close

Вход

Забыли?

вход по аккаунту

?

el%3A19910921

код для вставкиСкачать
function P[$ro s v 1 a] in eqn. 1 was then obtained by numerical integration.
Numerical results: The numerical results presented here
assume that the receiver filter is a third order ( N B = 3) Butterworth filter and BT = 1 which is the minimum baseband
bandwidth when statistical independence of the two samples
in the differential frequency detector is a good approximation.
Fig. 3 shows the probability of error as a function of E J N ,
for various values of modulation index. The error probability
0
I
I
0
5
I
I
I
IO
15
20
Eb IN,
.dB
ploi,l
Fig. 3 Probability oJ error as Junction of E$N, Jor sampling discriminator and differential frequency detection offiltered narrowband Manchester coded FSK with ET = 1.0
__ sampling discnminator
_
_ differential
~
~ frequency detection
A h=0.1
0 h=0.2
0 h=0.3
shown for the sampling discriminator detector is the lower of
the results obtained from sampling at T, or 7 ; . It can be
observed in the diagram that for low modulation index where
the modulated signal is narrowband, thc differential frequency
detector performs better than the sampling frequency detector.
The superior performance is observed whether sampling is
performed at T~ or at 7; in the sampling frequency detector.
As the modulation index is increased, the error probability
performance of the two detectors becomes nearly the same.
Conclusion: Differential frequency detection of Manchester
coded FSK has been proposed, analysed and compared with
the sampling frequency detector. It has been shown that the
differential frequency detector can perform better than the
sampling frequency detector under the same conditions.
Acknowledgment: This investigation was carried out at the
University of New South Wales, Australia where the author’s
many informative discussions with I. Korn are gratefully
acknowledged.
P. B. T. NYIRENDA
4th June 1991
University oJMalawi
Chancellor College
PO Box 280
Zomha, Malawi
PERFORMANCE CHARACTERISTICS OF
SURFACE PLASMON LIQUID CRYSTAL
LIGHT VALVE
Indexing terms. Liquid mystals, Optical modulation
An optically addressed light valve operating on the surface
plasmon resonance effecthas been developed and tested. The
structure consists of a thin amorphous siliwnfliquid crystal
sandwich, with an integral silver film to support the surface
plasmon mode. The performance characteristics have been
measured, and compare favourably with those of conventional liquid crystal light valves.
Introduction: The physical effects used to modulate light tend
to be weak, and therefore optical modulators typically require
large optical path lengths. This is particularly impractical
where large numbers of beam paths need to be modulated in
parallel, as with spatial light modulators of which the liquid
crystal light valve is perhaps the best known. One way to
increased modulation is to use an active material in an optically resonant structure, such as a Fabry-Perot cavity; this
effectively folds the optical path many times through the
material and multiplies the phase shift or attenuation accordingly. Another type of resonance that can be used in this way
is a resonantly excited guided mode. Such a mode can have a
field intensity much higher than that of the signal used to
generate it, and this increased intensity can also be used to
obtain enhanced modulation. Liquid crystal (LC) materials
can provide strong modulation over relatively short paths
(several micrometres) without resonance effects, but such
effects can be used to improve their low speed.
The particular guided mode we have chosen to exploit is
the surface plasmon. This is a mode supported by a single
interface between a metal and a dielectric, and is associated
with longitudinal charge density oscillations a t the metal
surface. The field is evanescent in both materials, with the
(transverse) decay distance typically about 0.2pm in the
dielectric and much less in the metal. The field energy is contained mostly (typically > 95%) in the dielectric material,
which makes modulation very efficient if this material is
active. In addition, the single-interface nature of the guide
means that effective path length uniformity is more easily
achieved.
The excitation of surface plasmons by prism coupling has
been widely reported.’.’ This is typically characterised by a
resonance curve, which is the dip in intensity of internally
reflected light as the angle of incidence is scanned past the
optimum angle for coupling (see Fig. 1). which is that where
the parallel wavevector in the prism equals the plasmon propagation constant. We have previously demonstrated that
prism coupled surface plasmons can be expleited in a spatial
light m o d ~ l a t o r . ~These earlier devices were electrically
addressed LC cells, and we have also reported optically
addressed devices: using photodiode array backplanes. Problems encountered with these latter devices included the difficulty of mounting the delicate 10pm photodiode wafers flat
on a robust substrate, and their high cost.
Experimental results: We have now converted to amorphous
silicon (a-Si) photoconductive backplanes. The device structure is illustrated in Fig. 1. The a-Si film is deposited on I T 0
index ( 1 73) prism and 51 de
smon suppoil8ng silver fclm 6 0 n m
3Cnm
References
c. H., TJHUNG, T. T., and SINGH, H.: ‘Performance of narrowband Manchester coded FSK with discriminator detector’, IEEE
Trans., May 1983, COM-31, pp. 659-561
2 KORN, I . : ‘Biphase frequency shift keying with discriminator detector followed by matched filter’, Austral. Tefecommun. Res (ATR),
1987,21, pp. 11-20
3 PAWULA, R . P., RICE, s. o., and ROBERTS, J. H.: ‘Distribution of the
phase angle between two vectors perturbed by Gaussian noise’,
I E E E Trans., August 1982, COM-30, pp. 1828-1841
I
MgFz al~gnmenllayer
TAN,
ELECTRONICS LETTERS
1stAugust 1991
Vol 27 No 16
2 b m teflon spacer
liqud c r y i l a l B D H E-7
MgFz alignment layer
gloss 5 U b i t r a l e
w r e I ghl
30nm
photoaddressable
3
~m a.5
backplane
-~
r
“ ( I
Fig. 1 SurJace plasmon light valve structure
1471
coated glass, giving a robust structure, and the Ag is evaporated on the high index upper plates. MgF, aligning layers are
then deposited on both plates, using oblique angle evaporation, and the cell is then glued together with the spacers
inserted, and filled. The bias potential is split between the
photoconductor (PC) and the LC; when the P C is illuminated
by the write beam, the voltage across the LC increases,
causing the effective index to increase. This alters the plasmon
resonance condition at the Ag surface, and consequently the
fraction of the incident beam which is absorbed. The LC must
be thin to achieve the correct splitting of the bias potential
between the LC and the PC, and due to lateral spreading of
the bias field, the device thickness will also impose a limit on
the spatial resolution that can be obtained. Using microbead
spacers, a thickness of 2 p m was obtain in these experiments.
Typical reflectivity curves for the a-Si devices are shown in
Fig. 2. The optimum bias potential was found to be about 4 V
RMS; this was gave significant switching in the LC, and
commonly used for optically addressed S L M S . ~The results
are shown in Fig. 4, at 500pW/cmZ write power, for the cases
of grating wavevector parallel and perpendicular to that of the
plasmons. We attribute the difference between the M T F
responses for these two cases to the directional nature of the
guided wave modulation mechanism used.
0
5
IO
15
20
25
spatial resolution,lpimm
30
001141
Fig. 4 M T F response functions parallel and perpendicular to plasmon
propagation direction
(i) Parallel
(ii) Perpendicular
64
65
66
67
68
,
,
,
,
,
I
69
70
71
72
73
7L
angle o f tncidence,deg
Acknowledgments: We would like to thank the Science and
Engineering Research Council for financial support of this
work, and the GEC-Marconi Research Centre for kindly providing the amorphous silicon backplanes. We would also like
to thank M. Green for his assistance and support.
ioolizl
M. E. CALDWELL
E. M. YEATMAN
Department of Electrical Engineering
Imperial College
London SW7 ZBT, United Kingdom
Fig. 2 SPR response against angle of incidence
(i) Unbiased
(ii) Biased at 4 V RMS
(iii) Biased as in (ii) and fully written
higher potentials caused excessive increase of the reflectance
minimum in the dark (unwritten) state. Both read and write
beams were of 633nm wavelength, and the write intensity was
500pW/cm2. Maximum contrast attainable was about 5 : 1;
however, we have obtained better than 100 : 1 in electrically
addressed devices. The reduced performance in this case is due
to excessive voltage being applied to the LC in the dark state;
ideally the dark resistance of the a-Si should be substantially
more than the resistance of the LC layer, but in our present
devices the two values were of the same order. In previous
devices, extra dips have appeared in the reflectivity curve due
to bulk guided modes in the liquid crystal, but these were not
present in this case due to the reduced LC thickness. Fig. 3
shows the reflected read beam intensity as a function of write
beam intensity at the incident angle of maximum contrast.
Rise and fall times were found to be about 1 ms and 3 ms,
respectively; the latter is about 30 times faster than the cell in
its conventional transmission mode. The speed improvement
is due to the rapid relaxation of the LC molecules near the cell
surfaces, where the plasmon field energy is contained.
T o characterise the spatial resolution, the modulation transfer function (MTF) was measured by illuminating the PC with
an interference pattern (grating) of variable pitch, a method
10th June 1991
References
‘The determination of the optical constants of
metals by the excitation of surface plasmons’, Z . Phys., 1971, 241,
pp. 313-324
BURSTEIN, E., CHEN, w. P., CHEN, w. I., and HARSTEIN, A.: ‘Surface
polaritons’, J. Vac. Sci. & Technol., 1974, 11, pp. 1004-1021
YEATMAN, E. M., and CALDWELL, M. E.: ‘Spatial light modulation
using surface plasmon resonance’, Appf. Phys. Lett., 1989, 55, pp.
613415
CALDWELL, M. E., and YEATMAN, E. M’:‘Optically addressed surface
plasmon spatial light modulators’, Proc. SPIE, 1990, 1280, pp.
276288
ARMITAGE, D., THACKARA, J. I., and EAOS, w. D . : ‘Photoaddressed
liquid crystal spatial light modulators’, Appl. Opt., 1989, 28, pp.
47634771
KRETSCHMANN, E . :
SYSTEM CHARACTERISATION OF HIGH
GAIN AND HIGH SATURATED OUTPUT
POWER, Pr3+-DOPED FLUOROZIRCONATE
FIBRE AMPLIFIER AT 1.3pm
Indexing terms: Optical fibres, Amplifers
We report the first use of a high gain (24dB) and high saturated output power ( > 100mW) Pr’+-doped fluorozirconate
fibre amplifier in a 2.5Gbitls system. A sensitivity of
- 37-3dBm has been achieved when used as a preamplifier in
conjunction with a pin-FET receiver. This we believe represents the most sensitive reported I.3pm direct detection
receiver at this bit rate.
0
100
200
300
LOO
500
write beam power, p W l c m Z
Fig. 3 Read signallwrite signal response curue
1472
600
E
Introduction: The past few years have seen an upsurge of
interest in rare earth doped fibre amplifiers for use within
optical communications systems.’ Most of this interest has
been concentrated on the 1.55pm transition in erbium-doped
ELECTRONICS LETTERS
1stAugust 1991
Vol 27
No 16
Документ
Категория
Без категории
Просмотров
2
Размер файла
201 Кб
Теги
3a19910921
1/--страниц
Пожаловаться на содержимое документа