How to choose a 905 nm pulsed laser diode: from range target to part number

Start from the range, not from the peak power

The first question in most projects is "how many watts do I need?". It is the wrong first question, and it is why so many designs end up over-specified on peak power and under-specified on everything else.

Range is set by pulse energy on the transmit side and by receiver sensitivity on the receive side. Peak power only matters because of how it combines with pulse width to produce that energy:

A 180 W diode driven for 100 ns and a 90 W diode driven for 200 ns put the same energy into the air. They are not interchangeable, because the pulse width also sets your resolution and your duty cycle - but for reach alone, they are the same part. The derivation is in the range equation article; what follows is the procedure.

Step 1 - convert the requirement into pulse energy

Using the model published in the range equation article - 36 mm receive aperture, 90 % optical efficiency, matched-filter receiver, silicon APD - the energy you need is:

Read that table twice, because it contains the two facts that decide most 905 nm projects. First, range goes as the square root of energy: doubling the reach costs four times the energy. Second, a diffuse target costs roughly a factor of nine in energy compared with a retro-reflector, because about 10 % of the light comes back instead of about 90 %.

Step 2 - choose the pulse width

Two constraints pull in opposite directions.

Resolution. Light travels 0.3 m per nanosecond, so a single-pulse threshold receiver separates two targets about 15 cm apart per nanosecond of pulse width. A 100 ns pulse smears every return over roughly 15 m. If you need sub-metre resolution, you need a short pulse, and you need a receiver fast enough to preserve it.

Energy per pulse, and driver cost. Short pulses are harder to drive and harder to measure. Our LZ-905-PD01 driver has a 7.5 ns typical rise time, which is 15 % of a 50 ns pulse and 7.5 % of a 100 ns pulse - see the driver design article for what that does to the energy you actually get.

So: pick the longest pulse that still meets your resolution and blind-zone requirements, and take the energy you need from peak power rather than from pulse width. The whole 905 nm family is rated for pulses up to 100 ns, which is a deliberate design point - beyond that, duty cycle and thermal limits start to dominate.

Step 3 - convert energy back into peak power

Divide the energy you need by the pulse width you chose:

This is where the marketing number gets honest. A 180 W diode is a 1.8 µJ part at 10 ns and an 18 µJ part at 100 ns, and the second number is the one that determines your range.

Step 4 - check the duty cycle before you check the package

Pulsed laser diodes fail thermally. The limit is the duty cycle - the fraction of time the junction is actually conducting - and for the whole 905 nm family it is 0.1 %.

The arithmetic is unforgiving and it is the most common way a design fails late: a project that needs 30 kHz has a hard 33 ns pulse ceiling on this family, no matter how much peak power it buys. If the application needs 100 ns and 30 kHz, the answer is more emitters in parallel or a different laser technology - not a bigger diode.

Step 5 - the driver is a selection constraint, not an afterthought

Two parts in the same power class can need very different drivers. From our own datasheets, at 100 ns:

A 55 W part draws less than half the current of a 75 W part, because the stacks are in series rather than in parallel. If your driver tops out at 20 A, the 75 W device is simply not available to you, and the 55 W device may reach further. Voltage headroom matters too: the low-drift PLD905AL-70W-M wants 23 V, which a 12 V-only driver cannot supply.

Step 6 - package, collimation and drift

Once energy, width, repetition rate and drive current are fixed, the remaining choices are mechanical:

• Package. TO56 covers 15 W to 180 W, plastic SMD covers 25 W to 75 W for consumer volumes, and the 300 W and 600 W devices move to larger headers. See the pulsed laser diode family.
• Collimation. The fast axis of an edge emitter diverges at 25°, which at 1 km is a stripe hundreds of metres tall. A fast-axis collimator takes it to 6-8° and puts the light where your receiver can see it - FAC and fibre coupling.
• Temperature drift. The standard parts shift 0.10 nm/K. The low-drift PLD905AL and PLD905BL parts are for narrow-filter receivers that cannot afford that shift over -40 to +85 °C.

A worked example: 2 km on a road sign

1. Retro-reflective target, 2 km: the table gives 28 µJ.
2. Pulse width: the sign is large and the resolution requirement is coarse, so take the family maximum, 100 ns.
3. Peak power: 28 µJ / 100 ns = 280 W.
4. Duty cycle: at 100 ns the maximum repetition rate is 10 kHz. If the design needs more, shorten the pulse and recalculate - at 33 ns you would need 850 W, which does not exist in this family, so the answer is to accept 10 kHz.
5. The catalogue part is PLD905-300W-M at 300 W / 30 µJ, giving about 2.1 km on a retro-reflector.

Note what happened: the request was "2 km", and the answer is a 300 W part, not the 180 W part everyone reaches for. The 180 W device reaches about 1.6 km on a retro-reflector and 535 m on diffuse terrain. Both numbers are published next to the part on this site, with the assumptions printed alongside, for exactly this reason. Every one of those four roles - emitter, optics, detector and driver board - carries its own budget and its own part number; the 905 nm rangefinder teardown unpacks them one at a time.

What would change the answer

• A larger receive aperture. Range goes as the square root of aperture area, so doubling the diameter buys 1.41× the range at no laser cost. It is usually the cheapest range in the system.
• A better receiver. See APD versus PIN.
• Multi-pulse averaging. Averaging N returns improves the signal-to-noise ratio as the square root of N, which buys range at the cost of frame rate.
• Eye safety. Everything above is an optical calculation. Whether you are allowed to emit 30 µJ at 905 nm into the world is a regulatory question - see laser safety classes for ranging - and it is the constraint that most often overrides the energy budget.