Page 185 - Photodetection and Measurement - Maximizing Performance in Optical Systems

Stability and Tempco Issues

178 Chapter Eight

1.44mW, or about a 55 percent intensity modulation. It can be seen that because
it is the amplitude that counts, even surprisingly small interfering intensities
can causes large power variations. Effective suppression is therefore very
difficult.
An anti-reflection (AR) coating on one surface can be used to reduce the inten-
sity of the second reflected beam. Standard “quarter-wave” AR coatings of mag-
nesium fluoride on glass provide about 1 percent power reflectivity, which
sounds pretty good, until it is realized that this means a 10 percent amplitude
reflection coefficient. A power reflectivity of less than 0.1 percent can be
achieved for higher index substrates, or over a narrow wavelength range by
using more complex dielectric stack designs, but this would normally be con-
sidered to be very good performance. Even this is 3 percent amplitude reflec-
tion coefficient, giving possible interferences with a ±6 percent intensity
variation. An alternative AR coating is a textured glass surface. These can also
provide power reflectivities less than 1 percent over a broad wavelength range,
and very high resistance to optical damage from high power lasers. See Yoldas
(1984) for one example. AR coatings are also wavelength- and angle-dependent,
more scattering than virgin surfaces, more susceptible to mechanical damage,
and difficult to clean. Hence AR coatings should be considered only part of a
solution to interference. Interference in detector windows can easily be the dom-
inant source of instability with laser measurements. Avoidance of overlapping
collimated beams should be the first consideration. If they must overlap,
perhaps the intersection angle can be big enough that many fringes appear
across the detector. Last, if the application can work without a high-coherence
source, for example by substituting a multimode semiconductor laser, then
achieving stability will be much easier. This is one approach taken in optical
disk readout optics.
A thick glass plate is an alternative beam-splitter. With the larger lateral sep-
aration of the two beams, overlap is reduced and physical interception and
absorption of the unwanted beam is easier. The only problems likely here are
increased optical aberrations and scatter caused by the thick glass, beam-shifts,
and perhaps extra cost.
As thick plates are one approach, very thin beam-splitters (“pellicles”) are
another offered by several manufacturers. These are membranes of polycar-
bonate or other tough plastic a few microns thick and held in tension on a metal
frame. The small thickness means that the two reflected beams are fully over-
lapped, and most sources will suffer interference, but the small optical path
difference (OPD a few l) causes little change in OPD with wavelength or tem-
perature. Hence there should be no intensity variation. Perhaps the biggest
problem with pellicles is their high compliance; they tend to act as microphones
unless carefully protected against air currents, sound and vibration. The small-
est pellicle which will accept the beam should be chosen, although the major-
ity of products are large (ª25 to 50mm diameter). Five millimeter pellicles
would be much stiffer, and more useful for much of modern optoelectronics.
Perhaps someone supplies them. Perhaps there is a market too for single crystal

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)
Copyright © 2004 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.