2

0
1
0
:

I
n
s
t
i
t
u
t
o

d
e

A
s
t
r
o
n
o
m

a
,

U
N
A
M

-

A

L
o
n
g

W
a
l
k

T
h
r
o
u
g
h

A
s
t
r
o
n
o
m
y
:

A

C
e
l
e
b
r
a
t
i
o
n

o
f

L
u
i
s

C
a
r
r
a
s
c
o

s

6
0
t
h

B
i
r
t
h
d
a
y
E
d
.

E
.

R
e
c
i
l
l
a
s
,

A
.

L
u
n
a
,

&

Y
.

D
.

M
a
y
y
a
RevMexAA (Serie de Conferencias), 37, 156159 (2009)
A DESIGN OF AN ACOUSTO-OPTICAL SPECTROMETER
G. Herrera-Martnez,
1
A. Luna,
1
L. Carrasco,
1
A. Shcherbakov,
1
D. S anchez,
1
E. Mendoza,
1
and F. Renero
1
RESUMEN
La radioespectroscopa se ha convertido en una herramienta fundamental para estudiar objetos astron omicos
en la banda de las microondas. Por lo tanto, es necesario el dise no y construcci on de instrumentos con alta
resoluci on espectral, bajo consumo de potencia y dimensiones lo m as compactas posibles que hagan f acil su
manejo y transporte. Aqu presentamos el dise no y pruebas de un espectr ometro acusto- optico para uso en
radioastronoma solar y para el estudio de la variabilidad de m aseres c osmicos con una antena de 5 metros de
di ametro (RT5) que se esta reinstalando en Sierra Negra. Mostramos las primeras evaluaciones del desempe no
de los componentes y el ensamble de laboratorio.
ABSTRACT
The radio spectroscopy has became a fundamental tool to study astronomical objets at the microwave band.
Therefore, the design and construction of instruments with high spectral resolution, low power consumption
and compact size for easy handle and transport are necessary. Here we present the design and the tests of
an acousto-optical spectrometer for use in solar radio astronomy and for variability studies of cosmic masers
sources with a 5 meter antenna (RT5) which is being reinstalled at the Sierra Negra site. We present the rst
evaluations of the performance of the components and the laboratory assembly.
Key Words: Spectrometers Acousto optics Instrumentation
1. INTRODUCTION
The reinstallation of a 5 m dish radio telescope
give us the oportunity to explore alternatives for
spectrometer techniques. Three spectrometers types
are common in radio astronomy: lter-bank, auto-
correlator and acousto-optic. The lter-bank-type
sprectrometers have several limitations: the main-
tenance and calibration required for every channel,
and the xed channel width. The autocorrellation-
type is limited mainly by the power consumption
and bandwidth because of current computer speed.
The acousto-optical spectrometer (AOS) has been
extensively used in radio astronomy, since Lambert
described its principle (Lambert 1962). They are an
ecient option for satellite telescopes (Klumb et al.
1994), due to the low power consumption and com-
pact size which make easy to handle and transport.
Our main interest is the study of dierent astro-
nomical objects at the millimeter band ( 43 GHz).
Figure 1 shows, for some of such objets, the required
resolution and bandwidth. The plot shows that AOS
is very well positioned for stellar formation and in-
terstellar medium studies at these frequencies.
The principle of performance is as follows: the
intermediate frequency (IF) of an heterodyne re-
ceiver of a radio telescope is injected into a piezo-
1
Instituto Nacional de Astrofsica

Optica y Electr onica,
Mexico (gherrera@inaoep.mx).
Fig. 1. Spectrometer bandwidth and spectral resolution
at 43 GHz. The plot, adapted from (Harris 2002), shows
the required velocities and frequencies for typical sources.
The dashed line indicates the region of spectral resolution
and bandwidth for this spectrometer according to the
evaluation shown in 3.
electric transducer that mechanically modulate a
Bragg cell illuminated by a collimated light beam.
This Bragg cell, also known as Acousto-Optical De-
ector (AOD), is an acusto-optical crystal of TeO
2
which deects a laser light beam as a function of
the injected IF. The TeO
2
crystal is known for its
acousto-optical properties (Uchida & Kiizeki 1973).
156

2
0
1
0
:

I
n
s
t
i
t
u
t
o

d
e

A
s
t
r
o
n
o
m

a
,

U
N
A
M

-

A

L
o
n
g

W
a
l
k

T
h
r
o
u
g
h

A
s
t
r
o
n
o
m
y
:

A

C
e
l
e
b
r
a
t
i
o
n

o
f

L
u
i
s

C
a
r
r
a
s
c
o

s

6
0
t
h

B
i
r
t
h
d
a
y
E
d
.

E
.

R
e
c
i
l
l
a
s
,

A
.

L
u
n
a
,

&

Y
.

D
.

M
a
y
y
a
AN ACOUSTO-OPTICAL SPECTROMETER 157
Fig. 2. Light intensity distribution in the Bragg regime,
where the interaction lenght is large, this regime provides
an oportunity of realizing 100% of light scattering.
The propagation of the acoustic wave through the
photo-elastic medium produces traveling periodic
variations in the refractive index of this medium.
The diraction pattern depends essentially on the
length L of the interaction zone between the light
beam and the acoustic waves. So the regimes of
diraction have two limiting cases, corresponding to
a short and a large lengths of interaction. The Bragg
diraction regime, that we use, occurs with a large
L. In this case the dynamic acoustic grating is rather
thick. So, during the analysis of the diraction pro-
cess, one has to take into account the phase relations
between waves of dierent orders. Such regime can
be realized only when the angle of incident light
B
on a thick acoustic grating meets the Bragg condi-
tion
sin
B
=

2n
;
and
Q =
L

2
1.
Here, Q is the Klein-Cook paramter (Klein &
Cook 1967), is the laser wavelength, n is the re-
fraction index of the crystal and is the acoustic
wavelength. The light intensity in orders of scatter-
ing is shown in Figure 2, where
q = (cos )
1

2M
2
PS
1
;
M
2
is the acousto-optic gure of merit, and P/S is
the acoustic power density. It is desirable to perform
the Bragg regime since it reaches 100% eciency of
light scattering.
Fig. 3. Comparison of the simulated spot size (subframe)
with OSLO EDU 6.4.5 and a measurement of a large
format lenses system of Edmund optics with FWHM
24 m.
TABLE 1
SPOT SIZE EVALUATION
Stock Lens Focal Spot Airy
number diam. distance diam. diam.
(mm) (mm) (m) (m)
Edmund 40 200 6.5 16
u32-925
Melles Griot 50 250 6.5 18
lao-631
Thorlabs 50.8 250 6.4 14
AC508-250-A1
Newport 50.8 250 6.4 14
PAC088
2. FOCUSING SYSTEM AND SPOT SIZE
EVALUATION
We have theoretically investigated the potential
size of an individual resolvable spot inherent in vari-
ous lens sets in a view of optical matching the AOD
aperture (about 40 mm) with a pair of the CCD-
camera pixels (about 14 m). For this purpose, we
have used the OSLO EDU 6.4.5 ray tracing software
and analyzed the spot size of more than 20 various
types of lens sets. The result are given in Table 1,
were it may be seen that we have obtained two poten-
tially acceptable lens sets from Thorlabs and New-
port with a same spot size of about 14 m and a
focal distance of 250 mm. They both are potentially
suitable for the goals of the AOS under construction.
Figure 3 shows the spot size measured in the focal
plane of the focus lens system. This size denes the

2
0
1
0
:

I
n
s
t
i
t
u
t
o

d
e

A
s
t
r
o
n
o
m

a
,

U
N
A
M

-

A

L
o
n
g

W
a
l
k

T
h
r
o
u
g
h

A
s
t
r
o
n
o
m
y
:

A

C
e
l
e
b
r
a
t
i
o
n

o
f

L
u
i
s

C
a
r
r
a
s
c
o

s

6
0
t
h

B
i
r
t
h
d
a
y
E
d
.

E
.

R
e
c
i
l
l
a
s
,

A
.

L
u
n
a
,

&

Y
.

D
.

M
a
y
y
a
158 HERRERA-MART

INEZ ET AL.
Fig. 4. Injected signal of the AOD driver signal sweeped
from 55 to 90 MHz, with a 5 MHz step.
spectrometer band resolution. For a Edmund Optics
focusing lens system a spot size of 4 pixels at FWHM
was reached ( 28 microns).
3. AOD EVALUATION AND PERFORMANCE
The AOD, manufactured by Molecular Technol-
ogy GmbH, consists of a driver, which manages the
injected signal to the transducer, and the optical
block, where the piezoelectric transducer, the crystal
and the absorber for the acoustic wave, are placed.
First, we perform computations for the spectral
resolution (f) and the bandwidth (F) intrinsic to
our AOD. With a large optical window and a low
sound velocity into crystal we can improve the re-
soution. The data sheet of AOD specics the sound
velocity in the crystal, 650 km s
1
, and the optical
window, 4010 mm. A linear CCD array with 3000
pixels (#CCD) and a pixel size of 14 200 m was
used. We obtain
f =
V
sound
L
crystal
=
650ms
1
40mm
= 16.24 KHz ,
and the bandwidth
F =
f#CCD
2
= 24.3 MHz .
In terms of the velocity at the 43 GHz fre-
quency (
0
= 7mm) the results are 0.1 km s
1
and
168 km s
1
respectively (dashed line of Figure 1).
In the practice, we rst evaluate the dispersion
angle() of the AOD at dierent frequencies using
the aproximation:
tan =
f
V
sound
;
TABLE 2
EVALUATION OF THE DIFFRACTION
ANGLES
Frequency Distance on Expected Measured
(MHz) screen (mm) value (

) value (

)
55 79 3.07 3.08 0.03
60 86 3.34 3.35 0.03
65 93 3.62 3.62 0.03
70 101 3.90 3.93 0.03
75 107 4.18 4.16 0.03
80 114 4.45 4.43 0.03
85 122 4.73 4.74 0.03
90 128 5.01 4.98 0.03
Fig. 5. Qualitative spectrum of the AOS without fo-
cusing optics and a CCD integration time of 1s. For
frequency calibration the better t was linear, with this,
the bandwidth by a pair of pixels are 32 KHz.
where f is the frequency of the acoustic wave, is
the laser wavelength and V
sound
is the sound velocity
in the crystal.
A He-Ne laser module of 633nm@12mW with lin-
ear polarization was used to illuminate orthogonally
the AOD window. The projection of the diracted
beam was measured in a screen at 1470 mm away
from the output of the cell. A signal frequency
sweeped in frequency from 55 MHz to 90 MHz with
a 5 MHz step was used. The results are shown in
Figure 4.
Measurements of the diraction angle are listed
in Table 2. The percentage error is less than 10%
of the expected value. The scanning angle in the air
for all the bandwidth is 2.2

0.03, in good agree-

). The AOD was

2
0
1
0
:

I
n
s
t
i
t
u
t
o

d
e

A
s
t
r
o
n
o
m

a
,

U
N
A
M

-

A

L
o
n
g

W
a
l
k

T
h
r
o
u
g
h

A
s
t
r
o
n
o
m
y
:

A

C
e
l
e
b
r
a
t
i
o
n

o
f

L
u
i
s

C
a
r
r
a
s
c
o

s

6
0
t
h

B
i
r
t
h
d
a
y
E
d
.

E
.

R
e
c
i
l
l
a
s
,

A
.

L
u
n
a
,

&

Y
.

D
.

M
a
y
y
a
AN ACOUSTO-OPTICAL SPECTROMETER 159
Fig. 6. Actual assembly of AOS, all the components are
mounted in a optical bench of 1 2 ft, the expansion of
laser beam is doing by a galilean telescope with magni-
cation of 15x.
constructed so that the incident light beam should
be orthogonal to the input optical plane of the cell,
and the light polarization plane should be parallel to
the base of optical block. This construction provides
the maximum scanning angle value.
We can see that the range of diraction angle is
appropiate for our purposes. To get a better fre-
quency resolution we need a focusing lens with a
good spot size at the focal plane.
4. WORK IN PROGRESS
To get the resolution closer to the desired one of
16 KHz it is necessary to expand and collimate the
laser beam to fully illuminate the optical window
of the AOD. Our rst approximation to expand the
beam is using a Galilean telescope with a magni-
cation of 15 and a laser diode module with beam
diameter of 1.55.25@1/e
2
. We will analyze the use
of a set of prisms to expand the laser beam. Software
of the driver control of the AOD and the acquisition
data software of the CCD camera will be improved to
do it more ecient and obtain a faster performance.
In Figure 5, the measurement of the intensities
at dierent angles of diracted beam on the CCD
camera without focusing optics are shown. For the
frequency calibration a gaussian t for every beam
was made so we locate the pixel postion where the
maximums are. Using least squares we calculate the
better t, there is a lineal relationship between the
separation of the maximums and the deection angle
of the beam. With this, the spectral resolution in a
pair of pixels is 64 KHz. Furthermore we can see
a surrounding shape like a function of the deected
angle. If we change the frequency of the acoustic
wave, we will see how it varies the intensity of the
deected beam too.
The laboratory assembly is shown in Figure 6.
Two mirrors are needed to bend the optical path
due to the dimentions of the optical bench. Some
additional mechanical parts have been made to get
the best alignment of the optical path.
REFERENCES
Harris, A. I. 2002, in Proc. Far-IR, Sub-mm and Millime-
ter Detector Technology Workshop, ed. J. Wolf, J.
Farhoomand, & C. R. McCreight, NASA CP-211408
Klein, R. W., & Cook, B. D. 1967, IEEE Trans. Sonics
and Ultrasonics, 14, 123
Klumb, M., Frerick, J., Tolls, V., Schieder, R., & Win-
newisser, G. F. 1994, Proc. SPIE, 2268, 305
Lambert, L. B. 1962, IRE Int. Conu. Rec., 6, 69
Scalise, E. 1999, ASSL, 241, 447
Schieder, R. T., et al. 2003, Proc. SPIE, 4855, 290
Shcherbakov, A., Luna, A., Ledeneva, Y., & Maximov, J.
2007, in Proc. XXII Annual Congress Soc. Mexicana
Instrum.
Shcherbakov, A., Luna, A., & S anchez, D. 2008, in Proc.
XXIII Annual Congress Soc. Mexicana Instrum.
Uchida, N., & Kiizeki, N. 1962, Proc. Inst. Radio Elec.
Electron. Engrs., 61, 1073