Breadcrumb: Home > Solutions > Speed Measurement
Updated:2026-11-17
Views: When a speed sensor fails it rarely goes dark. More often the reading comes and goes, the PLC sees no pulses, or the display jumps. The maintenance window is one or two hours at most, and the engineer on site usually has a multimeter and a screwdriver. Fitting a new sensor is the easy move, yet the new unit often behaves exactly the same, because the fault was never in the sensor. This article separates the W and P variants, maps six field symptoms to their root causes, gives a three-minute bench test, and defines the four wires.
A speed sensor of this kind rarely fails by going dark. More often it reads intermittently, sends no pulses to the PLC, or makes the display numbers jump. The window for a maintenance stop is often no more than one or two hours, and the engineer on site usually has nothing but a multimeter and a screwdriver. Plugging in a new sensor is the convenient route, but it is not rare to find that the new sensor does not work either, because the fault was never in the sensor to begin with.
The four models ROS-W/ROS-P and ROLS-W/ROLS-P share most of their construction: the same 303 stainless steel threaded barrel, the same 1-250 000 rpm range, the same three coloured leads plus shield, and the same pulse for every reflection. The differences sit in two places - whether the light source is an LED or a laser, and whether the cable ends in tinned wires or a 1/8 inch plug. Once those two points are clear, troubleshooting follows one common path.
This article follows the order in which work can actually be done on site: look at the lamp, then measure voltage, then check the wire sequence, then decide whether the sensor is alive, and leave return to factory or replacement to the end.
Confirm which version you are holding before anything else. The symptoms of W and P are almost identical, but the entry point for troubleshooting is different: on the W version you check the terminal block and the supply inside the cabinet, on the P version you check the plug, the socket and the supply of the instrument it is plugged into.
| Model | Cable end | Cable length | Supply | Output | Typical downstream device |
|---|---|---|---|---|---|
| 4 tinned wires (brown/blue/black + shield) | 8 ft (2.4 m) | 3.0-15 Vdc, approx. 40 mA | Negative pulse (+V pulled to 0) | Terminal block, ACT series panel meters, F2A converter, DataChart | |
| 1/8″ (3.5 mm) male plug | 8 ft (2.4 m) | 3.0-15 Vdc, supplied by the instrument it is plugged into | Negative pulse | PLT200, Nova-Pro 300/500, PLS stroboscopes | |
| 4 tinned wires | 8 ft (2.4 m) | 3.3-15 Vdc, power consumption ≤0.4 W | Open collector (internal 4 700 Ω pull-up to 3.3 V, load ≤25 mA) | Terminal block, PLC input module, data logger, panel meter | |
| ROLS-P (laser) | 1/8″ male plug | 8 ft (2.4 m) | 3.3-15 Vdc | Open collector | Handheld tachometer, stroboscope, SPSR-IM power module |
| ROLS24-W (laser) | 4 tinned wires | 8 ft (2.4 m) | 24 Vdc ±10% | Open collector | On-site 24 Vdc systems, no step-down needed |
The optical figures of the four models are worth remembering, because every distance and angle question depends on them: the LED version (ROS) is rated at 0.9 m with a permitted offset angle of 45°; the laser version (ROLS) is rated at 7.6 m with 60°, 650 nm Class 2, maximum output 1 mW, beam divergence below 18 mrad (spot about 4×7 mm at 2 m). Both versions cover 1-250 000 rpm and are rated for −10 to 70 ℃.
The full model and cable-length table, and how to choose W or P by the downstream interface, are covered in more detail on two collection pages: ROS-W/ROS-P optical speed sensors, ROLS-P/ROLS-W laser speed sensors.
Classifying what the field calls a failure is more useful than taking the sensor apart. The table is ordered by frequency, and the top three rows account for most real cases.
| Symptom | Usual root cause | Check before repair | Fixable on site |
|---|---|---|---|
| No pulses at all at the meter or PLC | No supply, reversed wire order, target outside the field of view, failed sensor | Green lamp flashing → terminal voltage → wire order | The top three can all be fixed on site |
| Reading exactly twice the real speed, or another integer multiple | More than one reflective tape on the target, or a second bright reflector passing the beam | Count the bright spots per revolution | A single tape is enough |
| Reading drops in and out, counts occasionally lost | Dirty or lifting tape, excessive offset angle, distance at the limit, loose terminal | Whether the green lamp flashes regularly | Cleaning, re-taping or re-aiming |
| No reading at low speed (below 500 rpm) | Tape too wide: the probe is square to the target and the spot covers the whole tape, giving continuous reflection | Tape width and aiming angle | Narrow tape (about 5 mm), aimed square |
| Reading low, with error growing as speed rises | Pulses filtered out downstream, or too long a filter constant on the PLC input | Downstream input filter setting | Change the parameter |
| Still no signal after a new sensor is fitted | Insufficient supply capacity in the cabinet, VFD interference on a shared circuit, damaged PLC channel | Measure voltage off-load, try another channel | Solvable, but often misdiagnosed as a sensor fault |
The table implies one rule: as long as the green lamp is on target, the sensor itself is most likely good, and the problem lies in the tape, the angle, the wiring or the downstream device. The sensor itself becomes the suspect when the green lamp does not flash at all.
Every ROS and ROLS unit has a green On-Target indicator in the cable-end cap. Its meaning is unambiguous: it toggles once for every valid reflection received. At low speed the eye can see it flash in step with rotation; as speed rises the flash rate exceeds persistence of vision and the lamp appears continuously lit.
The lamp is therefore the direct way to tell whether the optical path is working, and much faster than guessing from the meter. Three states, three conclusions:
When the lamp stays dark, work through the four points below in order; in most cases the third one solves it:
One point that is easily overlooked: if the target has keyways, bolt heads, polished or brushed surfaces, the change in reflection as they cross the field of view can be as strong as the tape, and the symptom is one pulse per bolt. On bright metal targets, place the tape on a relatively rough area instead, or mask the interfering zone with matte black paint.
The sensor is off the machine and there is no spare of the same model on site. How do you confirm whether it is good or bad? The LUYOR manual gives a practical method worth writing on the bench: aim the probe at an older fluorescent lamp and the output should show a square wave at twice the line frequency.
The reason is that a fluorescent tube extinguishes once per half cycle at line frequency: a 60 Hz supply gives 120 Hz flicker, a 50 Hz supply gives 100 Hz. The sensor treats that flicker as one reflection per revolution, so:
Two conditions apply. The method needs tubes on a conventional magnetic ballast; the electronic ballasts and LED tubes now in wide use run at tens of kHz and give completely wrong readings, so they cannot be used for this check. The probe also has to be close enough to the tube that the tube fills the field of view.
It also works without an oscilloscope: set a multimeter to DC volts and measure the signal wire against common. While aimed at the tube the voltage swings between two levels (between the high level and 0 on the negative-pulse version, between the internal 3.3 V pull-up and 0 on the open-collector version); when the tube is moved away it returns to the idle level. A swinging reading means the photoelectric stage and the output stage still work; a dead reading sends you on to the supply and the cable.
The value of this bench test is that it separates a failed sensor from a badly installed one. Often the sensor is good and the problem is a new bracket that changed the angle, or tape smeared with oil. Proving the unit alive at the bench before looking for installation faults on site saves an unnecessary return.
In the official manual the W version has four leads: three coloured wires plus a shield. Our product page writes "3 tinned wires (shield included)", meaning three functional wires; the shield is brought out separately for grounding. Do not cut it off as a spare tail, because it serves both as the shield and as the case ground.
| Wire colour | ROS-W | ROLS-W | ROLS24-W |
|---|---|---|---|
| Brown | Supply + 3.0-15 Vdc | Supply + 3.3-15 Vdc | Supply + 24 Vdc ±10% |
| Blue | Common | Common | Common |
| Black | Signal (negative pulse +V→0) | Signal (open collector) | Signal (open collector) |
| Shield | Case ground | Case ground | Case ground |
Cutting the plug off the P version and wiring it to a terminal block is common practice, but the wire colours differ between the plastic and the metal plug, and this is where miswiring usually starts. On the metal plug the tip is signal (black), the middle band is supply (brown) and the base is common (blue); inside the plastic plug the tip is signal (red), the middle is supply (black) and the base is drain/common. Wiring by the rule of thumb "red is positive, black is negative" reverses supply and signal on a plastic plug.
The second trap is the output type. All ROS versions output a negative pulse: at rest the signal wire sits at the supply voltage and is pulled to 0 once per revolution. ROLS is open collector by default: at rest an internal 4 700 Ω resistor pulls it to 3.3 V, and it is pulled low when conducting, with a load current limit of 25 mA. TTL or negative-pulse versions must be specified when ordering.
The difference between the two outputs at a PLC input is practical:
Wiring details, how to identify the PLC input type and the terminal numbers of common downstream devices are listed item by item in Laser speed sensor wiring and outputs explained, and are not repeated here.
A flashing green lamp with a dead meter, or pulses far below supply amplitude on a scope, is almost always a supply or cable problem. Measure three points in this order:
When the cable is too short, a 25 ft (7.6 m) extension EC-25P (plug to plug) and bulk three-core shielded cable are available. To lengthen a W version, make the joint inside a terminal box and carry the shield through the joint; do not twist wires together inside a cable tray. If pulses are lost after lengthening, look at the shield at the joint.
The sensor is a single sealed assembly of optics, optoelectronics and analog circuitry, so the boundary of what can be repaired is clear.
| Fault | Fixable on site | Action |
|---|---|---|
| Oil, dust or water marks on the lens | Yes | Wipe gently with anhydrous ethanol and a lint-free cloth; no acetone or other strong solvents (the lens is acrylic) |
| Tape aged, lifting or missing | Yes | Clean the surface and re-apply T-5; use T-5WP waterproof tape in wet or washdown areas |
| Loose bracket, shifted angle, loose nut | Yes | Re-aim until the green lamp flashes regularly, then lock with the two M16 (ROS)/M18 (ROLS) lock nuts |
| Loose terminal, reversed wire order, shield not connected | Yes | Re-crimp, check against the wire colour table, bond the shield to the cabinet earth |
| Supply voltage too low, interference on a shared circuit | Yes | Use a dedicated supply, separate the routing, add isolation |
| Cable jacket damaged, cores intact | Depends | Seal the damaged section against water; if the damage is at the cable exit it is structural and the unit goes back |
| Broken core inside the cable (snapped at the exit) | No | The cable exit is potted into the body; no reliable field repair |
| Laser or LED dark, fogging or mould inside the lens | No | Sealed optical assembly; return for inspection |
| Liquid inside the housing, corroded terminals | No | Moisture has entered; return for assessment or replace |
| Output stage failed (no swing, stuck low or high) | No | The board is potted inside the barrel; return to factory |
| Systematic reading error | No | Pulse counting has no zero drift; a steady error usually comes from the tape or the downstream device, so rule those out before returning |
In one sentence: anything you can tighten, wipe or reconnect is handled on site; anything involving the optics, the circuit or the cable exit inside the body goes back. Opening the body is not an option: the 303 stainless barrel is potted and press-fitted at both ends, so opening it scraps the unit and loses its protection.
The environmental limits of these sensors are often overestimated. The ROS manual gives a humidity condition of 80% relative humidity at 31 ℃, falling linearly to 50% at 40 ℃, which means it is not a waterproof unit for continuous spray or high-pressure washing. The stainless barrel resists corrosion, but the cable exit and the end-cap indicator are the weak points.
For permanent outdoor or wet-shop installation the practice is:
The laser version also has a safety boundary to respect: Class 2, maximum output 1 mW, a brief sweep is not harmful, but do not stare into the beam and do not view the beam path through optical instruments such as binoculars or magnifiers. Switching off before working on the beam path is a simple and effective habit.
Routine checks can be built into the machine maintenance sheet: once a week confirm the green lamp flashes regularly, once a month wipe the lens and the tape, once a quarter re-tighten the lock nuts and check the cable for wear at the conduit entry. Doing these three things lowers the chance of an unplanned stop.
Once the fault is confirmed to be in the sensor itself and the warranty has expired, the standard route is a factory repair. The LUYOR process is: submit the repair/calibration request online (model, serial number and symptom) → receive an RMA number by email → ship as instructed with the RMA number on the outside of the package → assessment and a formal quotation on arrival → repair or calibration after payment is confirmed → return shipment. Talk to technical support once before shipping: many suspected faults are eliminated at that step.
A few practical details matter. State in the request whether the device has been in contact with hazardous materials (coolant, chemicals, radioactive substances), because that decides whether it can be opened on arrival. Instruments with batteries must be handled to transport rules. A sensor repair does not necessarily include calibration, but if the instrument is inside a quality system and a traceable certificate is required, arrange it at the same time. A traceable calibration (NIST traceable) typically takes 10-15 days, the standard certificate is valid for one year, and a two-year interval can be requested.
When replacement is the better economics:
| Situation | Preference | Reason |
|---|---|---|
| Liquid inside, internal corrosion | Replace | Assessment plus repair approaches the cost of a new unit, and reliability is lower afterwards |
| Cable snapped at the exit, body must be opened to replace it | Depends on age | Repairable under two years of service; older units are better replaced |
| Discontinued model, spares no longer supplied | Replace or upgrade | An equivalent model can take its place, but interface and output must be checked again |
| Few units in service, high cost of downtime | Spare plus replacement | Keep one spare for critical points so production continues while the repair is away |
| The fault is actually installation or downstream | No repair | Work through sections 4 and 5 before blaming the sensor |
Choosing a replacement follows different logic from repair: select for the site conditions rather than settling for the old model. Where ambient light is strong, sunlight falls directly on the target or the lens condenses frequently, consider the frequency-modulated laser model with better light rejection, ROLS-5PW (7 m class, built-in high-frequency filter). For portable inspection a plug version is more flexible; if the signal has to reach a control room, use the LUYOR-912 speed transmitter to convert to 4-20 mA/RS485. Where local display and alarm are enough, the LUYOR-910 panel tachometer connects with three wires. For how to choose by downstream interface, cable length and light source class, see the laser speed sensor selection guide; the section overview is speed sensors.
The flashing lamp shows that the optical path is working, so the fault is in the electrical section. Check in turn: is the signal wire on the correct terminal (black is signal), does the polarity match the downstream device, does the input need sinking or sourcing wiring, is a pull-up resistor required, is the PLC channel damaged? The quick test is to connect the sensor to a spare meter of the same model or to an oscilloscope: a reading there means the downstream configuration is at fault.
Two pulses are produced per revolution. The usual cause is two tapes on the target, or one tape split in two by the spot, or a keyway, bolt head or polished surface adding a reflection as it passes. Stop the machine, turn it over by hand and count how many times the green lamp flashes per revolution.
Not advisable. The 303 stainless barrel is potted and press-fitted at both ends; opening it removes the sealing and the protection, and the optics and circuit cannot be reset without the proper tooling. Faults inside the body, in the optical assembly, in the output stage or at the cable exit should all go back to the factory.
Set a multimeter to DC volts and measure the signal wire against common, then aim the probe at a fluorescent tube on a magnetic ballast. The voltage should swing between two levels (between the high level and 0 on ROS, between the 3.3 V pull-up and 0 on the open-collector ROLS) and return to the idle level when the tube is moved away. A swing means the photoelectric stage and the output stage are working. Electronic ballasts and LED tubes do not work for this test.
Yes, but the joint must be inside a terminal box, with the shield carried through and grounded at one end. Twisting wires together in a tray and taping them introduces interference and loses pulses. If the break is at the cable exit it is structural damage and the remedy is a factory cable replacement. For longer runs a 25 ft (7.6 m) finished cable and a plug-to-plug extension are available.
No. ROLS-W accepts 3.3-15 Vdc and 24 V will damage it. Two options: use ROLS24-W (the 24 Vdc ±10% version), or add a 24 V to 12 V isolated supply module in the cabinet. ROS-W has a 15 V upper limit and likewise cannot be connected to 24 V directly.
Start by finding where the water enters: most often at the cable exit and the end-cap indicator, next at the seam between cable jacket and body. Remedies are to move the probe and wiring under a roof, turn the terminal box entries downward and add desiccant, apply secondary sealing at the end cap, and change the tape to a waterproof type. If water has already reached the inside, the unit must go back for assessment.
A pulse-type speed sensor has no zero drift: it counts pulses and the downstream device converts them, so a functional check is usually enough after optical or output parts are replaced. If, however, the machine is inside a quality system and test reports cite traceable data, a traceable calibration should be performed and the certificate kept; the standard certificate is normally valid for one year.
This article belongs to the speed and stroboscopic measurement section.
Other sections on the same product line include laser tachometers (aim at reflective tape and read RPM directly, one-handed), industrial test instruments (the entry point for the four lines: UV inspection, speed measurement, stroboscopic observation and surface inspection).
Models mentioned in this article: LUYOR-912, LUYOR-910, LUYOR-925. Ranges, accuracy and accessory lists for each model are listed on the product pages.
Email: info@xinysun.com
Tel: 131-2093-6518
Address: 1525 Minqiang Road, Bldg 26, Songjiang, Shanghai
Copyright 2026 Xinysun Instrument (Shanghai) Co., Ltd. All rights reserved.
ICP:ICP-2026046447
