Breadcrumb: Home > Solutions > Speed Measurement
Updated:2026-11-02
Views: A UAV propeller does not run at a nominal speed; the speed changes continuously with throttle, voltage and load. A ground test stand has to answer three questions at once: how much thrust and what speed this propeller gives at the present voltage, where the motor and the propeller are matched, and whether vibration or noise comes from the motor or from the blades. Using the ROLS-5PW as the example, this article describes three measurement points on the stand, explains why a frequency-modulated laser is used where sunlight and light flicker interfere, and sets out six typical tests and how speed data is used in each.
A UAV propeller does not run at a nominal speed; it runs at a speed that changes continuously with throttle, voltage and blade load. To reproduce flight conditions on a ground stand, the stand has to answer at least three questions at the same time: how much thrust and what speed this propeller gives at the present voltage; where the motor and the propeller are matched; and whether vibration and noise come from the motor or from the blades. The common precondition for all three is a trustworthy speed reading. This article uses the LUYOR ROLS-5PW fixed-mount laser speed sensor as the example and explains the specific role it plays in propeller and motor stand testing.
Distorted speed data usually comes from the stand environment itself: an open thrust stand in direct sun, indoor runs with LED fill light and mains-frequency lighting, the electromagnetic environment created by brushless motors and ESCs, and blade clamps and fixtures made of bright metal. A probe that skips, counts high or loses pulses under these conditions distorts the thrust-speed curve, and the test has to be repeated. This article treats measurement points, installation, the signal chain and on-site interference rejection as one subject.
Treating speed as one more number on the display wastes what it can do in a test. In the test chain for a UAV rotor system, speed carries at least four functions:
There is another requirement that is often underestimated: batch consistency. Fit the same batch of blades and the same batch of motors to the same airframe, and the spread in speed reflects blade geometry deviation and motor reluctance differences - a ready-made indicator for incoming inspection and line sampling.
On one propeller, the choice of measurement point decides which speed you read. A stand usually offers three positions, each with its own purpose:
Why a laser rather than an ordinary photoelectric probe: propeller speeds reach the ten-thousand-rpm range, and the stand has strong light from motors and ESCs plus reflections from drives and fixtures. The ROLS-5PW uses a 650nm frequency-modulated laser (Class 2, output below 2mW) with a built-in high-frequency filter and a large-diameter receiving lens, so a weak return is still received reliably in strong ambient light; the measuring range reaches 7 m, so the sensor can stand outside the safe distance and nobody needs to approach the measurement point while the propeller is turning.
The four kinds of interference on a propeller stand arrive together, and dealing with one of them alone is rarely enough:
A conventional photoelectric probe (including most laser-diode types) judges by return intensity: a reflection above the threshold counts as a pulse. That works in a laboratory and fails on a stand - sunlight is far brighter than the target return, the light ripple every 10 ms is enough to cross the threshold, and stray reflections from metal are counted too. The symptom is a reading that jumps, a pulse count that runs high or drops pulses, a curve drawn with sawtooth edges, and two runs of the same condition that do not agree.
The ROLS-5PW replaces that criterion: the laser is emitted at a fixed modulation frequency and the receiver demodulates that frequency alone. Sunlight, room lighting and indicator lamps on neighbouring machines do not carry that frequency signature and are rejected in the filter stage, leaving the signal that genuinely comes from its own target. That is why it still reads steadily on an open stand in outdoor sunlight.
| Comparison | Conventional photoelectric / intensity threshold | ROLS-5PW modulated laser |
|---|---|---|
| Identification basis | Whether the return exceeds an intensity threshold | Whether the return carries the modulation frequency the sensor emits |
| Direct sunlight | Sunlight is brighter than the target return, so the threshold cannot discriminate | Sunlight does not contain that modulation frequency and is rejected by the filter |
| Indoor flicker and fill light | Periodic brightness crosses the threshold, producing regular false triggers | Non-modulated frequencies never reach the demodulator |
| Nearby metal reflections | Stray reflections from bright surfaces can be counted as pulses | Accepts its own modulated light; stray reflections are not demodulated |
| Working distance | Typically of the order of 1 m | 7 m nominal (use within 7.6 m) |
| Weak return at high speed | The return is extremely short and low in energy, easily missed | A large-diameter collecting lens with high-frequency filtering still identifies it |
| Output form | Mostly a switching output, needing reshaping | 5 V TTL/PNP pulses, feeding data acquisition, panel meters and PLCs directly |
Two physical factors are often overlooked. One is return time at high speed: at ten-thousand-rpm the target crosses the beam in a few tens of microseconds, the return energy is low, and catching it takes a collecting lens with narrow-band demodulation. The other is aerodynamic fidelity from non-contact measurement - the probe does not touch the shaft and does not enter the propeller disc flow, so it adds no load to the object and does not change the flow near the blades, and the thrust and torque figures are the true behaviour of that propeller.
The 7 m range has extra value on a UAV stand: the sensor can be mounted outside the plane of rotation and behind a guard net, so nobody approaches the measurement point while the propeller is turning, and when the stand is modified or the propeller changed the probe position does not have to be reworked.
Start from low throttle and load up in steps, recording thrust and speed at each steady point, then plot the curve. The shape of the curve drives the selection strategy: a shallow climb means the blade is inefficient in that range; if thrust stops rising, or falls back, beyond a certain speed, the propeller has entered stall. The curve is also the basis for motor-propeller matching - the point where the motor characteristic crosses it is the sensible operating point.
Both the motor efficiency curve and the propeller efficiency curve need speed data before they can be drawn. The aim of optimisation is to find the efficiency peak within the usable speed range by superimposing motor efficiency on propeller efficiency, rather than looking at full throttle alone. Record thrust, torque and input voltage and current in step with speed, and the losses can be apportioned across motor, ESC and propeller.
Feed the speed pulse into the speed channel of a vibration analyser and the rotational components in the spectrum, including the blade passing frequency, are aligned to speed automatically. A high 1× component usually points to imbalance or misalignment; a peak near the blade passing frequency relates to blade count and stiffness; fractional orders suggest interference or resonance. If the speed reference is wrong, none of this reading is possible - for the reasoning behind harmonic interpretation, see stroboscopic observation of high-speed moving objects.
On a balancing stand the sensor supplies one pulse per revolution as the phase reference, and the machine uses it to calculate the phase angle of the unbalance and propose a correction. A stand measuring propellers also needs suitable fixturing and a steady speed segment; take the reference from the steady segment, not from acceleration or deceleration.
Once a blade has a crack, coating loss or edge damage, its mass distribution changes and the imbalance response at a given speed changes with it. On the line or in maintenance, combining a confirmed speed reference with stroboscopic visual inspection finds the abnormal part without dismantling it; for the stroboscopic side see predictive maintenance and stroboscopes and speed and stroboscopic inspection for industrial equipment.
A ground run with the propeller fitted has to confirm that the actual speed falls in the expected range. This is where a fixed-mount sensor shows its value: mounted once, nobody has to hold a handheld meter near the blades for the whole run, and the reading can go straight to a display or a data acquisition system for logging and over-limit alarms.
The ROLS-5PW outputs a 5V TTL PNP pulse; brown is +5V, blue is ground and black is the signal. That pulse chain can go in three directions as required:
Where blade phase, spacing and rotation need visual confirmation, a stroboscope works alongside the speed measurement: the LUYOR-906L laser-tracking stroboscope sets its own rate from the rotating pattern and makes the propeller appear "still", which makes tip path and any interference or rubbing easy to see.
What actually takes time on a test stand is running the whole envelope and then going back to analyse it, so the acquisition side has to record continuously rather than show an instantaneous reading. Route the ROLS-5PW pulse into host acquisition software (or upload through the RS485/Modbus RTU of the LUYOR-904 four-channel system or the LUYOR-910) and three things follow:
The output is a standard TTL pulse that mainstream acquisition cards, data loggers and PLCs accept directly; when integrating into an in-house stand, confirm the input type (PNP source into a sinking input) and the level match (5 V TTL into a 24 V digital module needs conversion, or choose the 24 V supply version). Wiring detail is in laser speed sensor wiring and outputs explained.
The same sensor is used at different stages by airframe manufacturers, eVTOL teams and university laboratories: manufacturers care about batch consistency and line cycle time, eVTOL teams about personnel safety distance and data traceability, laboratories about range coverage, channel count and budget. What they share is that nobody wants a person holding a meter next to a fast-turning propeller.
Organisations already using these sensors in volume for propeller and rotor speed measurement include the aerospace power laboratories of CH UAV (Aerospace CH UAV), XPENG AEROHT (eVTOL), Nanjing University of Aeronautics and Astronautics, Tsinghua University and Beihang University.
The names above indicate the actual application range of the product; project details remain subject to agreement between the parties, and permission should be confirmed before any long-term public reference.
The practice in these settings is very similar: the probe is fixed to the side of the stand or outside the guard net, aligned once and left alone for the whole test campaign; speed goes into the acquisition system and is logged synchronously with the thrust channel, and abnormal conditions are located by replay; changing a propeller or a motor means moving the target, not the sensor. That is the difference between a fixed probe and a handheld meter, and the precondition for a long test campaign to run continuously.
UAV thrust test stands, propeller dynamic balancing, motor-propeller matching and calibration, ESC speed-control performance testing, multirotor synchronous speed acquisition, wind-tunnel rotor tests, university aerospace power laboratories, and end-of-line quality checks on production powertrains.
| Stand requirement | Recommended configuration | Key points |
|---|---|---|
| Speed display with over-limit alarm | ROLS-5PW + LUYOR-910 panel tachometer | The panel meter can supply 5 V to the sensor; its two relays drive an audible/visual alarm or an interlock stop |
| Four-rotor or four-station comparison | ROLS-5PW ×4 + LUYOR-904 four-channel acquisition system | One time base, four colour curves on one screen, pulses per revolution and alarm limits set per channel |
| Integration into an in-house stand or PLC | The TTL/PNP pulse from the ROLS-5PW connects directly | PNP source output into a sinking input; 5 V into a 24 V digital module needs conversion or the 24 V supply version |
| Long-term outdoor installation, rainy season | ROLS-5PW + waterproof junction box and sunshade | The body is IP60, dust-tight but not water-resistant; probe and terminals need secondary protection |
| Temporary points, no fixed supply on site | SPLS-5PW self-powered module | Built-in battery about 60 hours, BNC output, integral display |
For stand testing, always non-contact. Contact types (magnetic, Hall) need the probe mounted at the shaft end or coupling, where the contact face and mounting concentricity affect the reading directly, and fitting and removal are awkward; laser and photoelectric probes work from a reflective target, go on and come off quickly, can be moved, need no realignment when the measurement point changes, and measure at a distance on fast-turning parts that a person cannot approach.
In most cases it is not. Photoelectric (LED) types usually work within about 1 m and within 45° off axis, and outdoors they need additional filtering or modulation; at high speed the return from the target is also weaker. A laser type has a fine beam, longer range and wider off-axis angle, and with high-frequency filtering it resists ambient light, which makes it the usual choice for high-speed propeller work. For a selection comparison see laser speed sensor selection guide.
It is. Uneven motor speeds bring attitude and efficiency problems, and measuring one machine at a time cannot compare sequence or amplitude within one run. Four-channel synchronous acquisition gives four curves and their differences at once, and with independent alarms per channel it points straight to the channel at fault; wiring and software are covered in LUYOR-904 four-channel speed acquisition system.
Yes, but the pulses per revolution must be set to match. With 4 targets each revolution gives 4 pulses, and if the back end still scales as 1 pulse/rev the reading is multiplied by 4. Where high resolution is needed, prefer a measurement point that already gives multiple pulses per revolution (a toothed disc or an encoder wheel) and keep reflective targets for temporary work.
On the shaft - motor output shaft, coupling or the root of the hub. On the blade surface the blade blocks the beam on every pass, pulses are lost and the reading jumps, and order analysis is especially sensitive to this. To measure the true propeller speed, add a second probe on the hub and compare it with the shaft speed.
The ROLS-5PW is rated IP60, dust-tight but not water-resistant, and the signal cable and terminals must also be kept out of the rain. On an outdoor stand, add a waterproof junction box and a sunshade, and in rain move the sensor and cabling under a roof before continuing; for a permanent outdoor installation, arrange the probe and cabling under a roof or give the connections secondary protection so that the terminals do not stay damp.
It can, provided a modulated laser type is chosen rather than an intensity-threshold type. Sunlight is far brighter than the target return, so a probe that judges by intensity largely fails in direct sun; the ROLS-5PW demodulates its own modulation frequency and sunlight, lacking that frequency, is filtered out. Two things still matter in use: keep the target clean, since oil and dust weaken the return, and fit a sunshade on the probe to extend its life - the body is IP60, dust-tight but not water-resistant, so in rain it has to move under a roof.
Three routes: the pulse goes into a high-speed counter channel or a data logger and the acquisition software saves it with timestamps; it is recorded through the LUYOR-904 four-channel system with four curves on one screen and export to CSV/Excel and statistical reports; or it is uploaded through the RS485 (Modbus RTU) of the LUYOR-910 panel meter. All three put speed on the same time base as thrust and torque for post-processing, which matters especially for throttle steps and load changes - otherwise the curves can be aligned by hand annotation alone.
The accuracy floor of a UAV propeller test depends on whether the speed reference is trustworthy. A fixed-mount laser speed sensor such as the ROLS-5PW settles three things at once: the 7 m range keeps people at a safe distance from a fast-turning propeller, the pulse output feeds panel meters, four-channel acquisition and vibration analysers directly, and the 1-300 000 rpm range covers the speeds of everything from agricultural blades to racing quadcopter propellers. If you need measurement-point planning and wiring confirmation for a specific stand, contact the Xinyisheng technical team for a proposal.
This article belongs to the speed and stroboscopic measurement section.
Other sections on the same product line: laser tachometers (aim at a reflective target and read RPM, one-handed), speed sensors (wired to a PLC or acquisition card for continuous monitoring, delivering an electrical signal) and stroboscopes (freezing rotation and reciprocation by persistence of vision, to see the process rather than a number).
Models mentioned in this article: LUYOR-910, LUYOR-ROLS-5PW, LUYOR-906L. Range, accuracy and accessory lists for each model are given 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
