How this instrument works
Divergence half-angle is how fast a laser beam's radius grows once it moves past its narrowest point, the waist, measured as an angle from the beam axis. The formula θ = λ ⁄ (πw₀) is not an empirical fit; it falls straight out of Fourier optics. A beam's angular spread in the far field is the Fourier transform of its spatial profile at the waist, and for a Gaussian profile that transform is itself Gaussian, with a width set by wavelength divided by waist size. Narrow the aperture and the transform necessarily broadens.
That inverse relationship between w₀ and θ is a space-bandwidth trade, the same mathematics that produces Heisenberg's position-momentum uncertainty relation, here applied to a light wave instead of a particle. A beam focused to a smaller waist carries a wider spread of transverse wavevectors, so it must fan out faster once released. There is no clever coating or exotic glass that gets around this; it is a property of waves, not of workmanship, which is why the same formula appears whether the source is a laser diode, a telecom fiber output, or a radio telescope feed.
The result assumes an ideal TEM00 Gaussian beam with a beam-quality factor M² of 1 — the best any real laser can do, never the average. Multimode diode bars and high-power fiber lasers typically run M² of 1.1 to well over 20, and their true divergence is θ multiplied by that factor. The formula also relies on the paraxial approximation, which holds as long as the waist is several wavelengths across; push w₀ down toward λ itself, as with some nanophotonic emitters, and the small-angle geometry this equation depends on breaks down.
- Enter the source wavelength in the Wavelength field; nm is the default and covers most visible and near-IR diodes, with µm available for mid-infrared sources like CO₂ lasers.
- Enter the Beam waist radius — the 1/e² intensity radius at the beam's narrowest point, not the full spot diameter and not the FWHM. Halve a measured diameter before typing it in.
- Read Divergence half-angle, mrad directly; it needs no unit conversion since the output field is fixed in milliradians.
- To get the full cone angle instead of the half-angle, double the reading.
- To estimate spot radius at a distance d well past the waist, multiply the half-angle in radians (divide the mrad figure by 1000) by d and add it to w₀.
Worked example — a 650 nm pointer with a 0.5 mm waist
A red diode laser is specified at 650 nm and, after its collimating lens, leaves a beam waist radius of 0.5 mm. Converting to metres, λ = 6.5 × 10⁻⁷ m and w₀ = 0.0005 m. The formula gives θ = 1000 × (6.5 × 10⁻⁷) ⁄ (π × 0.0005) = 0.413802852039, which the field displays as roughly 0.4138 mrad — about 0.0237°.
That figure is the diffraction floor, not a defect: a perfect collimating lens on a perfect cavity still cannot beat it, because it is set by the waist and wavelength alone. At 10 metres from the waist, the beam has spread by roughly (0.413802852039 ⁄ 1000) × 10 ≈ 4.14 mm of extra radius on top of the 0.5 mm it started with — the arithmetic behind why laser-pointer specification sheets bother quoting a divergence figure at all.
Questions
Why does focusing a beam to a smaller waist make it diverge faster?
Because waist radius and divergence angle are locked together by diffraction: their product is fixed at λ ⁄ π for a given wavelength. Squeeze the waist down and the far-field angle must grow to compensate — a wave-optics trade-off, not a flaw in the optics. It is the same space-bandwidth relationship that underlies Heisenberg's uncertainty principle, applied here to a light wave rather than a particle.
Is this the half-angle or the full divergence angle?
Half-angle, measured from the beam's central axis to one edge. The full cone angle a beam actually opens into is twice this figure — 2θ. Datasheets are inconsistent about which one they quote, so always check whether a spec sheet's divergence number is full-angle before comparing it against this field's output.
Does this formula predict the true divergence of any laser I own?
Only for an ideal, diffraction-limited TEM00 beam with beam-quality factor M² = 1. Real lasers, especially multimode diode bars and high-power fiber lasers, diverge more than this by a factor of M², so a datasheet's actual divergence is θ × M². Treat this calculator's output as the physical best case any beam of that wavelength and waist could ever achieve.
What exactly counts as the beam waist radius?
The 1/e² intensity radius at the beam's narrowest point along its path — the distance from the axis where irradiance drops to 1/e² (about 13.5%) of its peak value. It is a radius, not a diameter, and not the same as the FWHM spot size; entering a full beam width instead of this radius will roughly double the calculated divergence.
Why does a longer wavelength diverge more for the same waist?
Because θ scales directly with λ. A 1,550 nm telecom laser at the same 0.5 mm waist used in the worked example diverges at 0.98676064717 mrad — more than double the 650 nm figure — for the identical waist radius. It is the same reason radio-telescope dishes must dwarf optical telescopes to match their angular resolution: longer waves diffract more for a given aperture.
Can I use this for a fiber-optic output instead of a free-space laser?
Yes, as long as you supply the fiber's actual mode-field radius as the waist, not its core radius; the two differ because light leaks slightly into the cladding. Single-mode fiber datasheets usually list mode-field diameter directly — halve it before entering the value here, the same rule that applies to any measured beam diameter.