Pinhole spatial filter
A lens brings light travelling along its axis to one point and light at any other angle to another; a pinhole at the first point passes the beam and stops the rest.
Open in the canvas →Background
A laser beam that has passed a few optics is rarely clean. Dust, scratches and stray reflections add light that travels in slightly different directions from the beam itself, and it shows as rings and speckle on the beam's profile. A spatial filter removes it, and is commonly used to clean up the output of lasers[1].
The filter rests on what a lens does to direction. A lens brings all rays parallel to its axis to one point, the focal point, and all rays parallel to each other at an angle θ to the axis to another point, a distance f tan θ to the side. In the focal plane, position stands for direction. In wave terms the lens forms there the Fourier transform of the beam: the centre of the pattern corresponds to a perfect, wide plane wave, and light further from the centre to finer structure across the beam[1].
A small aperture in that plane, the pinhole, passes only the central spot[1]. A second lens, one focal length behind the pinhole, makes the light parallel again, so the filter is also a beam expander, wider by the ratio of the two focal lengths. The size of the pinhole is a compromise, chosen from the focal length of the lens and the wavelength: too small a hole improves the beam further but costs power[1]. The beam that leaves has a smooth profile, which can be almost a perfect Gaussian[1].
What this setup demonstrates
A 633 nm laser sends a 4 mm beam of 10 mW along the axis of two lenses, f = 50 mm and f = 100 mm, placed 150 mm apart. A second laser of 1 mW, at 6° to the axis and aimed at the centre of the first lens, stands for light that travels in another direction. In the focal plane of the first lens stands the pinhole: a Slit with a gap of 0.5 mm.
The beam along the axis comes to a focus in the gap and goes through. The light at 6° comes to a focus 50 mm × tan 6° = 5.3 mm to the side, on the plate, and stops there; the enlarged view shows both. The second lens makes the beam parallel again, now 8 mm wide. The power meter reads 10.0 mW: all of the beam, none of the other light.
Things to try. Open the gap to 11 mm or more: the light at 6° gets through as well and the meter reads 11.0 mW. Set the gap back to 0.5 mm and move the pinhole 0.3 mm to the side: the meter goes dark, because the focus is now on the plate. Put the pinhole back on the axis and move it along the axis instead: up to 6 mm from the focus nothing changes; at 10 mm the beam is 0.8 mm wide where it meets the 0.5 mm gap and the meter reads 6.00 mW. A spatial filter is aligned by exactly these three movements.
This is a top view, in two dimensions: the pinhole is a slit, a gap in a line, and it filters directions within the plane of the drawing only.
The setup shows the geometry of a spatial filter, not its effect on a beam profile. A traced beam is a bundle of rays that meet in a point with no size, so any gap passes it completely, and the 0.5 mm gap, the smallest the Slit offers, is far wider than a real pinhole. A real focus has a size set by diffraction, and the pinhole is chosen a little larger than it. Light scattered at small angles lands within that spot's surroundings, close to the axis; here the unwanted light is one beam at one large angle, chosen so that the drawing shows it. The smooth, nearly Gaussian profile a real filter produces, and the power it costs, are not computed.