Photoelectric sensors are an important part of any industrial automation process. Along with the inductive proximity switch and capacitive sensor, they allow the reliable detection of objects without involving physical contact. As just one of the myriad of applications to which these could be put, I will focus on the detection of a manufactured part as it travels down a conveyor belt, a common and simple task.
Photoelectric sensors, or photoeyes, began as crude and primitive devices compared to today's available technology. They were simply a flashlight and a photoresistive sensor calibrated to ignore ambient light and detect the brighter glare of the flashlight. When a part passed in front of the detector, blocking the flashlight beam, it would register the presence of the part and send a signal to a bank of relays or a PLC to continue the automation process. These devices could be fooled by other bright lights in the area and fail to send their signals.
Still, they were better for some applications than the mechanical switches they were intended to replace. Mechanical switches, by necessity, made physical contact with the part they were to detect. Their moving parts were subject to wear and had a limited cycle life. Additionally, they could in some cases damage delicate parts, or fail to detect parts that were too light or flexible to trigger their switching lever. Photoeyes solved these problems by not touching the part at all. However, mechanical switches still find application in situations where their mechanical action can directly break electrical connections, such as for safety devices, and also when light and flexible material (such as stray or loose packaging material) is intended to be ignored.
Modern photoeyes have improved the original concept remarkably. Light emitting diodes and lasers are capable of emitting a much tighter, focused, and frequency-specific beam than a simple light bulb, resulting in a sensor that is much more difficult to fool with other bright lights in the area. Infrared LEDs can even offer detection with light invisible to the human eye (infrared LEDs are also used in television remote controls). Alignment is sometimes easier with a visible beam though. Their maximum range has been increased considerably, and models with intentionally short ranges are also available (sometimes you will want to detect an object which is close and ignore one far away). A number of other models have been developed for specific applications as well, improving on the simple transmitter-receiver version represented by the obsolete model described above. All models can be fooled by very clear objects, and they must be kept clean. A dirty sensor will not pass light properly and may trigger unreliably or constantly.
As described above, the transmitter-receiver model has two parts, a simple LED or laser light source, and a sensor to detect its beam. Both parts require a power source, although they can come from different supplies. The detector piece is the only one which needs to be connected to the rest of the automation equipment. When a part passes between the transmitter and receiver, the beam is broken and the sensor sends a signal to the automation controller, usually a PLC these days.
Disadvantages: Consists of two units, both of which require a power supply. A clear spot must be available on both sides of the conveyor to put the two units. The transmitter must be aligned directly at the receiver and will fail if either part is bumped or moved out of position. Two sensors close together may experience "crosstalk" in which one sensor detects the signal from the wrong transmitter. This can be solved by orienting the receiver-transmitter pairs in opposite directions across the conveyor from each other so the transmitters are pointing at each other, rather than each others' receivers.
The fiber optic model is very similar to the transmitter-receiver model, except that it uses fiber optic technology so that power only has to be supplied to one side. In our conveyor example, the fiber optic cable is then pulled to the other side of the conveyor and pointed back at the sensor unit. When a part passes between them, the beam is broken and the sensor is triggered.
Disadvantages: Much like the transmitter-receiver, except that fiber optic technology means the other side does not require electrical power. However, there must be a good spot to run the fiber optic cable across without getting in the way.
The retroreflective model is a departure from the previous models in that the transmitter and receiver are built in to the same unit. The beam is bounced off a reflector on the other side of the conveyor back at the receiver on the same unit which transmitted the beam. When a part passes between it and the reflector, the beam is broken and a sensor is triggered.
Disadvantages: Much like the transmitter-receiver, except that the reflector is less forgiving of misalignment (although the retroreflective nature of the reflector makes the alignment of the reflector easier than if it were a simple mirror). However, the reflector has no need of a power source or any physical connection to the sensor unit, making it useful for applications in which it is not feasible to run wire or fiber optics to the other side for whatever reason. The retroreflective model does have the additional disadvantage of being fooled by highly reflective objects, which may bounce the beam back instead of triggering the sensor.
The diffuse unit has a significant advantage in that it does not require a second receiver or reflector to aid in detection. Instead it relies on light bounced back off of the part it is intended to detect. A broad-beam diffuse photoeye has a wide sensing range for reliable detection of parts that may not be lined up the same way every time, and a narrow-beam diffuse photoeye can ignore things around the part or detect misalignment in precision applications.
Disadvantages: Although there is no need for a second receiver unit that requires alignment or power, the diffuse unit can sometimes be fooled by dark objects which do not reflect enough light to detect. Additionally, the diffuse photoeye is less fail-safe than the transmitter-receiver units. There is no difference between a situation in which no part is present and a situation in which the photoeye is not functioning properly — in either case it is simply not receiving light. The other units should be constantly receiving light when no part is present, making it obvious when they have failed.
The light curtain is a device consisting of two long rails with several transmitter-receiver units along their length, typically packed closely together (less than an inch) so nothing can fit between the beams without triggering at least one. Light curtains provide detection in a large rectangular area and are usually used to detect something passing through an opening or as a sort of invisible fence. They are very touchy and the two rails must be lined up with each other with respect to pitch, roll, and yaw in order to function properly. To enclose an area in an invisible fence, they can be placed back-to-back with mirrors at the corners of the enclosed area. To cover an opening, they are simply placed at the sides of the opening, facing each other. Some models can be programmed to ignore objects passing through that only break a few of the beams, but trigger when larger objects pass through. This is useful for when small manufactured objects must leave the enclosed area but it would be dangerous for a person to enter.
Disadvantages: As the transmitter-receiver model, with additional sensitivity to misalignment. They are also very expensive, especially the models with built-in redundancy which are intended for safety applications.
A laser caliper is the logical extension of the diffuse model, using a laser instead of an LED. Not only does it detect the light reflected off of an object, but also calculates how far away the detected object is. Given the angle at which the light is transmitted and the distance across the sensor at which it is reflected back, it is a trivial problem to calculate the distance to the object. Objects farther away will reflect the light proportionally farther across the sensor by the law of similar triangles. Laser calipers can be very precise and accurate and are often used to measure the thickness of manufactured objects, such as paper, which have tight caliper tolerances.
Disadvantages: Care must be taken to ensure that the zero point remains at the same distance from the sensor at all times. This means preventing material buildup or wear on the zero point which could artificially bring the object closer to or farther from the sensor, and also ensuring that the sensor itself is mounted rigidly to prevent movement. Highly precise applications sometimes get around this by using two laser calipers mounted at a known distance from each other on opposite sides of the measured object. The difference between the sensor readings is the thickness of the object.
There are also dozens of highly specialized sensors for detecting color, very small objects, quality of reflection, glare, and even ones manufactured specifically for detecting clear objects. More powerful laser models are used to burn through smoke, haze, and dust which would confound the basic models, or provide a vastly improved range. These devices have come a long way from the simple flashlight and photoresistive sensor setup that started it all. They have become an invaluable addition to factory automation and are quite literally the eyes of a modern manufacturing process.
By the way, you've probably got a photoelectric sensor if you've got a modern automatic garage door opener. It's a transmitter-receiver setup at the bottom of the door, set up to stop the door from closing if something is blocking the beam, and therefore in the way of the door.