A downward-facing laser rangefinder for altitude hold must stay valid through every flight phase: take-off, hover, cruise climb, descent and landing. The three modules compared here—LRF22VB, LRF50H and LRF50VB—share a 0.05 m minimum measuring distance but differ in maximum range, update rate, output unit and form factor. Those four differences determine which fits your design.
Minimum distance on landing
All three modules publish a minimum measuring distance of 0.05 m (5 cm). This is a detection-circuit property: any target closer than 0.05 m produces the same reserved out-of-range sentinel the module uses when no echo returns at all. At 3 cm above the ground, the host cannot distinguish “target too close” from “no ground detected.”
For a landing sequence, the control logic must handle the reserved sentinel as a valid operating condition and not as a fault. A graceful descent strategy falls back to inertial or time-based altitude management before the sensor floor is reached; rangefinder altitude data cannot be trusted for the final 5 cm.
The 0.05 m floor does not change with baud rate, output frequency setting or supply voltage. It is not a software threshold and cannot be adjusted through the protocol command set.
Update rate for altitude hold
The three modules cover very different update-rate envelopes:
| Model | Update rate | Supply voltage | Weight |
|---|---|---|---|
| LRF22VB | 200 Hz max; selectable 200/100/50/20/10/1 Hz | 3.0–3.6 V DC | 1.5 g |
| LRF50VB | 100 Hz | 3.3–5 V DC | 1.3 g |
| LRF50H | 20 Hz–10 kHz; default 1 kHz | 9–36 V DC | 15 ± 2 g |
Select a rate that meets or slightly exceeds your altitude-hold control loop period. A loop running at 50 Hz needs at least 50 valid readings per second; at 200 Hz, the LRF22VB is the only module here with a matching output rate. The LRF50VB’s 100 Hz suits loops up to that frequency. The LRF50H’s default 1 kHz is well above any practical altitude-hold loop rate.
Setting the rate higher than the host consumes does not increase stability. A UART receive buffer that fills faster than it is drained introduces irregular read latency—worse for a PID loop than a lower, steady cadence. Measure the actual valid-reading cadence at your chosen rate under operating conditions before tuning the altitude-hold loop.
The frequency divisor formula differs between families. For LRF22VB: f = 10000 / (divisor + 1). For LRF50H: f = 1000000 / (divisor + 1). Confirm the formula from the model-specific protocol section before writing configuration commands.
Units and out-of-range codes
Each module outputs distance in a specific unit and signals an invalid measurement with a reserved sentinel. Parsing errors from unit confusion cause the host to interpret the sentinel as a large but plausible altitude reading.
| Model | Distance unit | Protocol range | Out-of-range sentinel |
|---|---|---|---|
| LRF22VB | mm | 0–20000 | 20000 |
| LRF50VB | mm | 0–50000 | 50000 |
| LRF50H | cm | 0–65535 | 65535 |
Three points follow directly from this table:
- Different units across families. LRF22VB and LRF50VB output millimetres; LRF50H outputs centimetres. A parser written for a VB-family module and reused on the LRF50H reads every centimetre value as millimetres—reporting a 5 m reading as 500 m—and interprets the out-of-range sentinel 65535 as 655.35 m rather than discarding it.
- The sentinel is not always outside the numeric range. For LRF22VB, 20000 mm equals exactly 20 m, the stated maximum. For LRF50VB, 50000 mm equals exactly 50 m. Only LRF50H uses 65535, which exceeds its documented maximum of 50 m (5000 cm). The validity check must compare the raw decoded integer against the documented sentinel before scaling, not after.
- Byte order matters first. All three use 2-byte little-endian distance fields. Decode byte order before the sentinel check; a big-endian read of a small distance value produces a large nonsense number that may pass a naive sentinel check.
Mounting and vibration
LRF22VB and LRF50VB are compact PCB-level boards. The LRF22VB measures 21 × 15 × 7.43 mm and weighs 1.5 g; the LRF50VB measures 18.7 × 11.8 × 13.3 mm and weighs 1.3 g. Neither has a cased mounting surface; both require a carrier bracket and isolation from motor vibration before the beam axis is considered aligned.
LRF50H is a housed module: 38 × 20 × 30 mm, 15 ± 2 g, with four Ø2.3 mm through holes for M2 screws on a 33 × 12 mm pattern. The housing provides a stable optical axis and simplifies alignment, but adds 10–13 g compared with either VB unit. On a multi-rotor with a tight payload budget, that difference is load-significant.
Field of view affects which ground surface dominates the return at altitude:
- LRF22VB: 3° — beam footprint approximately 26 cm diameter at 5 m altitude
- LRF50VB: 1.9° — beam footprint approximately 17 cm diameter at 5 m altitude
- LRF50H: ≈4 mrad — beam footprint approximately 2 cm diameter at 5 m altitude
A wider field of view integrates over a larger ground area, which averages across small surface features and can improve reading consistency over rough terrain. A narrower field of view is more sensitive to beam alignment and to surface variation directly below the sensor. Neither is universally better; the choice depends on the terrain type and the accuracy the control loop requires.
Vibration from rotors can produce occasional erroneous distance values. At 200 Hz (LRF22VB), a three-reading sliding-window median adds approximately 15 ms of latency and rejects single-frame outliers before the value enters the altitude-hold loop. At 100 Hz (LRF50VB), the same filter adds approximately 30 ms. Use soft-mount isolation on the carrier bracket for all three models and verify the filtered valid-reading cadence with the drone at hover before closing the altitude PID.
Altitude-hold selection checklist
Use these seven gates before committing a module to an altitude-hold design. Each is independent; a fail on any one is a stop.
- Maximum operating altitude: LRF22VB covers 0.05–20 m at 90% reflectivity. LRF50VB and LRF50H cover 0.05–50 m at 90% reflectivity. If your cruise altitude exceeds 20 m over typical terrain, LRF22VB does not qualify.
- Update rate vs control loop: Confirm the module’s output rate meets or exceeds your loop frequency. LRF22VB (200 Hz max) suits loops up to 200 Hz; LRF50VB (100 Hz) suits loops up to 100 Hz; LRF50H (1 kHz default) suits all practical altitude-hold rates.
- Supply rail: LRF22VB requires 3.0–3.6 V DC. LRF50VB accepts 3.3–5 V DC. LRF50H requires 9–36 V DC. Confirm the available regulated rail before ordering.
- Mass budget: LRF22VB: 1.5 g. LRF50VB: 1.3 g. LRF50H: 15 ± 2 g. For multi-rotor UAVs, include connector and cable mass in the estimate.
- Unit and sentinel: Maintain separate parsers for cm output (LRF50H, sentinel 65535) and mm output (LRF22VB, sentinel 20000; LRF50VB, sentinel 50000). Validate the sentinel before unit conversion, not after.
- Mounting: LRF50H has four M2 holes on a 33 × 12 mm pattern. LRF22VB and LRF50VB require a custom carrier bracket. Soft-mount all three to attenuate rotor vibration at the sensor.
- Ambient light: LRF50H is documented at 45 m ambient-light immunity; LRF50VB at 10 m; LRF22VB at 8 m. Bright outdoor sunlight degrades valid-reading rate most on the VB-series units; evaluate over the intended terrain in daylight conditions.
At first power-on, place the drone on the ground and confirm the host reads the reserved sentinel (not zero or a small positive value). Then lift to 1 m and verify a valid reading clears the sentinel. Log raw frames at this stage before closing the altitude-hold loop.
Contact ERDILRF with your operating altitude range, loop frequency and available supply rail if you need a configuration recommendation for a specific flight profile.
FAQ
What blind zone is acceptable for landing?
The LRF22VB, LRF50VB and LRF50H all document a 0.05 m minimum measuring distance. Below that threshold the module outputs the same reserved sentinel it uses when no echo returns, so the host cannot distinguish 3 cm above ground from no ground at all. Design the landing sequence to switch away from rangefinder altitude data before reaching the 5 cm floor—inertial or time-based fallback is the standard approach. Confirm the host handles the sentinel without stalling the PID loop, and test the transition before first flight.
Millimetres or centimetres: does it matter?
Yes, at the parser level. LRF22VB and LRF50VB output distance in millimetres; LRF50H outputs in centimetres. Reusing a VB-family parser on the LRF50H causes every centimetre reading to be treated as millimetres—5 m becomes 500 m—and passes the out-of-range sentinel 65535 as a valid large value instead of discarding it. Confirm the unit and sentinel against the specific model’s protocol section before writing any decode logic, and do not share a single decode function between families.
How fast should the rangefinder update?
Match the update rate to your altitude-hold loop frequency. A loop at 50 Hz needs at least 50 valid readings per second; at 200 Hz, only the LRF22VB (200 Hz max selectable) supplies that cadence directly from this group. The LRF50VB provides 100 Hz; the LRF50H defaults to 1 kHz with a selectable range of 20 Hz–10 kHz. Setting the rate higher than the host reads does not improve stability and can introduce read-latency jitter from UART buffer overflow. Measure the actual valid-reading cadence under hover conditions before tuning the altitude-hold PID.
Category pages: Laser rangefinder module
Models mentioned in this article
- LRF22VB905 nm · up to 20 m · UART · I2C · 1.5 g1 pc $12 · shipping includedSee price & buy
- LRF50H905 nm · up to 50 m · UART · 15 ± 2 g1 pc $100 · shipping includedSee price & buy
- LRF50VB905 nm · up to 50 m · UART · I2C · 1.3 g1 pc $70 · shipping includedSee price & buy