optics

For numerous applications, such as semiconductor and micro-parts fabrication, conventional photoelectric components simply don't fit into the available space. Fiber-optic components are ideal for such cramped installations. In addition, plastic and glass fiber optics can function under conditions that would destroy semiconductor sensors. Thus, fiber-optic devices are abundant in applications hampered by high temperature extremes, volatile chemicals, nuclear activity, and excessive humidity. A fiber-optic cable may contain dozens of hair-thin strands of glass or plastic enclosed in a thin, flexible sheath. For photoelectric applications, you can use these fiber-optic sensors for distances of 2 ft or less. However, adding specially designed lenses to the tip of the sensor increases the range to 20 ft.

To scan and detect small, diffuse parts, you can use bifurcated fiber-optic cable. This cable combines fibers from the emitter with those of the receiver bundle at the sensor tip. Although the color, texture, and reflectivity of the target surface affect the amount of reflected light energy, this arrangement suits transparent objects because the sensor tip can almost touch the target. Plastic fiber is less expensive, and you can easily trim it to the necessary size. However, glass fibers are necessary where temperatures extend to 600°F or in strong-chemical environments. Fiber-optic devices are relatively expensive and susceptible to breakage, and the tips are subject to moisture and dirt. However, these devices perform when nothing else does the job.

Selecting the proper sensor

Selecting the proper photoelectric sensor is not quite as simple as selecting a transistor, for example, because of the many variables involved in the final application. Among the key factors to consider are:

• The target you need to detect, count, or monitor, including its size, orientation, color, and texture;
• The placement of the sensor;
• The speed of the target's movement;
• The available power source;
• Environmental restrictions, such as dirt, dust, harsh chemicals, temperature extremes;
• The cost, installation, and maintenance of the system hardware.

The target is obviously the first factor to consider, and you need to answer a number of questions. Is the target a brick or a thin magnet wire? If it is a brick, how will it be facing the sensor (broadside or at an oblique angle?). In other words, how much of a target will the brick, wire, or other material present? Does the brick have a rough texture, or does it have a highly reflective gloss-paint coating? Will the brick be as reflective as the background when viewed from the sensor, or will it be darker, requiring background suppression? Does the application involve merely detecting the presence or absence of the brick or checking whether a mounting hole has been precisely drilled near one edge of the brick?

The color of the target is important. Some colors absorb a substantial amount of transmitted light energy and, thus, require high-sensitivity sensors. Black, for example, is a poor target in retro-reflective or proximity sensors. It may be difficult to see black targets placed against many industrial backgrounds. Green targets are also difficult to detect with commonly used infrared light sources. Therefore, green LED versions of some sensors are available. The texture of a target influences sensor choice much the same as color; coarse-textured surfaces can absorb and attenuate light energy, and highly reflective materials can result in sensor errors.

Because most photoelectric systems use modulated light beams, ambient light is generally not a serious factor. However, you must take proper precautions if you mount the sensor where temperatures may exceed the normal ratings of 0 to 50°C. Sensors are available to survive a range of temperature and humidity extremes, and special fiber-optic assemblies can function to 400°C.

Target speed is key

How fast the target moves translates to how much time the emitter's beam has to view the target. A related factor is sensor-response time, which is the time required to react to the sudden appearance and departure of an object. Response time can be different for light-to-dark and dark-to-light transitions. For example, a sensor may react to a target's appearance in 3 msec but take 4 msec to respond to its removal. Thus, the total sensor response time is 7 msec. You must also consider the interface following the sensor. Solid-state dc interfaces that use transistor output stages are much faster than ac-operated systems containing electromechanical relays. Thus, you should not specify the fastest available sensor without first analyzing the other time-sensitive components.

Using Figure 4, you can calculate the presentation time that the target offers the emitter beam. Assume that the target is larger than the sensor spot (beam diameter), the duty cycle (target vs background) is 50% or less, and the speed of the target is S in./msec (Figure 4a). The presentation time, T, then equals the length, L, of the target in inches minus the spot width, W (assuming that half the target must be in the beam for the sensor to respond), divided by target speed, S, or

Obviously, the presentation time, T, must be equal to or greater than the sum of the response time of the sensor plus the minimum input time of the PLC or other devices involved in the system. If the target is equal to or smaller than the sensing area and the sensor can see the target only at one point in its travel (Figure 4b), the target must stop to ensure a sufficient presentation time. To determine the maximum acceptable speed of a small target that the sensor can detect for some length of its travel, you must determine the length of travel (in inches) that is visible. The maximum presentation time, T, is again equal to the sum of the response of the sensor and the minimum input time of the PLC or other recording device. The maximum acceptable speed, S, is equal to the length of travel, L, divided by the minimum presentation time, T, or

In systems that involve rotary motion, a casual consideration of speed can be misleading. For example, an application may involve scanning a marker or dot on a cylinder rotating at 10 rpm (Figure 4c). This task doesn't seem very difficult because 10 rpm is rather slow. However, if the cylinder has a diameter of 10 in., then the speed of the dot equals the circumference ([pi]×d) times the speed of rotation, which equals 3.14×10×10=314 ipm, or 5.23 ips.

Consider package styles and mounting options

In addition to the speed of the target, you must also determine a number of physical characteristics early in the system-design phase. You must determine how close you can place the sensor to the target and where you can mount the sensor to best strike the target. You need to determine if there is sufficient space to perform any necessary optical alignment and how easily you can remove the sensor for cleaning or replacement.

Photoelectric-sensor packages are available in a variety of forms: panel-mounted with remote optics, frame-mounted with remote optics, and direct-mounted with self-contained optics. Factors affecting the package choice include available space, environmental limitations, and cost. Self-contained, fiber-optic assemblies are relatively expensive but are ideally suited when space is at a premium.

The flexibility of the mounting brackets included with the sensor has a significant effect on installation and the time and cost of maintenance. You need to determine the following: how many holes must be drilled and with what precision to secure the sensor, how rapidly you can complete alignment, and how difficult it is to disassemble the unit when repair or replacement is necessary.

You must take precautions for sensor installations in ovens and furnaces where radiant heat can elevate sensor ambient temperature levels above the maximum rated value. For example, a heat shield with a hole large enough to pass the sensor beam may offer a solution. Models are available that are explosion-proof, resistant to harsh chemicals, and able to survive radiation exposure.

Don't forget wiring considerations

Don't overlook wiring costs because they can add up. Through-beam arrangements may require wiring runs over hundreds of feet. Factors that can compound cost and time include the labor to pull wiring through conduits and the need for junction boxes that involve splices. If a failure occurs in the photoelectric system, a fully loaded production line may be brought to a grinding halt. You need to consider how long it will take to trace the trouble if wiring is the culprit. Think about how fast and easily you can replace a faulty sensor. Can the faulty sensor situation be corrected by merely snapping in a replacement with a convenient connector, or will it be necessary to splice and solder a replacement? Also, consider the available power, and, if it is dc, be sure that the supply has an appropriate filter.

A final consideration is the amount and type of wiring housed in the conduit that carriers the sensor signals. If a large bundle of wires is adjacent to the sensor wires, electromagnetic coupling is inevitable. Should wiring from heavy-duty motors, drill presses, or other high-current switching equipment coexist near sensor wires, extraneous voltage spikes will appear at the sensor receiver and wreak havoc with photoelectric-system performance. Thus, you need to analyze the sensor wiring run and all other wiring that the conduit encloses. A separate conduit for sensor wiring may be justified.

Once you select a sensor, you must set up a prototype system and check it thoroughly. Unknown glitches, such as unexpected reflections from nearby equipment or intermittent high-level pulses from distant motors, can result in unpredictable performance. You need to isolate and correct these bugs before final installation. Don't hesitate to discuss such problems with the vendor; chances are that the vendor's application engineers have seen similar problems before and can immediately suggest a number of remedies. There may be areas of compromise for such parameters as placement of the sensor, location from the target, or speed of target motion.