This lesson builds knowledge. Completing it does not authorize field work, equipment servicing, LOTO, energized testing, or customer-site activity. Site rules, your supervisor, and required LEO Safety Training govern all work. Stop-Work Authority is absolute.
§01 — Learning Objectives
Learning Objectives
By the end of this lesson, the technician will be able to:
Objective 1 (Cognitive/Understanding): Explain the physical operating principle of each major discrete sensor family — inductive, capacitive, photoelectric, and ultrasonic — and identify which target materials and environments each can and cannot detect reliably.
Objective 2 (Diagnostic/Analytical): Differentiate PNP (sourcing) from NPN (sinking) output configurations, predict the wiring consequences of mismatching a sensor output to a PLC input card, and interpret sensor status LEDs during fault isolation.
Objective 3 (Field/Practical): Execute a structured replacement and alignment of a failed discrete sensor, including verification of output switching with a DMM before returning the machine to service.
§02 — Field Scenario
Field Scenario
The Phantom Box: A high-speed case packer keeps faulting with "Product Not Present at Station 3." The operator swears product is flowing. You arrive and watch a cycle: cases stream past Station 3, and the diffuse photoelectric sensor's yellow output LED flickers — solid on some cases, dead on others.
The junior tech reaches for a spare sensor. You stop him and look at what changed. The cases passing through today are a new customer's product — matte black cardboard instead of the usual white. A diffuse sensor detects light bounced off the target, and a matte black surface returns a fraction of the light a white one does. The sensor isn't broken; it's been pushed past its excess gain margin by a darker target.
You verify with the sensitivity adjustment: at maximum gain it detects the black cases — barely, with no margin for dust buildup. The correct fix isn't a twin replacement; it's a sensing-mode change. You install a retro-reflective sensor with a reflector across the conveyor: now the case is detected by blocking the beam, and target color becomes irrelevant. The fault never returns. A part-swapper would have replaced an identical sensor and inherited an identical problem.
§03 — Concept Overview
Concept Overview
Discrete sensors are the control system's yes/no senses. Each one answers a single binary question — part present? guard closed? cylinder retracted? — by switching its output ON or OFF. They feed the PLC's digital input cards (Module 7 covers the controller side) and outnumber analog instruments in most plants by a wide margin.
Selecting or diagnosing a discrete sensor always comes down to three questions: What is the target made of? What is the environment doing? And what does the electrical interface expect?
Inductive Proximity Sensors
An inductive sensor generates a high-frequency electromagnetic field from a coil in its face. A metal target entering the field induces eddy currents, loading the oscillator; the sensor detects that load change and switches. Key field realities:
Metal targets only. Best response to ferrous steel; non-ferrous metals (aluminum, brass, copper) reduce the effective sensing range by a published correction factor (often 0.3–0.5 of rated range for aluminum on standard models).
Short range: typically 1–30 mm depending on barrel diameter. An 8 mm barrel sensor may sense only 1–2 mm.
Extremely robust: immune to dust, oil mist, and most washdown — the default choice for machine position detection.
Shielded (flush-mount) vs. unshielded: shielded models can be mounted flush in metal but have shorter range; unshielded models need clearance from surrounding metal or they self-trigger.
Capacitive Proximity Sensors
Capacitive sensors detect the change in capacitance caused by any material — metal, plastic, wood, liquid, powder — entering the field. Sensitivity is usually adjustable via a potentiometer.
The non-metal specialist: the standard choice for detecting product through a non-metallic wall — e.g., liquid level through a polyethylene tank or pellets through a sight window.
The trade-off: because they respond to anything with mass and dielectric constant, they also respond to condensation, material buildup on the face, and humidity swings. A capacitive sensor in a steamy washdown area is a nuisance-trip generator.
Photoelectric Sensors
Photoelectric sensors detect targets with a beam of modulated light (typically infrared or visible red). The three sensing modes matter more than the brand:
Through-beam: emitter and receiver are separate devices facing each other. The target is detected by breaking the beam. Longest range, highest reliability, immune to target color — at the cost of wiring two devices and aligning them.
Retro-reflective: emitter and receiver share one housing; the beam bounces off a reflector. Detection = beam blocked. Polarized retro-reflective models reject false returns from shiny targets (a critical spec around stretch-wrapped or metallic product).
Diffuse: emitter and receiver share one housing; the beam bounces off the target itself. Easiest to install (no reflector), shortest range, and — as the Field Scenario showed — hostage to target color, finish, and angle. Excess gain (the ratio of received light to the minimum needed to switch) is the spec that predicts reliability as lenses get dirty.
Ultrasonic Sensors
Ultrasonic sensors emit high-frequency sound pulses and time the echo. Because sound doesn't care about color or transparency, they detect clear glass, liquids, and visually difficult targets photoelectrics miss.
Deadband: every ultrasonic sensor is blind for a short zone directly in front of its face (commonly 5–20% of max range) — targets inside the deadband produce unreliable output.
Environmental sensitivities: strong air currents, steep temperature gradients, and sound-absorbing surfaces (foam, fabric) degrade the echo.
§04 — How the Principle Works
How the Principle Works: The Electrical Interface
A discrete sensor that senses perfectly but is wired into the wrong input type does nothing. Three interface concepts cover nearly every field situation:
PNP (Sourcing) vs. NPN (Sinking)
Three-wire DC sensors use a transistor to switch the output:
PNP (sourcing): the sensor switches +24 VDC onto the output wire when actuated. The load (PLC input) sits between the output and 0 V. PNP is the dominant standard in North American and European machine-building.
NPN (sinking): the sensor switches the output wire to 0 V when actuated. The load sits between +24 V and the output. Common in equipment of Asian origin and legacy installations.
A PNP sensor must land on a sinking input card; an NPN sensor on a sourcing input card. Mismatch the pair and the input never sees a valid transition — the classic "brand-new sensor, dead input" service call.
Output Logic: NO vs. NC
Normally Open output turns ON when the target is present; Normally Closed turns OFF. Safety-related detection (guards, light-curtain style presence) conventionally uses NC logic so a cut cable looks like an actuated condition and fails the machine safe rather than blind.
Load Limits
Transistor outputs are small switches: typical continuous ratings are 100–200 mA. They drive PLC inputs and pilot relays comfortably; they do not drive solenoids or motor contactor coils directly. Exceeding the rating kills the output transistor — often intermittently at first, which mimics an alignment problem.
§05 — Reference Matrix
Discrete Sensor Selection Matrix
Sensor Family
Detects
Typical Range
Strong In
Weak Against
First Diagnostic Check
Inductive
Metals only
1–30 mm
Dust, oil, washdown, vibration
Non-metal targets; reduced range on non-ferrous (apply correction factor)
Gap distance vs. rated range; surrounding metal on unshielded models
Capacitive
Any material
3–40 mm (adjustable)
Detection through non-metal walls; powders, liquids
Field Application Checklist: Replacing and Verifying a Discrete Sensor
Confirm the machine is in a safe state for the work. If guards must be opened or you will be inside the motion envelope, apply LOTO per Module 1 (Lesson 1.2) before touching anything. Sensor work near live automation is how hands meet unexpected motion.
Record the failed sensor's identity before removal: part number, wiring configuration (2/3/4-wire), output type (PNP/NPN), logic (NO/NC), and the exact mounting position and gap. Photograph the alignment.
Verify the replacement matches all electrical specs — not just the thread size. A physically identical sensor with NPN output in a PNP system is a guaranteed callback.
Land the wiring per the sensor's diagram (brown = +V, blue = 0 V, black = output is the IEC convention for 3-wire DC sensors — but verify against the datasheet, never assume).
Set the mechanical gap/alignment to the recorded position. For inductive sensors, target the manufacturer's rated sensing distance at roughly 75% or less for margin. For photoelectric, center the beam and verify the alignment indicator LED.
Verify switching with the status LED first: introduce and remove the target; confirm the output LED follows cleanly with no flicker at the actual process speed.
Verify electrically at the sensor output with a DMM (Lesson 3.6 discipline): for PNP, output-to-0V should read ~24 V actuated and ~0 V released. Confirm the corresponding PLC input indicator follows.
Run the machine in a supervised test cycle and watch several product cycles before closing the work order. Document the sensor model, settings, and gap in the CMMS.
§07 — Knowledge Check
Knowledge Check
A technician replaces a failed 3-wire inductive proximity sensor on an infeed conveyor with a spare of the same brand, barrel size, and sensing range. After wiring brown to +24 V, blue to 0 V, and black to the PLC input, the sensor's own output LED switches perfectly when metal passes — but the PLC input LED never turns on. A DMM from the black wire to 0 V reads ~24 V steady whether or not a target is present. What is the most probable cause?
Detailed Explanation
The sensor's own LED switching proves the sensing function and power wiring are healthy — eliminating answer C and making B untenable (reversed supply on most 3-wire DC sensors yields a dead or damaged sensor, not inverted logic; NO/NC is a model characteristic, not a wiring choice). Answer D ignores the measured evidence. The 24 V constant reading on the output conductor is the tell: in a sinking-input architecture wired for PNP sensors, the input terminal expects to be driven up to +24 V by the sensor. An NPN unit instead tries to pull the line down to 0 V — and depending on the input circuit, may not be able to produce a transition the card recognizes at all. Output polarity mismatch after a "like-for-like" swap is one of the most common discrete sensor service errors, because PNP and NPN versions of the same sensor often differ by a single character in the part number.
§08 — Source List
Source List
IEC 60947-5-2 — Low-voltage switchgear and controlgear — Control circuit devices and switching elements: Proximity switches (governing standard for proximity sensor characteristics and DC wiring color convention)