Photoelectric Sensor Not Detecting an Object? Find the Fault Systematically

What should you check first when the sensor stops detecting?

Compare the sensor’s model-defined operation indicator with the PLC input while the actual object enters and leaves the sensing point. This separates a missing optical decision from a signal that is being lost downstream. Do not replace the sensor or reteach it before recording the original behavior.

First make the machine safe under its isolation and testing procedures. Remove power before changing connections. Energized measurements and moving-machine trials require qualified personnel and a controlled test state. Do not bypass an interlock or use an ordinary process photoeye for personnel protection.

Read the indicator name, not just its color. A power LED, incident-light LED, stability LED and output LED report different things. A stability indicator can remain on in both a stable light state and a stable dark state. Record what each named indicator does; do not assume that green means “object detected.”

Is the sensor receiving the right power and configuration?

Check the full model code and its supply diagram, then verify voltage at the sensor connection—not only at the power supply. A loose contact or cable fault can leave a healthy supply reading at the cabinet while the sensor loses power.

Inspect connector seating, damaged cable sections and terminals with power isolated. During an authorized energized test, compare the supply with the model’s operating range, including the event that produces the miss. A steady meter reading may not capture a brief voltage dip; intermittent faults may need a suitable recording instrument.

Do not assume every sensor is a 24 V, three-wire device. Two-wire, AC, relay-output and separate-amplifier designs need their own circuits. Familiar brown, blue and black conductors are not a substitute for pin assignments. On a through-beam pair, verify the emitter’s supply as well as the receiver’s.

Also check whether the unit is in RUN rather than teach mode, whether an external input disables emission, and whether a configured timer delays the output. Those functions exist only on some models. If the fault began after replacement or a settings change, compare the complete suffix and saved parameters before changing the optical setup.

Why is a powered sensor still not reacting to the object?

The receiver must see a meaningful difference between the required object and the clear state. What creates that difference depends on the sensing method. Clean and inspect the optical surfaces, restore the intended path, then adjust using the real target.

Through-beam: does the object interrupt enough light?

Confirm that the emitter faces its receiver and that the object crosses the effective beam. A small part, hole or transparent section may leave enough light reaching the receiver to prevent switching. Test the actual sensing height and lateral path; waving a hand through the beam can conceal a coverage problem.

A shiny rail, floor or wall can also reflect light around an opaque object. In a controlled optical test, shielding the suspected reflection path without moving the sensor or target helps distinguish bypass light from simple misalignment. Use a permanent geometry or accessory permitted for the model if the result supports that cause.

Retro-reflective: is the return coming from the reflector?

Use the specified reflector and inspect both outgoing and return paths. If a shiny product sends enough light directly back to the sensor, the receiver may mistake it for the reflector. This is often called proxing. Polarized retro-reflective sensing can reduce this problem, but the selected sensor, reflector and actual product finish still need testing.

Diffuse or background suppression: can target and background be separated?

Diffuse sensing relies on light returned by the object. A dark target may return too little light, while a bright background may return enough even when the target is absent. With background suppression, check the cutoff setting and the allowed near/far target zone. Moving an object closer is not always a cure: some optical designs have a near-field dead zone.

Alignment should leave a stable operating margin after the bracket is tightened, not merely produce one LED transition. Use the manufacturer’s stability indication, signal display or teach result when available. For cleaning, follow the lens-material instructions; OMRON specifically warns that alcohol and other solvents can cloud or crack its plastic optics.

Why are black, shiny, clear or small objects missed?

Different products can create different optical signals at exactly the same distance. Compare a detected sample with a missed sample at the same position and speed before changing several settings at once.

• Black or matte: in diffuse sensing, compare the weakest return at the farthest expected position. A white reference target’s range is not the guaranteed range for a black product.
• Shiny or curved: rotate real samples through their permitted orientations. A useful reflection may disappear—or an unwanted return may appear—as the angle changes.
• Transparent: check empty, filled, labeled or wet conditions that actually occur. If the light change is too small for stable separation, evaluate a clear-object sensing method rather than repeatedly increasing sensitivity.
• Small or perforated: check beam coverage along the complete travel path. A smaller spot, slot arrangement or fiber-optic head may help, subject to the target and mounting constraints.

What if the sensor LED changes but the PLC input does not?

Keep the optical setup unchanged and trace the signal downstream. A changing output indicator makes the electrical path worth checking next, but it does not prove the output transistor is delivering a valid signal under load.

Confirm the output circuit and intended state

A three-wire PNP output supplies current to a compatible sinking input; an NPN output sinks current from a compatible sourcing input. Confirm the PLC input-group circuit and common connection from its manual. Do not apply this wiring model to a two-wire or relay-output sensor.

Light-ON/Dark-ON controls the relationship between received light and output state; it does not change NPN into PNP. Light-ON usually activates for an object returning light in diffuse sensing, while Dark-ON usually activates when an opaque object blocks a through-beam or retro-reflective path. Record the two states your program actually expects.

Find the point where the signal disappears

1. Sensor output: a qualified person checks the loaded output against its specified ON/OFF behavior. Consider leakage, ON-state voltage drop, load limits and protection status.
2. PLC terminal: if the local output changes but the terminal signal does not, inspect the conductor, connector, terminal and reference path.
3. Input data: if a valid terminal signal reaches the module but the tag does not change, inspect input filtering, channel mapping, module diagnostics and I/O communication.
4. Program response: if the tag changes but the machine does not respond, inspect the sequence condition, interlocks and event-handling logic. Do not bypass those conditions to make the machine run.

A floating, unloaded transistor output can give a misleading meter reading. Use the manufacturer’s specified circuit and test method. Even a successful output test proves only that part of the chain; it does not establish reliable product detection.

Why does a hand test work while production misses objects?

A hand test often presents a larger target for longer. At production speed, the actual object may cross the edge of the beam or create a pulse too short for the sensor or controller to register. Check both the target interval and the clear gap between objects.

First determine whether the sensor produces a repeatable output pulse at speed. Human observation of a small LED and a normal meter display cannot verify a millisecond waveform. Use suitable diagnostic capture or a properly rated recording instrument under the authorized test procedure.

Passing the filter is not the whole timing check. Sensor operating/reset response, hardware delays, remote-I/O updates and the program’s sampling or event handling can still matter. Slowing the conveyor may help isolate timing from optics during an authorized trial, but does not validate the final production speed.

How can you isolate an intermittent fault?

Reproduce one suspected influence at a time while keeping the target position and settings fixed. Record the sensor indication and PLC response together. A fault occurring when a motor starts is a clue—not proof of electrical noise, because the same event may also move the bracket or disturb the supply.

• Lighting: shade a suspected external-light path without blocking the intended beam. A changed result supports investigating orientation, shielding and the model’s ambient-light limits.
• Adjacent sensors: under the controlled test procedure, isolate one neighboring emitter at a time. If behavior changes, investigate optical cross-talk, reflections and supported anti-interference functions.
• Machine movement: observe the bracket, target path, cable strain relief and connectors. Recheck after final tightening; a stable stationary setup can shift during operation.
• Switching loads: compare fault timing with supply and output recordings. Cable routing, grounding and suppression changes must address the demonstrated entry path and follow the equipment instructions.
• Dirt or moisture: compare signal or stability before and after approved cleaning. Recovery suggests lost optical margin, but the application still needs an appropriate maintenance interval or environmental change.

Do not fit arbitrary capacitors, defeat grounding or select a larger filter simply because interference is suspected. A remedy that hides an unwanted pulse can also hide a real product pulse.

When should you replace the sensor—and what should the retest prove?

Replacement is justified when damage or a specified functional test points to the unit, or when a controlled comparison with a compatible known-good sensor isolates the fault. If no configuration can separate your real target from its background, the problem may instead be an unsuitable sensing method.

For substitution, match sensing mode, supply, output, logic, range, response, connector and environmental requirements. Preserve the mounting geometry and settings. Disconnect with power removed, then repeat the same target and clear-state tests. If changing the sensor also moves the bracket or reseats a bad connector, the apparent recovery does not isolate sensor failure.

Before returning to normal operation, verify the real product family, minimum target/gap intervals, maximum approved speed and relevant lighting, vibration and contamination conditions. Confirm both the sensor output and the controller’s required response—not only a successful hand test. Use the machine’s commissioning and return-to-service procedure.