Coil, ferrite, electrode, optics, electronics, active-face position and potting all affect range.
Sensor manufacturing and supplier guide
How Factories Control Sensor Switching Distance Tolerance
A sensor factory controls switching distance by controlling the product and the measurement system together. The target, approach, mounting, voltage, load, temperature and timing must be fixed before coil, electronics, assembly, calibration or process data can be judged.
Direct answer
Capable factories define one reproducible test condition, control the sensing element and mechanical stack, trim the electronics where the design permits, measure operate and release points on a validated end-of-line station, and monitor continuous distance data for drift. An LED-only pass check is not enough.
Target material, size, thickness, angle and motion determine the measured switch point.
Axis, fixture, voltage, load, sampling and decision logic can add false variation.
Continuous values, SPC, traceability and reaction rules prevent drift from reaching shipment.
“Tolerance” can mean four different limits.
Separate them before comparing suppliers. A catalog value, a standard relationship, a factory control limit and the machine's working gap protect different decisions.
The range or tolerance publicly stated for the exact orderable model under named conditions.
Answers: what the supplier promisesThe meaning and relationship of operating-distance terms for the applicable sensor class.
Answers: how terms are comparedA tighter internal boundary that triggers adjustment, investigation or segregation before specification failure.
Answers: when the process must reactThe usable installation distance after target, motion, environment, tolerances and risk margin are included.
Answers: where the machine should runSwitching-distance tolerance is the allowed spread of a defined switch point.
Plain-language definition
Switching distance is the gap between a sensor's reference surface and a defined target when the output changes state. Its tolerance is meaningful only when the target, movement, mounting, electrical conditions, temperature and measurement rule are also known.
The target can approach the sensor or move away from it. Those two positions are normally different because hysteresis separates the operate and release points. A factory may also repeat the movement several times to measure short-term repeatability.
OMRON's official proximity-sensor terminology defines sensing distance using a standard object moved by a specified method. It separately defines set distance as a stable-use distance that includes temperature and voltage effects; its general illustrated guidance places this at about 70% to 80% of rated distance. That percentage is manufacturer guidance, not a universal rule for every product.[3]
Specify material, dimensions, thickness, surface, flatness, angle and target identification.
Define axis, direction, starting point, speed, repetitions, debounce and edge logic.
Control flush class, surrounding clearance, sensor seating, torque and neighboring sensors.
Specify terminal voltage, ripple, output type, load or PLC simulator, cable and warm-up.
Set a temperature band, stabilization time and family-specific light, humidity or acoustic limits.
State limits, guard band, uncertainty treatment, rounding, retest and disposition before testing.
Buyer warning: Do not copy a tolerance from another M18 sensor because the housing diameter and nominal distance look similar. Shielding, target, active face, mounting and environmental conditions may be different.
The familiar inductive case
Sn, Sr, Su and Sa answer different distance questions.
IEC 60947-5-2:2019 covers several proximity-switch technologies. The numerical relationships below are the familiar relationships published for standard inductive proximity sensors and should not be applied automatically to capacitive, photoelectric, ultrasonic, magnetic or special designs.[1][2]
0.9 Sn ≤ Sr ≤ 1.1 Sn
0.9 Sr ≤ Su ≤ 1.1 Sr
0 < Sa ≤ 0.81 Sn
Use the exact selected model's current datasheet and test conditions for acceptance.
Sensing faceIncreasing target distance
Sa assured zone
Su usable range
Sr effective range
Sn nominal reference
Worked example
An 8 mm nameplate value does not mean set the machine at 8 mm.
The nominal value identifies a distance class. The usable machine gap still depends on the real target and installation.
| Value | Calculation | Result | Correct interpretation |
|---|---|---|---|
| Sn | Declared | 8.00 mm | Nominal reference, not a guaranteed working gap. |
| Sr band | 0.9 × 8 to 1.1 × 8 | 7.20 to 8.80 mm | Allowed individual effective distance at the stated reference setup. |
| Two-stage lower boundary | 0.9 × 0.9 × 8 | 6.48 mm | Basis of the familiar 0.81 Sn assured boundary for the applicable inductive case. |
| Real machine gap | Model and application specific | Not calculable from Sn alone | Must include target, motion, mounting, temperature, voltage, contamination and risk margin. |
Do not transfer 6.48 mm to a small copper, aluminum or tilted target. The calculation is based on the applicable standard reference relationship. A real part can couple less strongly than the reference target and may need a smaller working gap.
Do not combine unlike specifications
Tolerance, hysteresis, repeatability and response speed are not the same.
| Characteristic | Simple meaning | Factory measurement | Why it matters |
|---|---|---|---|
| Operate distance | Where output changes as the target approaches. | Repeated defined approach. | Main value for range acceptance and process centering. |
| Release distance | Where output resets as the target leaves. | Repeated withdrawal on the same path. | Confirms reset behavior and enables hysteresis calculation. |
| Hysteresis | Difference between operate and release points. | |soperate − srelease| under stated conditions. | Prevents chatter but also affects positional precision. |
| Repeatability | Spread of repeated switch-point readings on the same unit. | Several cycles with setup and time controlled. | A unit can repeat tightly while being off center. |
| Response time / frequency | How quickly output follows a moving target. | Defined speed, target and spacing. | A correct static distance does not prove high-speed detection. |
Definitions of sensing distance, hysteresis, response time and response frequency: OMRON Proximity Sensor Explanation of Terms.[3]
Variation comes from a chain
Why does switching distance vary between sensors?
The finished switch point is the combined result of the sensing element, electronics, mechanical position, encapsulation, test target, electrical setup and measurement system.
Sensing element
Coil turns and geometry, ferrite position, electrode spacing, emitter output, receiver sensitivity or transducer behavior.
Electronics
Oscillator amplitude, gain, comparator threshold, reference, component tolerance, self-heating and output load.
Mechanical stack
Element depth, active-face thickness, lens position, housing deformation, seating and final tightening.
Potting and cure
Resin identity, fill, shrinkage, movement, stress, voids, cure profile and dielectric effects.
Reference setup
Target, alignment, nearby metal, terminal voltage, load, temperature, speed and software timing.
Measurement system
Axis backlash, encoder bias, fixture repeatability, target wear, channel timing and station-to-station agreement.
From requirement to shipment
The factory controls the distribution, not only failed units.
Final testing is necessary, but it cannot economically repair a process that continually produces units close to both limits.
Define the test
Lock model, target, mounting, movement, electrical conditions, temperature, repetition and acceptance rule.
Budget the tolerance
Allocate variation across sensing element, electronics, mechanics, environment, calibration residual and measurement uncertainty.
Control suppliers
Identify material and component characteristics that move range; retain lots and approve changes.
Control assembly
Use datum-based fixtures for element depth, alignment, active face, housing, potting and final torque.
Calibrate where designed
Trim analog or digital thresholds only when the architecture and validated process support it.
Test the finished sensor
Measure approach, release, hysteresis, repeat spread and electrical output after final assembly.
Monitor the process
Chart continuous values by line, fixture, material lot, cavity, recipe, shift and other suspected sources.
Trace and react
Segregate failures, control retest and rework, investigate drift and preserve shipment evidence.
Assembly and potting control
A correct circuit can still switch at the wrong distance.
The sensing element must remain at a controlled position relative to the active face. Potting and final tightening can move or stress that stack, so the finished product needs another functional test.
- Use physical datums for coil, electrode, lens, transducer or magnetic element position.
- Inspect alignment and depth before encapsulation while the structure is accessible.
- Control resin batch, dispense, critical voids, fixture restraint, time and cure temperature.
- Repeat target-based range testing after cure and final assembly.
- Correlate dimensional, electrical and functional data when the mean or spread moves.
Calibration is architecture specific
Not every sensor is individually calibrated in the same way.
A sensor can meet its specification through capable design and assembly, per-unit trim, teach functions or multi-point compensation. The correct question is what remains after the chosen method.
No individual trim
Design and process capability place units inside tolerance, followed by final functional test.
Needs: capable components and marginAnalog trim
A resistor, potentiometer or trimmed element adjusts threshold, oscillator, gain or offset.
Needs: drift and sealing controlDigital coefficient
Offset, gain or temperature terms are stored in memory and verified by software.
Needs: revision and memory integrityTeach setup
A target and background establish a threshold or window for the intended application.
Needs: controlled teach stateMulti-point compensation
Several reference points model offset, slope or temperature behavior over a wider range.
Needs: validated interpolationSequence matters: If potting, cure, burn-in or thermal cycling can move the switch point, engineering must decide whether calibration occurs before that operation, after it, or both. A pre-potting calibration label is not proof of the finished sensor.
Finished-product proof
How an automated end-of-line test measures switching distance
A repeatable fixture holds the sensor while a defined target moves on a calibrated axis. The station records target position and output state together instead of relying on an operator to watch an LED.
- Identify part number, revision, output and controlled test recipe.
- Verify target ID, fixture, axis home, supply/load, software and reference check.
- Mount the sensor, apply terminal voltage and allow defined stabilization.
- Approach from a confirmed non-detect position at validated speed.
- Capture operate point, then reverse and capture release point.
- Repeat, calculate hysteresis and spread, and verify electrical output.
- Store raw values, station, recipe, conditions, time and disposition.
Steel reference targetMaterial, size, thickness, surface and ID controlled
Motorized linear axisEncoder position + defined approach and withdrawal speed
OperateRelease
Sensor fixtureDatum, flush class, alignment, torque and surrounding metal
Regulated supplyVoltage measured at sensor terminals
PLC input / loadOutput voltage, current, leakage and state
DAQ and recordPosition, timestamp, edge logic and serial traceability
100% test and sampling can coexist. A short, nondestructive reference switch-point test may suit every unit. Temperature, humidity, EMC, vibration, ingress and life testing normally use qualification, change validation or risk-based sampling because they take longer or may consume product life.
The tester can create variation too
Control every element that can bias the result.
Calibration of the linear axis is necessary, but it is not the whole measurement system. Target condition, fixture alignment, terminal voltage, edge timing and software version can change the reported distance.
TargetMaterial, dimensions, flatness, cleanliness and replacement interval
Linear axisResolution, calibration, backlash, straightness, speed and home repeatability
FixtureReference surface, tilt, seating, surrounding metal and clamping force
Electrical setupTerminal voltage, ripple, wiring, connector and representative load
Data logicSampling, timestamps, debounce, interpolation, edge logic and software revision
EnvironmentTemperature, warm-up, airflow, humidity, light or acoustic noise as applicable
Measurement-system analysis
Resolution
Repeatability
Reproducibility
Bias and linearity
Stability
Part variation
Can the factory trust the number from its tester?
A wide result distribution can come from the product, the measurement system or both. MSA should evaluate the station over the actual distance range and include the factors that exist in production.
Can the system distinguish changes small enough for the tolerance and control limits?
Does the same station reproduce the same result on repeated cycles and remounts?
Do fixtures, stations, shifts, operators or sites agree?
Is the station high or low, and does that bias change across short and long ranges?
Does the measurement drift with time, maintenance, temperature or target replacement?
Does the study include representative low-, middle- and high-range units?
ASQ describes GR&R as a way to evaluate repeatability and reproducibility and discusses bias, linearity and stability as measurement-system characteristics.[5]
Guard band and decision rule
A value displayed at the catalog limit still has measurement uncertainty. An approved decision rule states how that uncertainty is considered when declaring conformity.[7]
RejectReviewInternal acceptReviewReject
Concept only: The guard-band size must come from the actual requirement, uncertainty, measurement capability and accepted false-accept / false-reject risk. Do not copy the proportions shown here.
ISO states that ISO/IEC 17025:2017 remains current and covers laboratory competence, impartiality and consistent operation. A production line station need not automatically be an accredited laboratory, but calibration, methods, uncertainty awareness and records still matter.[6]
Switching distance over timeConceptual control chart
Upper specification
Upper control limit
Center line
Lower control limit
Statistical process control
A pass counter can stay green while the process moves toward failure.
Continuous operate distance, release distance, hysteresis and repeat spread provide earlier warning than pass/fail totals. ASQ distinguishes statistically derived control limits from customer specification limits; they answer different questions.[8]
- Prove the process is stable before relying on Cp or Cpk.
- Stratify by line, fixture, winding station, PCB lot, resin lot, mold cavity, shift and recipe.
- Review MSA status before blaming the product distribution.
- Do not use one Cpk screenshot as proof of field reliability.
- Set capability acceptance from contract and risk, not a universal number.
Same logic, different physics
Each sensor technology needs its own reference test.
The control sequence stays the same: define, control, measure, analyze and trace. The target and disturbance variables change with the sensing principle.
Inductive
Control coil, ferrite, active face and oscillator; test with the specified metal target, axial approach and mounting class.
Capacitive
Control electrode, guard, face material and potting dielectric; define target, grounding, humidity, deposits and sensitivity.
Photoelectric
Control emitter, receiver, optics, alignment and gain; define target color, reflectivity, angle, background, ambient light and contamination.
Ultrasonic
Control transducer and compensation; define target angle, reflector, air temperature, echoes, dead zone and response behavior.
Magnetic
Control magnet grade, orientation and air gap; test operate and release points with the specified actuator and nearby ferrous material.
Factory conformity is not machine approval
Why can a conforming sensor still miss the real part?
The factory uses a controlled reference target and geometry. The machine may remove the remaining signal margin through target, bracket, motion, environment or controller conditions.
- Smaller, non-ferrous, dark, glossy, transparent or tilted target
- Nearby metal, neighboring sensors, lateral offset or incorrect torque
- Vibration, runout, bracket deflection or mechanical wear
- Temperature extreme, cable voltage drop or changed PLC input
- Dust, chips, moisture, oil, deposits or damaged sensing face
- Target pulse shorter than the sensor-plus-controller response chain
Commercial trade-off
Why does tighter switching-distance tolerance increase sensor cost?
Tightening one number can require changes across components, calibration, test capacity, yield and change control.
Narrower incoming distributions
Higher-grade or sorted coils, ferrites, electronics, optics and mechanical parts may be needed.
More station time and data
Per-unit trim or multi-point compensation adds equipment, cycle time, software and maintenance.
More operating corners
Temperature and voltage points consume chamber capacity and production time.
Lower tester uncertainty
Better stages, encoders, fixtures, targets, automation and MSA may be required.
Smaller internal window
Guard bands can reject more tail units unless the actual process spread improves.
More critical inputs
Materials, firmware, tooling, resin, fixtures and suppliers may need stronger traceability and validation.
Buying decision: If the target and bracket vary more than the sensor, paying only for tighter sensor tolerance may not solve the machine. A larger target, smaller gap, more rigid bracket, different sensing principle or analog displacement measurement can create more usable margin.
Supplier audit
Ask for evidence tied to the exact model and line.
A corporate quality certificate provides context. It does not by itself prove control of one model's switch point.
Strong evidence
- Controlled test method naming target, axis, mounting, voltage, load, temperature and limits
- Model-specific tolerance analysis and characterization across relevant corners
- In-process controls for sensing element, assembly position, potting and final tightening
- Automated operate/release data, tester identity, recipe and traceability
- Fit-for-purpose MSA covering station, fixture, remounting, bias, linearity and stability
- Time-ordered SPC and capability evidence after stability is demonstrated
- Controlled failure, retest, rework, change-notification and revalidation rules
Warning signs
- “We test with a metal plate” with no material, size, angle or drawing
- An operator moves the target by hand and watches only an LED
- Pre-potting range is treated as proof of the finished sensor
- “Calibrated” means only that a pass label was applied
- A calibration certificate is shown while fixture and software variation are ignored
- Only one unexplained Cpk screenshot or overall defect rate is supplied
- Unlimited retest, manual limit changes or data from a different product family
RFQ and sample approval
Give the supplier the conditions that define stable detection.
| RFQ block | Information to provide | Evidence to agree |
|---|---|---|
| Target | Material, minimum size, thickness, shape, finish, angle, speed and contamination. | Reference target plus worst real production samples. |
| Geometry | Min/max gap, offset, runout, vibration, bracket, mounting class, neighboring sensors and torque. | Approved bracket datum and tested installation envelope. |
| Electrical | Terminal voltage, NPN/PNP, NO/NC, load or PLC input, cable, response and startup behavior. | Output and timing test under the real interface. |
| Environment | Temperature, humidity, water, oil, chemicals, dust, light, shock, vibration and IP requirement. | Qualification or sample plan for relevant corners. |
| Product data | Exact distance term, tolerance, hysteresis, repeatability, target and assured/set-distance guidance. | Current model datasheet and written deviations. |
| Production control | 100% versus sampled checks, calibration, retention, traceability, retest, disposition and release. | Sample report tied to exact model and revision. |
| Change control | Components, sensing element, housing, resin, PCB, firmware, fixture, software, site and supplier changes. | Risk-based revalidation and notification agreement. |
For the product-family decision before test planning, review the industrial sensor supplier selection guide and the sensor datasheet guide.
Troubleshooting matrix
Separate product variation from tester and application variation.
| Symptom | Likely categories | First confirmation path |
|---|---|---|
| All units shift after changeover | Wrong recipe, target, fixture, element position, PCB lot, cure, calibration reference or tester bias. | Check reference result, recipe revision, target ID, first-off data, material lots and dimensional correlation. |
| One tester disagrees | Axis bias, alignment, target wear, voltage/load, software timing or stability. | Cross-check the same units and compare raw transitions, datums, reference artifacts and software versions. |
| Room test passes, hot test fails | Temperature coefficient, compensation, supply drop, expansion, cure or material shift. | Separate temperature and voltage tests; measure at sensor terminals and compare geometry and coefficients. |
| One unit repeats poorly | Chatter, marginal connection, unstable oscillator, loose fixture, vibration or contamination. | Record raw output and position over repeated cycles; remount and verify reference stability. |
| Factory passes, machine misses | Real target, angle, bracket metal, speed, vibration, PLC filter, voltage drop or deposits. | Recreate machine geometry and dynamic signal; measure actual gap, voltage, target dwell and output pulse. |
| Cpk falls before defects appear | Mean or spread movement, mixed streams or measurement drift. | Review stability, stratify line/fixture/lot, run reference checks and verify MSA. |
Need stable sensing on a target that does not match the catalog test?
Send the target material, dimensions, angle, speed, gap, bracket, temperature, supply and output, contamination and annual volume. xsz sensor can help organize model selection, sample validation and an acceptance test around the real machine condition.
Continue the evaluation
Related switching-distance and application guides
Frequently asked questions
Sensor switching-distance tolerance FAQ
What is a normal switching-distance tolerance for an inductive proximity sensor?
There is no single tolerance for every model. In the familiar standard inductive relationship, the effective operating distance Sr of an individual sensor at reference conditions lies between 0.9 Sn and 1.1 Sn. Compact or specialized products may publish different values. Always use the exact model's current datasheet and test conditions.
Is rated sensing distance the guaranteed working distance?
No. Rated operating distance Sn is a conventional nominal reference. It does not include every manufacturing and external influence. The real working gap may need to be smaller after target material, size, angle, mounting, temperature, voltage, vibration and contamination are included.
Do factories calibrate every proximity sensor?
Not always. Some designs receive individual analog or digital trim, while others use controlled components and assembly followed by final functional testing. Ask how the exact model meets its specification, what is adjusted and what data are retained.
How do factories measure sensor switching distance?
A typical station holds the sensor in a controlled fixture and moves a defined target along a calibrated axis. It records position when the output changes during approach and withdrawal, then checks operate distance, release distance, hysteresis, repeatability and electrical output against documented limits.
Why must the factory use a standard target?
Target material, size, thickness, shape and angle affect the physical signal that produces detection. A standard target makes production measurements comparable. If the customer's target differs, it still needs separate application validation.
Is 100% switching-distance testing necessary?
It depends on risk, automation, cycle time and process capability. A short reference test may be practical for every unit, while temperature, EMC, vibration, ingress and life tests are usually sampled or performed during qualification and change validation.
What is the difference between switching-distance tolerance and repeatability?
Tolerance describes the allowed spread relative to a specification across units or conditions. Repeatability describes how closely repeated measurements on the same unit agree under controlled conditions. A sensor can repeat tightly but be off center, or meet tolerance while repeating poorly.
Why can a sensor pass factory testing but fail after installation?
The factory uses a defined reference target and setup. The machine can introduce a smaller or different target, tilt, offset, nearby metal, vibration, supply drop, temperature, contamination, high speed or PLC filtering that removes the remaining signal margin.
What quality evidence should a buyer request?
Request the exact test definition, sample switch-point records, target and fixture description, calibration and traceability status, MSA summary, control-chart or capability evidence for the correct model and line, failure-disposition process and change-notification agreement.
Does a high Cpk prove the sensor will work in my machine?
No. Cpk relates a stable measured process to specification limits under stated assumptions and a defined measurement system. It does not validate the real target, mounting, response time, environment or lifetime. Use process capability for supplier control and application testing for machine suitability.
Technical references
Sources used for this guide
- IEC 60947-5-2:2019, Proximity switches — official scope, edition and lifecycle record.
- Pepperl+Fuchs, Operating Distance as Central Characteristic — standard target and familiar Sn, Sr, Su and Sa relationships for inductive sensors.
- OMRON, Proximity Sensors Explanation of Terms — standard object, sensing distance, set distance, hysteresis and response definitions.
- OMRON, Proximity Sensors Technical Guide — target, environment, mounting and selection considerations.
- ASQ, Gage Repeatability and Reproducibility — repeatability, reproducibility, bias, linearity and stability.
- ISO/IEC 17025:2017 — competence, impartiality and consistent operation of testing and calibration laboratories.
- NIST, Decision Rule — treatment of measurement uncertainty when stating conformity.
- ASQ, Control Chart — time-ordered data, control limits, process stability and special-cause variation.
- ASQ, Process Capability — interpretation of capability estimates and stable-process prerequisites.
Image credits: Factory, product, test and application images are xsz sensor website assets. Diagrams are original explanatory graphics created for this page and are conceptual rather than product specifications. This guide does not replace the current standard, exact model datasheet, controlled factory procedure, calibration record or finished-machine validation.