Access Official Technical PDF

Share Summary:LinkedInEmail

SiPM direct ToF LiDAR combines a pulsed laser, a silicon photomultiplier (SiPM) and timing electronics to estimate the distance to a target. Its appeal includes single-photon sensitivity, fast response and the potential for compact, low-voltage systems. But using an SiPM also makes optical design and ambient-light management central to system performance. Engineers need to consider the complete signal path, from emitted pulse to extracted range, rather than treating the detector as an isolated component.

How direct time-of-flight ranging works

A direct time-of-flight (ToF) system sends a laser pulse toward a target. Some of the light is reflected or diffused back and detected by the receiver. Timing electronics record when the returned signal arrives. The target distance is calculated from the round-trip travel time: D = c × Δt / 2, where c is the speed of light and Δt is the measured time of flight.

The receiver also detects photons that are not part of the laser return, including ambient light. The system must distinguish the time-correlated return from this background. A single laser pulse produces a single-shot measurement, but the detected return may be weak. Combining timestamps from many pulses into a frame can reveal the return more clearly and support a more precise range estimate.

Choosing a readout approach

The readout method determines how timing information is captured and how the system handles signal strength, background light and processing demands. Common approaches include:

  • Leading-edge discrimination (LED): Detects the rising edge of a multi-photon signal. Timing depends on identifying that edge; the method is not affected by laser pulse width in the same way as full-pulse timing.
  • Full-waveform digitization: Records the signal waveform, which can be oversampled to improve timing accuracy. Implementation can be difficult with very short pulses or high repetition rates.
  • Time-correlated single-photon counting (TCSPC): Records individual photon arrival times and can provide strong ambient-light rejection. It requires fewer than one detected signal photon per laser pulse, along with short pulses, a high repetition rate and fast timing electronics for fast measurements.
  • Single-photon synchronous detection (SPSD): A form of TCSPC that provides high ambient-light rejection, while requiring electronics designed to manage range ambiguity.

Why the ranging histogram matters

Timestamp data collected over a frame can be plotted as a histogram. A returned laser pulse appears as a peak at the corresponding ToF; uncorrelated background detections contribute noise across the measurement window. One useful measure is histogram signal-to-noise ratio (SNRH), defined as the signal peak divided by the maximum noise peak.

Laser repetition rate affects how many single-shot measurements can be collected during a fixed acquisition period. A higher rate can increase the number of detected returns and improve histogram quality, although noise counts also rise. Repetition rate must also be chosen with the unambiguous range in mind: pulses sent too close together can make returns from distant targets difficult to associate with the correct transmission. The frame acquisition time sets how long the system gathers data for each range measurement, creating a trade-off between data collection and frame rate.

Optical design is essential for ambient-light control

Collection aperture, sensor field of view and optical filtering determine how much light reaches the SiPM. Increasing the collection aperture can admit more ambient photons. If the detector approaches saturation, it may become less able to register the laser return, reducing the useful signal. A wider sensor angle of view similarly exposes the detector to more background light. The field of view should be limited to cover the laser’s target area rather than unnecessary surroundings.

A bandpass filter centered around the laser wavelength can reject light outside the range of interest. A narrower passband can reduce background, but the laser’s wavelength variation must be considered when selecting the filter. The design goal is not simply to maximize collected light; it is to collect enough return signal while keeping background at a level the receiver can handle.

Balance laser, sensor and operating conditions

Laser pulse width influences the width of the return peak. For leading-edge detection, the first arriving photons establish the timing edge, so a shorter pulse can produce a narrower timing feature. In practice, available laser sources and other system requirements influence the choice.

Wavelength selection involves more than matching the laser to the detector. Eye safety, laser availability, solar irradiance and the SiPM’s photon detection efficiency (PDE) all matter. A wavelength with lower ambient solar irradiance may reduce background, but a lower PDE can also reduce detection of both the laser return and ambient photons. The net effect depends on the sensor and the full system configuration.

Range and illumination also shape performance. As target distance increases, the returned laser photon density at the sensor falls, while ambient background may remain comparatively significant. Higher ambient light can increase noise counts and reduce SNRH. Longer-range operation may therefore require coordinated changes to laser power, optics, filtering, detector field of view and acquisition settings—not an adjustment to one parameter in isolation.

Use system modelling to assess trade-offs

A direct ToF model can help evaluate a proposed setup before hardware is finalized. An analytical stage estimates laser and ambient light levels at the sensor and checks them against saturation limits. A Monte Carlo stage can represent stochastic effects such as PDE and timing jitter, while accounting for pulse width, repetition rate and acquisition time. The simulated timestamps can then pass through a readout model, such as a discriminator and time-to-digital converter (TDC), to produce a histogram for range extraction.

In the reported demonstrator work, measurements from a Gen1 system were compared with model results to validate the model for exploring other configurations. A later Gen2 configuration was simulated for ranging to 100 m; those results should be understood as modelled performance, distinct from the Gen1 system’s measured results at shorter distances.

Design for the complete measurement chain

For industrial and embedded LiDAR teams, the practical lesson is to optimize the full chain: pulse generation, optical transmission, target return, collection optics, SiPM response, signal discrimination, timestamping and histogram processing. Use the histogram to assess whether the return can be separated from background, and evaluate repetition rate alongside unambiguous range and frame rate. Careful control of aperture, field of view and filter bandwidth can help protect the receiver from ambient-light saturation.

SiPMs can support compact direct ToF ranging, but successful integration depends on matching the sensor to the optics, timing architecture and operating environment. Modelling these interactions early—and validating the resulting design against measurements—helps teams make informed trade-offs in range, signal quality and acquisition speed.

Access Official Technical PDF

About The Author