Industrial sensor troubleshooting
Sensor LED Is ON but No Output: Find Where the Signal Stops
A lit sensor LED proves only the condition assigned to that indicator. It does not prove that a usable loaded signal reached the PLC, that the input module accepted it, or that the program used the right bit. Trace the same event through the sensor output, cable, PLC terminal, raw input and control logic before replacing hardware.
Start at the first point where the state stops changing.
If the LED changes but the loaded output at the sensor connector does not, check the indicator definition, NO/NC or Light-ON/Dark-ON mode, measurement reference, load path, overload protection and output driver. If the output changes at the connector but not at the PLC terminal, inspect the cable, connector and terminal.
If the correct signal reaches the PLC terminal but the channel or raw input does not change, focus on input common polarity, PNP/NPN compatibility, input voltage/current thresholds, filtering, configuration and channel health. If the raw input changes but the machine does not, the fault has moved into addressing, forces, mapping, interlocks or downstream control logic.
Signal path first
“ON” can describe five different points in one control chain.
Observe one target event at every checkpoint. The first point that does not follow the previous point narrows the fault to a small part of the system.
Sensor indicator
Shows power, sensing, output command, stability, teach or fault according to the exact model.
Loaded output
The output conductor changes electrically while connected to a compatible input or approved test load.
PLC terminal
The same state arrives at the correct input terminal with the required common and return path.
Raw input state
The input module recognizes the voltage/current after threshold, delay and channel configuration.
Program and action
The correct address or tag reaches the logic, interlocks pass, and the commanded machine action occurs.
Read the legend
The LED does not have one universal meaning.
Some datasheets call the light an operation indicator and show it following the control output. Other sensors have separate power, stability, teach, fault, alignment or signal-strength LEDs. The same color can mean different things across product families.
For example, OMRON’s E2E documentation provides model-specific timing charts that place the operation indicator beside the control output and sensing-object state. That is useful evidence for those exact variants, not permission to assume that every LED on every sensor proves the black-wire output.[1]
Find the exact model suffix and read its indicator table before diagnosing the circuit. A PNP NO model, PNP NC model, NPN model, 2-wire model and configurable photoelectric sensor may look nearly identical while producing different electrical states.
Power LED
Confirms some internal power condition. It may stay on regardless of target or output state.
Operation/output LED
Usually follows the commanded switching state shown in the model timing chart, but not downstream delivery.
Stability or margin LED
Reports sensing margin or a stable region. It may not represent the output conductor at all.
Teach or fault LED
May flash or use a code for teach status, overload, short circuit, configuration or diagnostic conditions.
Interactive fault locator
Select the first checkpoint that fails to follow the target.
Use this as a screening path, not as permission to work on energized equipment or bypass machine safeguards.
What is the first failed observation?
Output circuit, mode, load or measurement setup
The sensor is detecting, but the output conductor is not producing the expected loaded state. Do not condemn the sensor until the output diagram, operating mode, reference point and load path are confirmed.
- Verify the LED meaning and NO/NC or Light-ON/Dark-ON setting.
- Confirm PNP, NPN, push-pull or 2-wire output and the correct measurement reference.
- Check overload/short-circuit behavior and output-load limits in the manual.
- Test with the real compatible input or an approved load, not an arbitrary resistor.
Controlled test sequence
Trace the signal in order and change one condition at a time.
The safest useful measurement is one that answers a defined question. Record target state, LED state, test points, reference point, load and PLC observation together.
Control hazardous energy first. Wiring changes, continuity tests and channel moves should be made under the site’s approved de-energized procedure. Where an authorized energized diagnostic measurement is necessary, use trained personnel, suitable instruments, protected test points and the machine’s controlled commissioning procedure. A software stop or sensor state is not an energy-isolating device.[6]
Identify the exact components
Record the complete sensor suffix, cable or connector version, PLC input module, terminal group and supply. Similar housings can hide different circuits.
Decode the indicators
Use the exact manual to label each LED as power, operation, stability, teach or fault. Record target-present and target-absent behavior.
Verify supply at the sensor
Measure at the sensor or connector under the real operating condition. Compare with the model’s allowed supply range and polarity.
Read the output diagram
Confirm PNP/NPN, NO/NC, Light-ON/Dark-ON, push-pull or 2-wire behavior and the input common required to complete current flow.
Measure the loaded output
Compare output state at the sensor connector in both target conditions while connected to a compatible input or approved test load.
Repeat at the PLC terminal
If the connector changes but the terminal does not, de-energize and inspect the conductor, M8/M12 pinout, junctions, terminal and shared common.
Check the physical channel
Compare terminal state, channel LED, module diagnostics and raw input data. Verify common polarity, ON/OFF thresholds, filter and enable status.
Follow raw data into logic
Compare the raw input with the program tag, alias, mapping, force state, debounce logic, sequence conditions and interlocks.
Use a controlled spare channel
Only after de-energizing, move to a known-good channel with equivalent electrical/configuration conditions. Update the test address deliberately.
Substitute one known-good item
Use the same approved model and unchanged wiring. Replace the sensor only when the fault follows it and a known-good unit restores the same test.
Interpret the circuit
PNP and NPN outputs need different current paths and different reading logic.
Do not use “about 24 V when triggered” as a universal sensor test. That expectation fits many loaded PNP NO circuits measured from output to 0 V, but it does not describe an active NPN output, an NC model, a two-wire sensor or every push-pull device.
PNP transistor output
A PNP output sources current toward the load. A common three-wire NO arrangement uses brown for +V, blue for 0 V and black for output, but the exact diagram and connector view remain authoritative.
Active NO reading: output-to-0 V should enter the PLC’s ON region. It is normally below the supply by the sensor’s residual voltage and any cable/connection drop.
Inactive NO reading: the PLC input/load commonly pulls the output toward 0 V. An open-circuit meter may show a less useful floating value.
NPN transistor output
An NPN output sinks current toward 0 V. The PLC input or load must provide a path from +V; otherwise the output conductor can be measured in a circuit that has no usable current loop.
Active NO reading: output-to-0 V should fall into the low region while current flows through the input. A low reading is expected, not evidence of “no signal.”
Inactive NO reading: the compatible input/load commonly pulls the conductor toward +V. Without that load path, the conductor may float.
| Output arrangement | Target state for NO model | Useful output-to-0 V interpretation | Do not forget |
|---|---|---|---|
| 3-wire PNP NO | Target present | High enough to satisfy the exact PLC ON threshold under load | PLC/input group normally needs a 0 V common return path |
| 3-wire NPN NO | Target present | Low enough to represent the active sinking state under load | PLC/input group must supply current from +V through the input |
| PNP or NPN NC | Target present | Opposite switching state from the corresponding NO model | LED and timing chart must be read for the exact mode |
| 2-wire DC sensor | Model-dependent | Evaluate voltage across the sensor and load current as a series circuit | Minimum load current, leakage current and residual voltage can all matter |
| Push-pull or configurable output | Configured mode | Use the manufacturer’s truth table and loading conditions | Do not infer behavior from housing, wire color or a previous model |
Why an unloaded multimeter test can mislead
A digital multimeter usually draws very little current. An open-collector output, leakage path or input circuit can therefore show a voltage that disappears when connected to the real PLC, or show an undefined floating level when no pull-up or pull-down path exists.
Measure the actual installed input whenever the procedure permits. If a separate test load is required, use the sensor manufacturer’s method or an engineering-approved value derived from the output voltage, minimum and maximum load current, resistor power and fault condition. A fixed 2.2 kΩ resistor is not a universal test tool.
For more circuit background, see the XSZ guides to NPN vs PNP outputs, normally open vs normally closed logic and two-wire vs three-wire sensors.
Checkpoint 2
If the indicator responds but the loaded output does not, prove the circuit before condemning the sensor.
The fault may be a configuration or test-method problem, a protected overload state, a broken output conductor, or a damaged switching stage.
Wrong mode or wrong LED assumption
An NC output may open when the target appears. A photoelectric sensor may be set to Light-ON or Dark-ON. A stability LED may change while the output condition follows a different indicator.
First proof: compare the exact timing chart with target-present and target-absent states.
Overload or short-circuit response
Some sensors include load short-circuit protection, but the response and recovery are model-specific. The output may limit, shut down, flash a fault code or require the fault to be removed before recovery.
First proof: de-energize, disconnect the load under the approved procedure, inspect it, and follow the manual’s reset test.
Conductor, pin or output-stage failure
A broken black conductor, bent connector contact, damaged pigtail or failed transistor can leave the sensing electronics alive while the external output path fails.
First proof: continuity with power removed, then a same-model substitution on a known-good compatible branch.
Two-wire exception: a two-wire sensor is part of the load’s series circuit. Its OFF-state leakage and ON-state residual voltage are expected electrical quantities, not automatically evidence of failure. Check the exact minimum load, leakage and residual-voltage limits.[2] The related XSZ guides explain sensor leakage current and residual voltage.
Checkpoints 3 and 4
A correct terminal voltage is not enough unless the PLC sees valid voltage and current for long enough.
Digital inputs do not switch at one universal voltage. They have model-specific OFF and ON regions, current demand, common architecture and input delay. A sensor can produce a plausible meter reading while the input remains below its guaranteed ON region or lacks a complete current path.
As one model-specific example, Siemens documents an ET 200SP 24 VDC high-speed input with a signal-0 region of −30 to +5 V, a signal-1 region of +11 to +30 V, typical signal-1 current of 6 mA and selectable input delays from none to 20 ms. Those values explain why threshold and pulse width matter; they must not be copied to a different module.[3]
Input filtering can also hide a real pulse. Siemens states that a pulse shorter than the configured input delay is suppressed and does not appear in the process image for the cited module. Rockwell likewise documents configurable digital-input filters, with the reminder that the setting does not include all hardware, field-device and supply delays.[4]
Current path or common is wrong
- PNP sensor paired with an incompatible input/common
- NPN sensor has no +V-side path through the input
- Input group common is open or landed on another group
- Sensor and PLC supplies do not share the required reference
Electrical state misses the threshold
- Supply drops at the sensor under load
- Residual voltage or cable drop reduces ON margin
- Input current demand exceeds available output path
- Leakage keeps the signal inside an undefined region
Module rejects or never publishes the event
- Input pulse is shorter than the configured filter
- Channel is disabled, faulted or assigned differently
- Connector/terminal is loose despite a static meter reading
- Input channel is damaged and requires controlled comparison
Useful next reference: the XSZ PLC sensor input guide explains sourcing, sinking and common terminology without relying on wire color alone.
Checkpoint 5
If the raw input changes, stop troubleshooting the sensor output.
Once the controller’s raw physical input follows the target, the sensor, cable and input hardware have delivered an accepted state. The remaining issue is how that state is mapped and used.
Address or tag mapping
The program may read another slot, channel, alias or copied tag. Compare the raw module input with the exact tag used in the rung or function block.
Force or override
A force can override the actual input value seen by program logic. Rockwell documents that an input force replaces the received input value regardless of the physical state.[5]
Sequence and interlocks
The input may be valid while mode, permissive, timer, latch, safety state or downstream output logic intentionally blocks the machine action.
Do not use forcing as a casual field test. Follow the controller manufacturer’s procedure and the site’s change-control and machine-safety rules. Record and remove test forces deliberately. A forced input can make online logic look correct while hiding the actual field state.
Symptom to first proof
Match the observation to the next measurement.
Each row starts where the previous checkpoint already passed. This keeps the test focused and avoids changing several variables at once.
| Observed condition | Most useful first check | Likely zone | Avoid this shortcut |
|---|---|---|---|
| Power LED on; operation LED never changes | Confirm target, sensing margin, teach/mode and exact indicator legend | Sensing or configuration | Do not assume every illuminated LED means detection |
| Operation LED changes; loaded output does not | Verify output type, NO/NC mode, reference, load and protection state | Output/test circuit | Do not declare a failed transistor from an unloaded voltage alone |
| Connector output changes; PLC terminal does not | Compare both points during the same event; then continuity-test de-energized | Cable, connector or terminal | Do not move conductors while energized |
| PLC terminal is valid; channel/raw input stays off | Check common, threshold/current, filter, channel configuration and diagnostics | PLC input module | Do not copy another module’s threshold values |
| Channel LED reacts; event missing at speed | Compare pulse width with sensor response, module delay, update and scan path | Timing/filtering | Do not test only with a stationary target |
| Raw input changes; program tag stays fixed | Trace mapping, alias, copied data and force status | Software data path | Do not replace field hardware |
| Program tag changes; machine does not act | Inspect permissives, mode, interlocks, state sequence and commanded output | Control logic/downstream actuator | Do not bypass protective logic to “prove” the sensor |
Worked diagnostic example
A high output at the sensor but no PLC input does not immediately prove a dead channel.
24 VDC control system
Replacement decision
Replace the sensor only after the fault follows the sensor.
A lit LED with no PLC response is a symptom, not a diagnosis. Build enough evidence that a replacement will solve the same failed checkpoint.
Exact configuration confirmed
Model suffix, supply, output polarity, output logic, connector pinout and indicator meaning match the installed circuit.
Supply is valid under operation
Voltage and reference are measured at the sensor during the fault, not only at the cabinet power supply.
Compatible loaded test completed
The output is tested with the correct current path and load without exceeding sensor or PLC limits.
Cable path is proven
Connector pins, conductor continuity, junctions, terminals and common are checked under a de-energized procedure.
PLC input behavior is verified
Common polarity, guaranteed thresholds, input current, filter, enable status, diagnostics and channel mapping are known.
Controlled substitution is conclusive
The fault follows the suspect sensor, while an approved known-good same-model unit restores the same test branch.
For a useful supplier review, send evidence rather than only “LED on, no output.” Include sensor and PLC model numbers, wiring diagram, connector pinout, supply measured at the sensor, output readings in both target states, input common, input threshold/filter settings, cable length, load, photos and a short event video. The XSZ sensor datasheet guide shows which fields to collect, and the M12 pinout guide helps prevent connector-view mistakes.
Application and replacement review
Need to match an XSZ sensor output to your PLC input?
Send the exact sensor, PLC module, common wiring, supply, target state, required logic, connector, cable length and measured results. XSZ can help screen the output configuration and prepare a more complete sample or replacement specification.
Frequently asked questions
Sensor LED on but no output FAQ
Does a lit sensor LED mean the output is ON?
Not always. First identify whether the LED indicates power, operation/output, sensing stability, teach status or a fault. Even an operation LED that follows the commanded output does not prove that a compatible loaded signal reached the PLC terminal or passed the input threshold.
Why does a PNP sensor LED turn on while the PLC input stays off?
Common causes include an incompatible PLC input/common, an open 0 V return, wrong NO/NC mode, incorrect pinout, broken output conductor, voltage below the PLC ON threshold, excessive input filtering, a disabled channel or a failed output stage. Trace the loaded signal from connector to terminal before replacing the sensor.
Should a triggered NPN sensor output measure 24 V?
Not when measured from output to 0 V in a conventional active NPN NO circuit. The NPN transistor sinks the output toward 0 V while current flows from +V through the PLC input or load. Use the exact circuit diagram, state, load and measurement reference.
Can a broken output wire leave the sensor LED on?
Yes. The sensor electronics can remain powered and the indicator can respond while the output conductor, connector contact, junction or terminal is open. Compare the loaded state at the sensor connector and PLC terminal, then perform continuity testing only with power removed.
Can a PLC input filter hide a real sensor signal?
Yes. A pulse shorter than the configured input delay may be rejected and never enter the process image. Compare the real target dwell and sensor output pulse with the exact module filter, hardware delay, update path and controller scan requirements.
Why does the PLC channel LED change but the program tag does not?
The program may read a different channel, address, alias or copied tag, or an I/O force may override the field value. Compare the raw module input with the exact tag used in logic and inspect all force/override status under the controller and site procedures.
Should I add a resistor to test the sensor output?
Do not use one fixed resistor value for every sensor. Prefer the real compatible PLC input or a manufacturer-approved test method. If a separate test load is engineered, its resistance and wattage must respect supply voltage, minimum and maximum output current, residual voltage and fault conditions.
When is it reasonable to replace the sensor?
Replace it after the exact model/configuration, supply, loaded output test, cable path and PLC input compatibility are proven, and a controlled same-model substitution shows that the fault follows the suspect sensor while the known-good unit restores operation.
Technical references and image credits
- OMRON E2E Standard Proximity Sensor Datasheet — model-specific operation indicators, timing charts, NPN/PNP output circuits and protection specifications.
- OMRON Proximity Sensor Terms — output configurations, NO/NC behavior, two-wire leakage current and output residual voltage.
- Siemens ET 200SP DI 8x24VDC HS Equipment Manual — one published example of input voltage regions, input current and parameterized input delay.
- Rockwell Automation: 1756 Digital DC Input Configuration — configurable input filters and their scope.
- Rockwell Automation: Force — how an input force overrides the received physical input value in controller logic.
- OSHA: Control of Hazardous Energy (Lockout/Tagout) — responsibilities and procedures for controlling hazardous energy during service and maintenance.
Photography: electrical-panel diagnosis, loaded measurement and control-panel detail via Pexels. Product image: XSZ Sensor.
Scope: This guide covers ordinary industrial sensor and PLC input troubleshooting. Safety outputs, hazardous locations, mains-voltage circuits and validated machine-protection functions require the exact certified documentation and qualified engineering procedures.