OpenWorldLab

Loading interactive view…

Optics

A laser lab on an optical table

Lasers, mirrors, lenses, splitters, slits and detectors on a table you can rearrange. The light is computed as real beams with a width, a phase and a polarisation, so interferometers make fringes, slits diffract, and a pair of mirrors around a gain medium lases only if it would in a real lab.

What you are looking at

A top-down view of an optical table, with holes every 25 mm. Each beam is drawn as a ribbon as wide as the beam itself, made a few times wider so you can see it, and as bright as its power on a logarithmic scale: a 1 mW beam and a 1 µW reflection are both visible. The colour is the colour of the wavelength; infrared and ultraviolet beams are drawn in a dim false colour.

Detectors show what reaches them. A screen shows the light falling on it; a camera adds noise and saturation; a power meter reads watts; a spectrometer spreads the light by wavelength. Click one and its reading opens in the Inspect tab. Pin it to keep its card on the table while you work.

The experiments

18 ready-made benches, each with its own guide.

Interference

Diffraction

Polarisation

Rays and beams

Lasers

  • Build a HeNe laser: A discharge tube between two mirrors lases only when they are aligned and stable.
  • Passive Q-switching: A saturable absorber turns a continuous Nd:YAG laser into a train of nanosecond pulses.
  • Passive mode-locking: A fast saturable absorber locks a laser’s modes into picosecond pulses, c/2L apart.

Using the bench

Open the Parts tab and click a part to pick it up, or drag it onto the table. Select a part to see its settings, move it with the mouse or type its position, and turn it. Real optics sit on kinematic mounts with fine screws, and so do these: “Fine adjustment” tips and tilts a mirror by microradians and moves it by nanometres on a piezo, which is how you walk a beam into a pinhole or scan an interferometer.

Hover over a beam to read its power, width, wavelength and polarisation at that point. The Room tab sets the table’s isolation, air currents and temperature drift, and how the beams are drawn.

Everything you build is saved in your browser. Share copies a link that holds the whole bench, so whoever opens it sees exactly what you see; Save writes it to a file.

KeysWhat they do
Click a part in the Parts tabPick it up; click the table to put it down
Drag a partMove it, snapping to 5 mm (hold Alt to move freely)
Shift-clickAdd a part to the selection, or take it out
R, Shift+RTurn the selection 15° anticlockwise or clockwise
[ and ]Turn it by 1°
Arrow keysMove it 1 mm, or 0.1 mm with Shift
Ctrl+DDuplicate
DeleteRemove the selection
Ctrl+Z, Ctrl+YUndo and redo
FFit the table to the screen
EscapeDrop what you are holding, or clear the selection

How the light is computed

Each beam is a Gaussian beam: a narrow bundle with a width ww, a wavefront curvature and a direction. Its size and curvature live in one complex number, the beam parameter qq, and every lens, mirror and stretch of free space changes it by the ABCD rule

q′=Aq+BCq+D,1q=1R−iλπw2.q' = \frac{Aq + B}{Cq + D}, \qquad \frac{1}{q} = \frac{1}{R} - \frac{i\lambda}{\pi w^2}.

A beam carries its polarisation as a Jones vector and its phase as the optical path it has travelled. At a splitter it divides into two beams that remember they came from the same laser. Where beams from one laser meet on a detector, their fields are added, not their powers:

I=∣∑kEk eiϕk∣2.I = \Big|\sum_k E_k\,e^{i\phi_k}\Big|^2.

That sum is what makes fringes. Beams from different lasers, or from a lamp, add as powers unless their wavelengths are close enough to beat.

Slits, pinholes, gratings and other apertures need more than beams: the lab replays the light through them by Fresnel propagation on a grid, using the Fourier transform, and the pattern on the screen comes from that calculation, not from a textbook formula.

When mirrors around a gain medium send light back on itself, the lab finds the closed loop, works out the cavity’s mode from its round trip, and balances saturated gain against the losses. If the gain wins it lases and a beam comes out of the output coupler; tilt a mirror by too much and it stops.

What the lab leaves out

  • The table is flat: beams stay in its plane. Tip errors push a beam up or down and are counted as misalignment, but nothing is drawn in three dimensions.
  • Beams are Gaussian everywhere except behind apertures, where the Fourier engine takes over. Thick lenses and prisms have real surfaces and refraction, but aberrations show only as rays missing their focus.
  • Lasers run steadily, or in a steady train of identical pulses when a saturable absorber Q-switches or mode-locks them; the pulses are worked out in closed form, not built up round trip by round trip. There is no dispersion or Kerr effect in a cavity, no nonlinear crystals, and no electronics to drive a Pockels cell, a chopper or a lock-in.
  • Photon sources give count rates. Single clicks, coincidences and the correlations between a photon pair’s two arms are not simulated.

References

  • E. Hecht (2017), Optics, 5th edition, Pearson.
  • B. E. A. Saleh and M. C. Teich (2019), Fundamentals of Photonics, 3rd edition, Wiley.
  • A. E. Siegman (1986), Lasers, University Science Books.
  • H. Kogelnik and T. Li (1966), Laser beams and resonators, Applied Optics 5, 1550.
  • J. W. Goodman (2017), Introduction to Fourier Optics, 4th edition, W. H. Freeman.