Fiber Optic Sensor for Small Object Detection: Selection Guide

Which fiber arrangement should you try first?

For a small opaque part on a controlled path, start by evaluating a through-beam fiber pair if both sides are accessible. If you can mount on only one side, evaluate a focused reflective head at the actual working distance. Neither choice is complete until the head, amplifier, operating mode and target have been tested together.

A fiber optic sensor carries light between a compact sensing head and a separate amplifier. The head puts the optics into a tight space; the amplifier interprets the returned or interrupted light and provides the output. A small head does not, by itself, guarantee small-object detection.

Area coverage describes a sensing layout, not a guarantee of high resolution. Likewise, a narrow slot can simplify alignment if the part fits through it, but the slot width alone says nothing about the minimum detectable feature.

If the real task is measuring diameter, distinguishing touching parts or identifying a defect, a presence sensor may be the wrong tool. Define the required decision before comparing fiber heads with laser measurement, vision or other sensing methods.

How small an object can the complete system detect?

Use the manufacturer's minimum detectable object specification with its stated conditions, then verify your own target. Do not substitute fiber core diameter, housing diameter or beam diameter for that specification.

Read the size specification and its footnote together

Core diameter describes the light-guiding fiber. Beam or spot size describes an optical region, often at a specified distance. Minimum detectable object describes the target used to establish a detection capability. These are different quantities, even when each is given in millimeters.

Why can an object smaller than the beam still be detected?

A through-beam sensor can switch on a sufficient reduction in received light; the object does not always need to block the entire beam. But a very small obstruction may produce too little usable change, particularly when the receiver is saturated or the target crosses an unfavorable part of the field.

SICK's guidance on achieving minimum object size with a fiber amplifier identifies excessive received light and inadequate signal change as a possible obstacle. “More light” is therefore not a universal fix. Adjust power, gain or operating mode only as supported by the selected amplifier, and compare target-present with target-absent behavior.

For a reflective head, small-spot optics help concentrate the interaction on a feature, but finish, angle and background remain part of the detection problem. A shiny target can send light away from the receiver even when the spot appears correctly placed.

Will the smallest part cross the sensing area every time?

Check the entire permitted target path, not one carefully centered sample. A sensor can meet a minimum-object specification yet miss production parts that pass beside its useful sensing region.

Record the smallest projected feature across the beam, its possible lateral and vertical displacement, tilt, and the distance between head and target. For timing, separately record the feature length along the direction of travel. These dimensions answer different questions.

How fast must the amplifier and PLC input be?

The amplifier must resolve the optical event, the input must capture the resulting electrical pulse, and the machine must act before its process deadline. These are related checks—not one universal calculation that subtracts every downstream delay from target dwell time.

Estimate the event duration, then measure the output

For an opaque target crossing a narrow, fixed beam, a useful first estimate is:

Geometric crossing time [ms] ≈ feature length along travel [mm] ÷ speed [m/s]

For the illustrative contact above at 1.5 m/s, 0.8 ÷ 1.5 ≈ 0.53 ms. This is an approximate geometric time, not a guaranteed output pulse width. Beam extent, threshold, partial obstruction, response behavior and any output timer affect the electrical pulse.

Measure the shortest valid HIGH and LOW durations at the receiving input under the fastest and closest-spaced production conditions. A front-panel display can refresh too slowly to reveal a missed short event.

Check the mode you will actually use

Panasonic's FX-550 specifications illustrate why mode selection matters. With the FT-42 through-beam fiber, the published values include:

FAST is not available on FX-550L variants. Panasonic also notes that free-cut fiber range can decrease by up to 20% depending on the cut. These range figures do not establish minimum-object performance at those distances. Recheck the complete specification after selecting any interference-prevention or timer functions.

Do not confuse pulse capture with processing latency

Check input polarity and electrical compatibility, minimum accepted pulse duration, filtering, and whether the channel uses cyclic sampling, a latch, an interrupt or a counter. The correct capture method depends on how closely events can occur.

For example, Siemens documents that S7-1200 pulse catch can hold a short event until the next input update, but it operates after the input filter. It cannot recover a pulse that the filter removes, and it does not preserve every event when several occur in one scan.

Once an event is reliably captured, program, communication and actuator delays still have to meet the machine's action deadline. They do not necessarily all have to fit inside the original optical crossing time. If output pulse stretching is used, also verify that adjacent parts remain separate events.

How should you teach and mount the sensor for a reliable trial?

Teach the installed arrangement using representative target-present and target-absent conditions. A successful teach indication shows that the setup procedure completed; it does not establish reliability across all production variation.

Challenge the weakest valid signal and the strongest unwanted signal

Include the smallest acceptable part, worst permitted position and angle, surface variation, and actual background. Run missing-part or gap conditions as well as good parts. Where the amplifier provides suitable signal monitoring, look for separation between the two states rather than chasing the largest display number.

Test neighboring sensors together if their light could enter the receiver. Use the manufacturer's supported interference controls and repeat the timing check after configuration changes. If the two optical states overlap, changing a threshold alone cannot provide a dependable separation.

Keep the final fiber route and mounting conditions

• Head and bracket: secure alignment without exceeding the specified mounting torque; include machine vibration and access for cleaning.
• Fiber preparation: use the specified cutting and insertion procedure for free-cut fiber. Do not treat every fiber as field-cuttable.
• Routing: respect the model's bend radius, provide strain relief and keep the trial route representative of the final installation.
• Moving axes: obtain repetitive-flex suitability separately. A static minimum bend radius is not a cable-life rating.
• Environment: check the head, fiber, connectors and remote amplifier individually for the actual temperature, liquids and cleaning exposure.

Perform mounting and wiring changes with the machine isolated under the site's safe-work procedure. Dynamic testing should use an authorized, guarded test method.

Record the configuration and the actual controller result—not only the amplifier LED. Recheck after restart and representative disturbance or contamination conditions. Define the test count, operating envelope and allowable misses or false detections for the process before testing; a short error-free demonstration is not evidence of unlimited reliability.

What should a supplier confirm before you order?

Request a complete configuration that can be reproduced on your machine, together with the conditions behind its performance claim. “Fiber sensor for a 0.3 mm part” leaves too many important choices unresolved.

• Matched components: exact head and amplifier part numbers, suffixes, required lenses or adapters, cable preparation and included accessories.
• Optical task: actual target sample or drawing, material and finish, smallest relevant feature, working distance and permitted path variation.
• Signal and interface: chosen mode, response specification, speed and spacing, output type, PLC/input model and capture settings.
• Installation: mounting drawing, final fiber route, moving or static cable use, environmental exposure and maintenance access.
• Supporting evidence: applicable manuals and a sample-test record linking the target conditions and controller results to that configuration.

A useful selection is a conditional, testable choice: the sensing arrangement fits the access, the smallest feature remains detectable across its path, and the complete signal chain records distinct events at production speed. If one of those points is unproven, resolve that gap before treating a catalog match as a production solution.