= Faraday rotation in a glass-rod experiment
{c}
A William & Mary teaching experiment places a $10\,\mathrm{cm}$ glass rod inside a <solenoid>. Align a polarized <helium-neon laser> through the rod's centre onto a <photodiode> after a rotatable <polarimeter analyzer>. First scan the analyzer from maximum to minimum transmission in $5^\circ$ steps and record detector voltage, establishing the angular response. A <function generator> and <electronic amplifier> then drive sinusoidal <electric current> through the <solenoid>; a <multimeter> measures the <electric current> and an <oscilloscope> reads the <photodiode>. Trigger the <oscilloscope> from the generator, remove the large constant detector component by AC coupling, and average repeated traces. Convert the small synchronized voltage modulation to rotation using the calibrated angular slope. The manual fits rotation against <magnetic flux density>, using its stated coil calibration and converting measured <root mean square> current to peak current. This is a specified laboratory procedure, rather than measured results reproduced here.<Physics 251: Experimental Atomic Physics, Fall 2019/Faraday Rotation>[Physics 251: Experimental Atomic Physics, Fall 2019; William & Mary, Faraday Rotation laboratory chapter: printed pp. 2–7, apparatus, alignment, analyzer calibration, synchronized readout and equations (1.8)–(1.9)]{ref}
To test the inference, add matched zero-<electric current> readings before and after each reversal, hold the analyzer fixed during paired readings, and compare equal positive and negative <electric currents>. A blocked beam tests electrical pickup in the detector channel. These are proposed controls. <Joule heating> scales as current squared, so reversing <electric current> at equal magnitude helps separate an odd rotation from even heating effects; it does not remove drift between readings. Record <temperature> and allow repeatable settling. Changes in transmission, mechanical motion, stressed-glass <birefringence> and detector pickup can imitate a signal. Synchronized averaging suppresses unsynchronized noise, but cannot reject an artifact synchronized to the drive.
Quantitative extraction of the <Verdet constant> also needs the <integral> of <magnetic flux density> through the rod: a finite <solenoid> can have a nonuniform axial <magnetic field>. Use its calibrated path integral, or justify a uniform-field approximation; do not silently multiply a central field by the whole rod length. A signed reversal or phase-sensitive readout tests the part odd in applied <magnetic field>. Its agreement with the simple rotation law supports the model under these controls, rather than proving that every change of transmitted <light intensity> is <Faraday rotation>.
Back to article page
