Polarisation controller
A three-paddle polarisation controller: loops of fibre that act as a quarter-wave, a half-wave and a quarter-wave plate, each turned about the fibre. Together they take any polarisation in to any polarisation out.
Ports
Ports 1 and 2, one each side.
The model
A waveplate of retardance \(\Gamma\) with its fast axis at \(\theta\) from s is
\[\mathbf W(\theta, \Gamma) = \mathbf R(\theta)\begin{pmatrix} 1 & 0 \\ 0 & e^{i\Gamma} \end{pmatrix}\mathbf R(-\theta),\]with \(\mathbf R\) the rotation matrix. The controller is three of them, the paddles’ angles \(\theta_1, \theta_2, \theta_3\):
\[\mathbf M = \mathbf W\!\left(\theta_3, \tfrac{\pi}{2}\right)\mathbf W\!\left(\theta_2, \pi\right)\mathbf W\!\left(\theta_1, \tfrac{\pi}{2}\right), \qquad \mathbf S_{12} = \mathbf M, \qquad \mathbf S_{21} = \mathbf M^{\mathsf T}.\]Light going back through it sees the transpose, as in a real fibre: fibre birefringence is reciprocal. So a controller followed by a plain mirror sends back light whose polarisation depends on the paddles; followed by a Faraday mirror, it does not.
With the paddles at 0° the controller does nothing to s or p. Setting them is the job of a person at a real bench: the tool does not search for the angles, but each can be swept.
Settings
| Field | Default | Meaning |
|---|---|---|
a1 | 0 | Paddle 1, λ/4 (°) |
a2 | 0 | Paddle 2, λ/2 (°) |
a3 | 0 | Paddle 3, λ/4 (°) |
From the tool's own tables, file format version 1: the names are those in a .jekray file.
Assumptions
- Each paddle’s retardance is exact and flat in wavelength; it adds no loss.
Related
- Faraday mirror.
- In-line polariser and polarisation beam splitter.
- Waveplate, the free-space waveplate.