Interferometer Alignment
The previous section described how to align the
laser beam parallel to the holes of the optical breadboard/table.
- In the case of the Mach-Zehnder interferometer we put the first beam
splitter, aligning it so that it deflects part of the beam by 90 degrees.
- Then we put the two mirrors, aligning them so that the steered beams are
also parallel to the holes of the table. We put one of the mirror mounts on top of a translation stage pushed by a
micrometer screw. We then put a piezo stack (Thorlabs model AE0505D8) as a
spacer in between the screw and the stage. This piezo alows us to do a fine scan
of the
path length difference in the interferometer.
- We now have two beams intersecting in air. We put the second beam splitter
as well as we can so that the two beams meet at the partially reflective
surface. Align this beam splitter so that it steers the beams parallel to the
holes of the table.

Fig. 1. Steps to put together a Mach-Zehnder interferometer.
- We now look at the interference pattern by passing it through a diverging
lens and projecting it on a piece of paper. As we turn the screw of the linear
stage of the interferometer the fringes move in some direction (either out or
in). When we pass the point where the length of the two arms is exactly the
same, the fringes will switch from going in to going out, or vice versa. If
the aligning was done very carefully, maybe it is good enough to go to the
next step.
- We now put a white light source (a tungsten light bulb) at the input of the
interferometer. When we are close
to the point where the two arms have the exact length, we should see white
light fringes. This can be done visually, but we had a tough time doing this. We learned of a better method from Kwiat and co-workers at U.
Illinois. It consists of looking at the output with a spectrometer. We were
lucky that we had a small spectrometer connected to an optical fiber (Ocean
Optics). When we are close to the equal-arm-length point some
wavelengths will interfere constructively and some destructively. The spectrum
of the broad-band source would be continuous but with oscillations, with the
latter representing those interferences. The
frequency of these oscillations will change as we make small changes in the path
length. The latter is done by turning the micrometer of the stage where the
movable mirror is mounted. The spectra that we get as we turn the screw is shown in Fig. 2 below. The closer we are to the
equal length the broader the oscillations get.
Fig. 2 Spectra of a broad-band source taken after it passed through the interferometer. As
we get closer to
the equal-arm-length point the curves change in the progression: black, red, green and
blue. At the
equal-arm-length point the entire spectrum will go up or down depending on the
setting of the screw.
The Mach-Zehnder interferometer is easier to align than the Michelson
interferometer. This is because in the former the approximate equal-arm-length
point is roughly obtained when putting it together, as explained above. In the
case of the Michelson interferometer we have to get to the approximate position
by eye (Fig. 3), but could be far off target.

Fig. 3 Aligning the Michelson Interferometer can be tougher
than the Mach-Zehnder. A procedure similar to the one
described above can be used.
E.J. Galvez/Colgate U.
Return to main page