The laser range equation: how much pulse energy does a ranging link actually need?

The question engineers actually ask

Almost every laser ranging project starts with the same sentence: "I need to measure 2 km." It is the wrong first sentence, because range is not a property of a laser diode. Range is a property of a whole link: transmitter energy, transmit optics, target reflectivity, atmospheric transmission, receive aperture, optical filtering, detector gain and the detection threshold you are willing to run.

That said, you have to start somewhere, and the useful starting quantity is pulse energy, not peak power. This article builds the budget in energy, then converts back to the peak power a datasheet quotes.

The range equation for an extended target

For a target that overfills the laser spot - a wall, a hillside, a vehicle, a tree - the return follows the Lambertian form of the radar range equation:

where

Solve for R and substitute P_rx = P_mds:

Because P_tx = E / t_p, and because a matched-filter receiver's minimum detectable energy is roughly constant as long as the electrical bandwidth tracks the pulse width, the P_tx and the t_p largely cancel. What survives is the pulse energy:

This is the single most useful fact in laser ranging: shortening the pulse improves resolution, not reach. If you want more distance, buy energy. If you want better precision, buy a shorter pulse at the same energy - which means buying more peak power.

Calibrating k against real products

Theory gives you the shape of the curve but not the constant, because P_mds depends on a detector you have not chosen yet. Calibration against products with known performance is faster and more honest.

A 905 nm pulsed laser diode rated 180 W and driven at 100 ns delivers:

That part sits in handheld rangefinders that are specified at roughly 1.5 - 2 km on reflective targets and 500 m on dark terrain. Taking 1.6 km on a retro-reflector as the anchor:

The same anchor on diffuse terrain (roughly one third of the reflective range, a rule of thumb that matches almost every commercial LRF) gives k_diffuse ≈ 126.

For a 1535 nm eye-safe link the receive chain is an InGaAs APD with a higher noise floor, but the transmit energies are one to two orders of magnitude larger and the optics are bigger. Empirically, k_retro ≈ 600 and k_diffuse ≈ 200 fit published erbium glass rangefinder specifications.

The design table

Pulse energy needed for a given range, at 905 nm with a Si APD receiver:

Read the diffuse column and the conclusion is unavoidable: a 905 nm pulsed diode cannot reach 2 km on dark terrain at a 100 ns pulse. 250 µJ at 100 ns is 2.5 kW of peak power, which is not a TO56 device. Your options are:

• Move to 1535 nm, where 100 - 300 µJ modules are standard parts.
• Accept a longer pulse and more averaging, trading precision for reach.
• Improve the receive side: bigger aperture, narrower filter, lower-noise detector.

That last option is where most of the practical gains live, and it is covered in the receiver selection article.

Turning energy back into a part number

Once you know the energy you need, the datasheet conversion is:

So a requirement for 10 µJ is satisfied by any of:

• 100 W for 100 ns
• 200 W for 50 ns
• 1000 W for 10 ns

These are not equivalent systems. The 1000 W / 10 ns variant gives you 10× better range resolution and a much shorter blind zone, at the cost of a driver that has to switch 1000 W cleanly in 10 ns, tighter timing electronics, and a receiver with ~100 MHz of bandwidth. That trade - resolution against driver difficulty - is the real decision, and it is why the driver design article matters as much as the diode choice.

What the equation deliberately leaves out

The model above assumes single-pulse threshold detection, clear air, a target larger than the spot, and a well-aligned optical axis. Real designs lose range to:

• Atmospheric attenuation. Rain, fog and haze cost 3 - 30 dB depending on conditions. 1535 nm and 1550 nm sit in a water absorption band; that is a disadvantage in rain and an advantage for eye safety.
• Speckle. A rough target returns a speckle pattern, so the received energy fluctuates pulse to pulse. Budget 3 - 6 dB for a reliable detection probability.
• Beam walk and alignment. Small angular errors at long range move the spot off the receiver's field of view.
• Detector bandwidth. A receiver wider than the pulse integrates more noise than signal. Match the bandwidth to the pulse: BW ≈ 0.4 / t_p.
• Threshold choice. A fixed threshold that works on a retro-reflector is deaf on a tree. Constant-fraction or time-variable thresholds recover several dB.

Working numbers, applied

Take the PLD905-180W-M: 180 W peak, 1 - 100 ns, 6 emitters, 18 µJ at 100 ns.

For anything beyond that on real terrain, the honest answer is a longer wavelength. See 905 nm vs 1535 nm vs 1550 nm for that decision.