13.1 Discrete Field Devices: Proximity, Photoelectric & Mechanical Switches

13.1 Discrete Field Devices: Proximity, Photoelectric & Mechanical Switches

Quick Answer: Everything a PLC knows about the physical world arrives through a field device. Red Seal sub-tasks F-29.01 and F-29.02 cover installing and maintaining the discrete (on/off) end of that: limit switches, inductive and capacitive proximity sensors, photoelectric sensors, ultrasonic and magnetic sensors. Three facts carry most of the exam weight: an inductive sensor's rated range assumes a standard mild steel target and derates for every other metal; shielded sensors mount flush in metal but reach less far than unshielded sensors, which need a metal-free zone; and a PNP (sourcing) sensor pairs with a sinking input card, an NPN (sinking) sensor with a sourcing card.

---

1. Mechanical Limit Switches

The oldest discrete device is still the most reliable in the harshest conditions — hot, dirty, oily, high-vibration places where optics fog and electronics cook.

Actuator selection

The travel terminology that gets tested

The installation rule: design the cam or dog to carry the actuator through the operating point and into the overtravel range, but never to the switch's mechanical limit. Provide a positive mechanical stop for the machine independent of the switch. A limit switch destroyed by the machine it was watching is the single most common mechanical sensing failure in a plant.

Safety interlocks

Any switch proving a guard position must use direct-opening-action (positive-opening) contacts, marked with the arrow-in-circle symbol, so that opening the guard mechanically drives the normally closed contact apart even if the contacts have welded.

---

2. Inductive Proximity Sensors

An inductive sensor drives an LC oscillator whose coil produces a high-frequency field at the sensing face. A conductive target entering that field has eddy currents induced in it; those currents absorb energy, the oscillation amplitude falls, and a trigger circuit switches the output.

Therefore: an inductive sensor detects metal, and only metal. It is completely blind to plastic, wood, glass, liquid and cardboard, which is also its greatest advantage — a metal part on a plastic conveyor is trivially easy to detect.

Target material correction factors

Rated sensing distance ($S_n$) is specified with a standard target: a square of mild steel (Fe 360), 1 mm thick, with a side length equal to the sensor's sensing face diameter or three times $S_n$, whichever is greater. Every other material derates it:

Worked example. A design calls for detecting an aluminium cam at a 5 mm gap. A sensor with $S_n = 10\text{ mm}$ appears to have twice the range needed, but:

$$S_{\text{effective}} = 10\text{ mm} \times 0.35 = 3.5\text{ mm}$$

The aluminium target will not be detected at 5 mm. Worse, it may be detected intermittently on a warm day and not on a cold one, because sensing distance also drifts with temperature. The design needs a sensor with a larger $S_n$, a closer mounting gap, a steel target flag attached to the aluminium part, or a different sensing technology.

Two further derations to remember:

• Assured operating distance ($S_a$) is the distance at which detection is guaranteed over the full temperature and voltage range, and it is conventionally about 81% of $S_n$. Never design to $S_n$.
• Target size below the standard target reduces range as well.

Shielded versus unshielded

The most common field failure: mounting an unshielded sensor flush in a steel bracket. The surrounding steel is inside the sensing field, so the sensor sees the bracket, latches on, and never changes state. The symptom is a sensor whose LED is on permanently regardless of the target.

Mutual interference is the second: two inductive sensors mounted close together, face to face or side by side, couple into each other's oscillators and produce erratic switching. Manufacturers publish minimum spacing for both arrangements.

---

3. Capacitive Proximity Sensors

A capacitive sensor's face forms one plate of a capacitor; an approaching target changes the capacitance and starts an oscillator. Because the change depends on the target's dielectric constant rather than its conductivity, a capacitive sensor detects almost anything: metal, water, grain, plastic pellets, wood, powder, cardboard.

• Sensing through a vessel wall. The signature industrial application is point level detection — a capacitive sensor mounted on the outside of a plastic or glass tank or a sight glass detects the liquid inside, with no process penetration at all.
• Sensitivity adjustment. Most industrial capacitive sensors have a potentiometer or teach input so the threshold can be set to see the product but ignore the container.
• The weakness is everything else. Condensation, washdown water film, dust and product build-up on the face all look like a target. Capacitive sensors in dirty or humid areas need either regular cleaning on a PM route or a different technology.

---

4. Photoelectric Sensors

Photoelectric sensors detect the interruption or reflection of a light beam — usually infrared LED, sometimes visible red for easy alignment, sometimes laser for tiny targets.

Excess gain and contrast

• Excess gain is the ratio of light received to the minimum needed to operate. An industrial installation targets an excess gain of at least 1.5 in clean air, 5 in light dust, 10 or more in heavy dust, steam or spray, because contamination on the lenses consumes gain over time.
• Contrast is the ratio of received light with the target absent to received light with the target present. A through-beam sensor has enormous contrast; a diffuse sensor looking at a matte black part in front of a white machine guard may have almost none — and it is exactly this case that background suppression solves.

Design rule: choose through-beam whenever the physical arrangement allows it. Nearly every chronic photoelectric problem in a plant is a diffuse sensor doing a job that needed through-beam.

---

5. Other Discrete Sensing

• Ultrasonic sensors emit a sound burst and time the echo. They see nearly any material regardless of colour or transparency — including clear plastic and dark liquids that defeat optics — but they are affected by temperature (the speed of sound changes), by strong air currents, and by soft or angled surfaces that absorb or deflect the echo.
• Magnetic (reed) switches respond to a permanent magnet and are the standard method for sensing piston position in a pneumatic cylinder, mounted in the cylinder's extrusion groove.
• Hall-effect sensors are solid-state magnetic sensors with no contacts, used for speed and position pickups.

---

6. Wiring Discrete Field Devices

The IEC 60947-5-2 colour code

Two-wire, three-wire and four-wire

The two-wire sensor trap

A two-wire sensor must draw a small current through the load even when it is off, to power its own electronics — typically 0.5 to 2 mA of leakage. A modern PLC input card has very high impedance, so that leakage can develop enough voltage across the input to hold it on permanently. The symptom is a sensor whose LED correctly turns off while the PLC input stays lit.

The fix is a bleeder resistor across the PLC input sized to shunt the leakage without loading the real signal. Two-wire sensors also have a minimum holding current — they need a certain load current to stay operating — so an input card drawing less than that produces erratic behaviour. Three-wire sensors avoid both problems entirely and are the default choice for PLC interfacing.

PNP versus NPN

This is treated in detail in the PLC I/O section of this guide; the pairing rule is worth repeating here because it is a field-wiring decision:

• PNP (sourcing) sensor — when active, connects +24 V to the black output wire. It must feed a sinking input card whose common is tied to 0 V.
• NPN (sinking) sensor — when active, connects the black output wire to 0 V. It must feed a sourcing input card whose common is tied to +24 V.

PNP is the dominant convention in Canadian and European industrial equipment; NPN remains common on Japanese-built machinery. Mixing them on one machine is a documented source of intermittent faults, and the two types are visually identical.

---

7. Troubleshooting Discrete Field Devices

+---------------------------------------------------------------------------------+
|              DISCRETE SENSOR TROUBLESHOOTING — WORK OUTWARD                     |
|                                                                                 |
|  1. Does the SENSOR'S OWN LED change state when the target is presented?        |
|        NO  -> Sensor, gap, target material, alignment, contamination,           |
|               or supply voltage. Verify 24 V at brown-to-blue.                  |
|        YES -> The sensor is working. Continue downstream.                       |
|                                                                                 |
|  2. Does the PLC INPUT LED follow the sensor LED?                               |
|        NO  -> Wiring, wrong sourcing/sinking pairing, broken conductor,         |
|               two-wire leakage holding the input on, blown input fuse.          |
|        YES -> The I/O path is good. Continue into the program.                  |
|                                                                                 |
|  3. Does the PROGRAM data table bit follow the input LED?                       |
|        NO  -> Forced bit, wrong address, module fault, corrupt I/O map.         |
|        YES -> The fault is in the logic or downstream of it.                    |
+---------------------------------------------------------------------------------+

Symptoms and causes

Maintenance

• Clean optical lenses and reflectors on a scheduled route, and consider an air purge in dusty or wet areas.
• Verify mounting torque and gap — vibration walks sensors out of position, and a 1 mm change is significant at short range.
• Protect cables. Sensor cables broken by a moving machine member are among the most common plant faults; use drag chain, strain relief and abrasion-resistant jacketing.
• Keep spares to the same part number, because a replacement with a different sensing distance, output type or shielding is a fault waiting for the next shift.