What techniques are used in the Analyzer part of an Ellipsometer? Specifically, what technique is used where the analyzer does not involve mechanical rotation? I assume it is a device to electronically rotate the plane of polarization of incoming light. Faraday Effect?
 A: From the "technology" section of the manufacturer:

The FS‑1 Detector Unit provides the Polarization State Detector for the ellipsometer system.  The FS‑1 Detector design has no moving parts, and is based on the Division of Amplitudes Polarimeter (DOAP) concept first proposed by Azzam (‘‘Division-of-amplitude photopolarimeter (DOAP) for the simultaneous measurement of all four Stokes parameters of light’’, R. M. A. Azzam, Opt. Acta Vol. 29, page 685, 1982).  One disadvantage of the traditional DOAP design is that the polarization measurement accuracy can be affected by the alignment and uniformity of the incoming beam.  To overcome this disadvantage, the FS‑1 Detector implements a novel optical configuration* to cancel measurement errors caused by mis-alignment or uniformity of the measurement beam.  Since there are no moving parts, the ellipsometric data can be acquired very quickly (10 ms minimum acquisition time for 4 wavelengths of data), and long term measurement stability and robustness is also improved.  The computer which controls the data acquisition and performs the data analysis is also integrated into the FS‑1 Detector Unit, resulting in a compact sensor that requires no additional control electronics.

The abstract of the Azzam reference reads:

To measure all four Stokes parameters of a light beam simultaneously, the beam is divided into four separate beams using a beamsplitter and two Wollaston prisms. Linear detection of the light fluxes of the four component beams gives four signals that determine the four Stokes parameters. This division-of-amplitude photopolarimeter (DOAP) is completely characterized by a wavelength-dependent 4 2 4 matrix that can be determined directly by calibration. The instrument has a fast response (that is limited only by the photodetectors) as it has no moving parts or modulation.

So there it is. No Faraday effect; just measuring multiple components along fixed axes allows you to fit all the parameters of the ellipse.  I suggest that you read the Azzam reference to get the exact details.
