Picking the right discrete sensor for a machine application
When designing machinery, one of the first tasks for an engineer is to select discrete sensors for the application. Generally, discrete sensors can be divided into two categories: those that detect the positions of mechanisms and tooling (parts of the machine) and those that detect the parts being processed.
Types of discrete sensors
It’s easy for designers to select hall-effect sensors for pneumatic cylinders since these often are presented as accessories. Hall-effect sensors are designed to detect the magnetic piston inside of a cylinder body and determine whether it is extended or retracted. This works fine as long as the cylinder shaft can’t become unattached from the tooling; detection of tooling using an inductive proximity switch is the only foolproof way to determine that the tooling is in the correct position. Photoelectric sensors are useful for applications when the tooling is not made of metal.
Photoelectrics and inductive proximity switches often are used for part detection. Photoeyes are useful for designers since they can be located farther from the target. This opens up several sensing possibilities such as through-beam (separate emitter and receiver); retroreflective (using a corner-cube bicycle type reflector or reflective tape); or diffuse (using light reflected off of the target). Form factor, such as fiber optics that can reach into tight spaces or larger self-contained sensors that may have a longer range, also should be considered.
Visible beam photoelectrics are often easier to align and infrared sensors also have a longer range. Laser sensors, while more expensive, can be used to detect smaller parts and also extend the sensing range. The other important consideration that needs to be made is the voltage and output type of the sensor. With the exception of some large conveyor systems and process control facilities, discrete sensors are almost always 24 V dc. Positive-negative-positive (PNP) sourcing sensors are used most commonly in the U.S., but equipment from other countries may use negative-positive-negative (NPN) sensors, or sinking types.
Sensors with contact closure type outputs such as limit switches also are used occasionally, but like ac sensors, they are becoming less common. Contact closures are best for applications where the sensor does not switch often. Physical contacts wear and often have a specified lifetime number of operations.
Sensor specification depends on a lot of factors, such as:
- How much space is available for mounting
- What type of target is being detected
- With what type of control system it needs to be interfaced
- Sensor cost and availability of replacements.
Sensors for harsh conditions
In some cases, resistance to chemicals and water are also a consideration. Coolant in metal cutting applications and caustic washdown liquids in food processing are some common substances that may need to be evaluated. This also may contribute to the type of electrical connection the sensor may require; quick disconnect or hard conduit.
There are many resources online to check when selecting discrete sensors for a machine application. Keeping these considerations in mind when making a decision will make the process much simpler.
Frank Lamb, Automation Consulting LLC, is a Control Engineering Editorial Advisory Board member. Automation Consulting LLC is a CFE Media content partner. Edited by Chris Vavra, production editor, Control Engineering, CFE Media, firstname.lastname@example.org.
www.controleng.com keywords: discrete sensors
- Discrete sensors can be divided into two categories: those that detect the positions of mechanisms and tooling and those that detect the parts being processed.
- Photoelectric and hall-effect sensors commonly are used for discrete applications.
- Depending on the application, resistance to chemicals and water also should be considered when picking a discrete sensor.
What other considerations should be made when selecting a discrete sensor?
See related sensor articles from Frank Lamb linked below.