The 3.6 m Indo-Belgian Devasthal Optical Telescope:
Performance results on site
Nathalie Ninane, Christian Bastin, Carlo Flebus and Brijesh Kumar*
Advanced Mechanical and Optical Systems (AMOS s.a.), LIEGE Science Park, B-4031 ANGLEUR
(Liège), BELGIUM
* Aryabhatta Research Institute of Observational Sciences (ARIES), Manora Peak, Nainital, 263 129
India
ABSTRACT
AMOS SA has been awarded of the contract for the design, manufacturing, assembly, tests and on site installation
(Devasthal, Nainital in central Himalayan region) of the 3.6 m Indo-Belgian Devasthal Optical Telescope (IDOT).
The telescope has Ritchey-Chrétien optical configuration with one axial and two side Cassegrain ports. The meniscus
primary mirror is active and it is supported by pneumatic actuators. The azimuth axis system is equipped with hydrostatic
bearing.
After successful factory acceptance at AMOS SA, the telescope has been dismounted, packed, transported, and remounted on site. This paper provides the final performances (i.e. image quality, pointing and tracking) measured during
sky tests at Devasthal Observatory.
Keywords: telescope tests, active optics, image quality, pointing, tracking, IDOT
1. INTRODUCTION
(1)
After telescope design , AIV and factory acceptance tests(2) performed in AMOS assembly hall, IDOT was dismounted
and packed to be sent to its site, Devasthal in India. End of 2014, the building was ready for telescope remounting and
commissioning. Few months later, first light, and fine tuning could start. And finally, performance and acceptance tests
were conducted end of 2015.
After a short telescope overview and requirement discussion, some acceptance test results are presented.
2. TELESCOPE OVERVIEW
The main characteristics of the telescope are summarized in Table 1. As shown in Figure 1, the optical combination is a
Ritchey-Chrétien type with a Cassegrain focus where the light beam can be directed toward a main port designed for
interfacing an instrument with a mass up to 2 tons or toward two side ports for smaller instruments. The mount type is an
alt-azimuth. The telescope weights 150 tons. It rotates around the azimuth axis thanks to an hydraulic track(3). The
telescope is equipped with an active optic system(4) (AOS) that controls the primary mirror figure and the secondary
mirror positioning to keep the telescope wavefront error in the specification for any operational conditions. The primary
mirror is a meniscus 165 mm thick, 3700 mm diameter supported by 69 axial actuators and 24 lateral astatic levers.
While an Acquisition and Guiding Unit (AGU) aligned on a guide star at the edge of the telescope field of view
measures the wavefront and tracking errors, the set of forces applied by the actuators to the mirror is adjusted
continuously. In parallel, the Telescope Control System(5) (TCS) computes the telescope trajectory taking into account of
the weather conditions, the pointing model of the telescope and the tracking errors measured by the AGU.
Ground-based and Airborne Telescopes VI, edited by Helen J. Hall, Roberto Gilmozzi, Heather K. Marshall,
Proc. of SPIE Vol. 9906, 99064E · © 2016 SPIE · CCC code: 0277-786X/16/$18 · doi: 10.1117/12.2232775
Proc. of SPIE Vol. 9906 99064E-1
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 08/08/2016 Terms of Use: http://spiedigitallibrary.org/ss/termsofuse.aspx
Figure 2 shows pictures of the telescope on site. The telescope sizes are: height 13 m, width 7 m and total weight 150
tons.
The telescope is installed at Devasthal, North of India, at an altitude of 2540 m. A picture of the dome is in Figure 3; it is
a rotating cylinder equipped with as large slit and a wind screen. Large fans all around enable the venting and limit the
dome seeing.
1
3
2
6
4
5
Figure 1: ARIES telescope and ray tracing showing the side port and axial port optical path; with 1) the primary mirror, 2) the
secondary, 3) the side port focal plane, 4) the side port folding mirror, 5) the field corrector and 6) the axial focal plane.
Type:
Focal length:
Aperture stop:
2 focal plane configurations:
Field of View:
Operational waveband:
M1 characteristics:
M2 characteristics:
Distance M1-M2:
Back focal length:
Focal plane Radius of curvature:
Scale plate:
Ritchey - Chrétien
32.4 m (telescope F#/9 with M1 F#/2)
3.6 m on M1
Side Port and Axial Port
10 arcmin on side ports,
30 arcmin on axial port,
(35 arcmin for the AGU)
350 nm to 5000 nm
R = 14638.87 mm CC
K = -1.03296
Optical Φ = 3600 mm, mechanical Φ = 3700 mm
R = -4675.30 mm CX
K = -2.79561
Optical Φ = 952 mm, mechanical Φ = 980 mm
5.51 m
2.5 m
1935.78 mm
157 µm / arcsec (0.006 arcsec/µm)
Table 1: Summary of the telescope characteristics
Proc. of SPIE Vol. 9906 99064E-2
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 08/08/2016 Terms of Use: http://spiedigitallibrary.org/ss/termsofuse.aspx
°
Figure 2: IDOT on site
Figure 3: IDOT building
3.
REQUIREMENTS AND REQUIREMENT VERIFICATIONS
Table 2 summarizes the main performance requirements specified for 3.6 m IDOT.
Image quality
Optical main requirements
- Encircled Energy 50% < 0.3 arcsec,
- Encircled Energy 80% < 0.45 arcsec,
- Encircled Energy 90% < 0.6 arcsec,
For the waveband 350 nm to 1500 nm; without corrector
for 10 arcmin FOV.
Proc. of SPIE Vol. 9906 99064E-3
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 08/08/2016 Terms of Use: http://spiedigitallibrary.org/ss/termsofuse.aspx
Mechaniical main requ
uirements
Sky coveragge (elevation range)
r
15 too 87.5°
Pointing acccuracy
< 2 arcsec
a
RMS
Tracking acccuracy
< 0.11 arcsec RMS for 1 minute in open loop,
< 0.11 arcsec RMS for 1 hour in close loop,
< 0.55 arcsec Peak for 15 minutees in open loopp.
Table 2: Main
M telescope reequirements
The telescopee requirementts are defined to get an imaage quality degradation veryy small compparing to the contribution
c
of
the seeing. Moreover
M
the seeing
s
is variaable in time, depends
d
of thee pointing direection and cannnot be measu
ured directly at
a
telescope level. The imagee quality requuirements thuss cannot be diirectly verifiedd in pointing stars. Instead
d of measuringg
w
senssor is installedd in the focal plane
p
of the teelescope. In addjusting the in
ntegration timee
directly the sttar image, a wavefront
of the wavefrront sensor caamera the seeing is averaged and the wav
vefront error can
c be measurred. A relation
nship betweenn
the measuredd WFE and thee encircled eneergy requirem
ments shall thu
us be identifiedd.
A Monte-Carrlo analysis coonsidering a laarge panel of wavefront
w
erro
or sources dettermined the rrelationship off the WFE andd
the encircled energy at λ= 1500 nm. Figgure 4 shows the
t results forr of 50%, 80%
% and 90% encircled energy
y respectivelyy.
E
< 0.455 arcsec for λ = 1500 nm is reached for WFE
W < 210 nm
m; the criteria EE50% < 0.3
3 arcsec is lesss
The criteria EE80%
constraining than EE80% < 0.45 arcsecc and is reachhed for a WFE
E RMS up too 257 nm. Criiteria EE90% is difficult too
P and is veery dependentt of the spatiaal
work with beecause it is reaached in integgrating energyy that is in thee foot of the PSF
frequency of the WFE; thiss criterion cann thus not be verified
v
by test.
Tracking acccuracy can be critical for faint
f
object obbservations th
hat need long integration tiime. The mov
vement of thee
image at telesscope focus while
w
integratinng will enlargge the image and
a decrease thhe telescope rresolution. Ass for the imagee
quality, the effect
e
of the seeeing that willl produce a high
h
temporal frequency tipp/tilt of the im
mage will mak
ke the trackingg
requirements very difficultt to demonstraate. Temporall integration is
i needed to suppress
s
the seeing but by the same timee
p
to testinng the trackingg on the sky, factory tests(22)
could also hidde image jitteer produced byy the telescoppe itself. But prior
were perform
med for measuuring the axis control follow
wing errors th
hat are recordded at 200 hz.. In exploring
g the telescopee
field of regardd with the corrresponding axxis speeds andd accelerations; no vibrationn appeared.
The tracking errors were measured
m
in recording the im
mages of a star with a cameera at one of thhe foci of the telescope.
t
Thee
wing errors aree also recordeed.
image centroiids are compuuted and the asssociated statiistics. In paralllel axis follow
Pointing accuuracy specificaation is large enough comppared to the seeeing and is thhus easier to vverify. The ceentroids of starr
images are siimply recorded while pointting the telescope toward sttars covering the
t field of reegard. The disspersion of thee
centroids givees then the poointing error.
Spec: 0.33’’
Proc. of SPIE Vol. 9906 99064E-4
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 08/08/2016 Terms of Use: http://spiedigitallibrary.org/ss/termsofuse.aspx
Spec: 0.45’’
Sppec: 0.6’’
F
Figure
4: Montee-Carlo analysiss results; EE50%
%, EE80% and EE90% vs WF
FE
4. IMAG
GE QUALIT
TY TESTS
m
mountedd on a movingg arm can seleect a guide staar in the telesccope field of vview up to a diameter
d
of 355
A pick-off mirror
arcminutes. The
T guide staar beam is theen split in tw
wo wave ranges feeding a guider camerra and the AG
GU wavefronnt
sensor. The measurements
m
s given by the AGU WFS will
w drive the active optic system.
s
The w
wavefront erro
or is processedd
to get de-rotaated Zernike polynomials
p
c
cleaned
from field of view contributors. Focus and C
Coma terms arre corrected inn
moving the secondary
s
mirrror thanks to a hexapod. Astigmatism
A
and
a spherical aberration aree cancelled in
n adjusting thee
forces of the actuators suppporting the prrimary mirror.. A good imag
ge quality willl then be obtaained only if i)) the telescopee
q
iii))
is preliminaryy well alignedd; ii) the field of view contrributors in the wavefront measurement arre accurately quantified;
the hexapod and
a the actuattors are drivenn correctly.
The AGU WF
FS sensor getss an 11 X 11 lenslet
l
array thhat is a low sp
patial samplinng but is optim
mized for work
king with fainnt
guide stars.
The test WFS
S that is installled at telescope main focuss is equipped with a 30 X 30
3 lenslet arraay that gives a better spatiaal
resolution. Fiigure 5 gives a picture of thhe test WFS inn telescope foccal plane.
The image quuality measureements were performed
p
forr several testin
ng configuratiions; with andd without field
d corrector, inn
the center or in the field off view and for many telesccope elevation
n angles. Mostt of the time, the seeing wh
hile measuringg
a
As farr as local seeinng in the dom
me was reduceed using fans and WFS inteegration timess
was between 0.7 and 1.5 arcsec.
properly adjuusted, few minnutes after cllosing the actiive system lo
oop the waveffront requirem
ment was reacched. Figure 6
shows a typiccal plot of thee WFE measuurements acquuired with AG
GU WFS andd Test WFS siimultaneously
y vs time. Thee
wavefront errror measured with AGU WFS
W on whichh the active optic
o
is drivenn converges qquickly around
d 75 nm RMS
S
while Test WFS
W
stabilizes around 1255 nm RMS. The
T small diffference betw
ween those ressults comes mainly
m
by thee
accuracy of thhe AGU data processing annd by the finerr sampling of Test WFS. Fiigure 7 presennts a typical measurement
m
of
the telescope image qualitty. The spot pattern
p
on thee WFS cameraa, the computted WFE and the correspon
nding Zernikee
polynomials measured
m
in the
t focal planee of the telesccope at the cen
nter of the FoV without corrrector are giv
ven. The WFE
E
RMS is 88 nm
m. WFE’s dow
wn to 90 nm RMS
R
were reggularly achieveed in good seeeing conditionns.
Proc. of SPIE Vol. 9906 99064E-5
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 08/08/2016 Terms of Use: http://spiedigitallibrary.org/ss/termsofuse.aspx
Corrector
Test WFS
Figure 5: Test WFS mounted in telescope focal plane
ARIES - ACS In CL - WFE
H1P5722_ TesiCL_2015_11_03_1234116px1
600
AGU WFS
-4 -Test WFS
500
*
Test WES - focus removed
400
Ñ
F
cc
E 300
Lai
200
100
0
23 30
23:40
23:50
00:00
Time
00:10
00.20
00 30
Figure 6: Telescope WFE at the start of the closing of the active loop
Proc. of SPIE Vol. 9906 99064E-6
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 08/08/2016 Terms of Use: http://spiedigitallibrary.org/ss/termsofuse.aspx
Measurement, 151104 h00m11 s04 TEST_WFS.shz
Mean 3.76879e -018
Corr. wave - front, 151104 h00m11 s04 TEST_WFS.shz
Zernike coefficients (ISO)
RMS 0.0883415
Residuum:
P -V 0.618987
0.3166121 0.042922
Max 0.255609
CO
Min -0.363378
Cl
C2
C3
C4
C5
C6
C7
C8
C9
C10
0.2
0.1
C11
-0.1
C12
C13
C14
C15
C16
C17
C18
C19
C20
-0.2
-0.3
i
2
24.05.2016
i
3
i
i
4
5
iri
6
7
x/rnm
8
C21
9
10
11
12
C22
C23
C24
0.0333084
-0.00941796
-0.111324
-0.0870605
0.117482
0.0283068
0.00649092
0.162622
0.118827
0.0528059
-0.00338887
0.044095
0.0109573
0.00764345
-0.0265882
-0.0630979
-0.0158248
0.00578928
0.011438
-0.0165362
-0.0110114
0.00928714
-0.0539363
0.0255842
0.0696476
piston
tilt, x
tilt, y
defocus
Ast. 0°, 1st
Ast. 45°, 1st
Coma x
Coma y
Sph. ab.
trifoil 0°
trifoil 30°
Ast. 0 °, 2nd
Ast. 45 °, 2nd
radial term
tetrafoil 0°
tetrafoil 22,5°
radial term
Figure 7: Telescope WFE: WFS spot pattern, WFE map and Zernike coefficients (in µm).
Proc. of SPIE Vol. 9906 99064E-7
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 08/08/2016 Terms of Use: http://spiedigitallibrary.org/ss/termsofuse.aspx
To cross-checck the excellennt results, imaages of doublee stars were reecorded to verrify the accorddance of the measured
m
WFE
E
and the telesccope resolutioon that can be achieved durring good seeiing nights. Figgure 8 shows a picture of the double starr
HIP117695 separated
s
by 1.2 arcsec, pointed
p
at an altitude anglle of 60° aboove horizon w
with an expossure time of 4
seconds. Figuure 9, double star
s HIP1176446 separated by
b 0.6 arcsec, exposure tim
me 1.5 secondss at an altitudee angle of 45°°.
It has to be noticed
n
that thhe spot elongaation comes from
f
the air wavelength
w
diispersion (no narrow bandp
pass filter wass
used). The figgures show that 0.6 arcsec is
i clearly resolved despite of
o the seeing (which was goood that day).
1.2 arcsec
Figuure 8: Double sttar HIP117695 recorded
r
with IDOT
I
1.5
X: -0.3162
Y: 1
1
1
0.8
0.6 arcsec
Normalized intensitiy
Y [arcsec]
[
]
0.5
0
-0.5
-1
-1.5
-1.5
X: 0.2641
Y: 0.9811
X: -0.01341
Y: 0.7779
0.6
0.4
0.2
-1
-0.5
0
X [arcsec]
0.5
1
1. 5
0
-1.5
-1
-0.5
0
0.5
Double star axis [arcsec]
1
1.5
Figuure 9: Double sttar HIP117646 recorded with IDOT:
I
picture and
a intensity prrofile.
5. POINTING
G AND TRA
ACKING TE
ESTS
For the pointing tests, imagges of 30 starrs spread all over
o
the field of
o regard are recorded.
r
Thee telescope uses its pointingg
wards the seleected stars. Foor this test no pointing
p
correection is applied before reco
ording the starr
model to drivve the axes tow
images. The centroids
c
of thhese stars are computed. Thhe centroids dispersion
d
in rooot mean squaare gives the pointing
p
errorr.
The main porrt pointing tesst result is givven in Figure 10. The poin
nting axes of the
t star samplle and the disspersion of thee
centroids are shown. The pointing
p
error RMS is < 1.2 arcsec RMS for a requirem
ment of 2 arcseec RMS.
Proc. of SPIE Vol. 9906 99064E-8
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 08/08/2016 Terms of Use: http://spiedigitallibrary.org/ss/termsofuse.aspx
90
Main Port Pointing test 29/11/2015
Spec'. 2 eresec RMS, Achieved: I.1939eresec RMS
2
80
120
60
*
15
60
..
150
30
Va."'
+
D.5
co
_
180
0
U
W
-a5
++
210
+
330
-1 5
-
240
300
2
270
0
0.5
1.5
1
2
2.5
RA axis laresec]
Figure 10: Main port pointing test; at the left the star sample in the telescope field of regard, at the right the dispersion of the star
centroids while sighting them with the telescope.
The tracking error is specified in open and closed loop of the guider. The tracking error is measured in recording the star
image on a CCD camera in the telescope focal plane while tracking. The statistics on the movements of the centroids
give the tracking error. Centroid positioning were measured with integration time of 30 sec. The measurements were
repeated several times for covering the particular axis movements that appear in the field of regard of the telescope.
Close to zenith, azimuth and rotator axes accelerate and rotate larger angles in a given lapse of time; close to Polaris the
axes are nearly still or can change their moving direction. Close to horizon the seeing becomes worst, good
measurements are then difficult to catch. Except for measurements close to horizon the tracking requirements were met.
The discrepancy at horizon is obviously due to the higher seeing present there. Typical test results in open and close loop
tracking are given in figure 11 and 12.
Open Loop Tracking test - Specification analysis
Filename: 20151201 000200 Tcam MP OL
Telescope trajectory in Azimuth - Altitude coordinates
Measure dater 2015_12_1
A Start
- Trajectory
RMS on 1 min - p +2o: 0.062
spec on 1 min
Peak on 15 min - max: 0.22
spec on 15 min
0.5
- Limits
---
0.4 -
yN
0.3
`m
+;)
R
0.2 -
0.1
00:10
00:15
00:20
00:25
Time
Figure 11: Open loop tracking test; the star trajectory while measuring at the left, at the right the plot in blue gives the RMS error for
data acquired during 1 minute, the plot in green is the peak deviation of the centroid for sliding elapsed time of 15 minutes.
Proc. of SPIE Vol. 9906 99064E-9
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 08/08/2016 Terms of Use: http://spiedigitallibrary.org/ss/termsofuse.aspx
Closed Loop Tracking test - Specification analysis
Filename: 20151202 234800 Tcam MP CL
Telescope trajectory in Azimuth - Altitude coordinates
Measure date: 2015_12_2
N
RMS on 1 min - mean: 0.054
RMS on 1 hour - mean: 0.083
specification
Start
- Trajectory
- Limits
i
0.1
Ú
N
U
L
I
L
Il
Ilit
rf---
AI
A
E
V
(0
LL
I
0.05
1
S
o
00:00
00:10
V
II
1
y
00:20
V
00:30
00:40
00:50
Time
Figure 12: Close loop tracking test; the star trajectory while measuring at the left, at the right the plot in blue gives the RMS error for
data acquired during 1 minute, the plot in green is the RMS deviation of the centroid for sliding elapsed time of 1 hour.
6. CONCLUSION
IDOT commissioning and acceptance test campaign happened successfully despite of the difficult context. Indeed, the
specifications of the telescope are written such that the performance is limited by the seeing and the customer is willing
an end to end acceptance test. The first difficulty is to transform theoretical specifications into measurable parameters in
observing conditions. These parameters have to be sufficiently reliable despite of the seeing and in accordance with the
customer final need. The image quality requirements expressed in Encircled Energy was transformed in wavefront error
specification. The customer who are astronomers are not used to work with this criteria and would like to get an
independent way to prove the image quality with a measurement related to image resolution. Fortunately some good
seeing nights enabled to demonstrate the accordance of what was measured with the wavefront sensor and the telescope
quality in acquiring double star images.
It has to be reminded that a telescope is a complex machine with some very accurate sub-systems and with finely tuned
controls. The obtained quality is the result of an attentive work on every sub-system.
Before sending the telescope on site, it was completely mounted, tuned and tested in the factory integration hall (2).
Despite of the fact that the hall seeing was bad (2 arc seconds in the best case), the debugging and experience acquired at
that time were very precious on site. Time was spared and all the attention could be put on the fine tuning and the
alignments that are necessary for reaching the specifications.
IDOT is the first 4-m class telescope equipped with an active primary mirror realized by AMOS; its success is very
encouraging for the new projects granted to AMOS; DKIST Primary mirror cell(6) and the Doğu Anadolu Gözlemevi
(DAG) telescope(7) that uses this demonstrated technology.
7. ACKNOWLEDGEMENT
This work has been performed under ARIES contract reference 1985-14-02. AMOS is very grateful towards ARIES
team for having put their confidence in AMOS team for the design and manufacturing of the 3.6 m telescope.
Proc. of SPIE Vol. 9906 99064E-10
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 08/08/2016 Terms of Use: http://spiedigitallibrary.org/ss/termsofuse.aspx
At the end of the commissioning and test campaign, ARIES large field camera was installed at the focus of the telescope.
This gave us (ARIES and AMOS team) the opportunity to acquire a first “astronomic” image. Thanks for having spent
this moment together. The image is presented in Figure 13 and was acquired during the first working night of the
telescope Imager instrument (which was thus not tuned nor aligned at that time). Seeing during this night was quite bad
(~1.5 arcsec).
Figure 13: Crab Nebula; composition of 2 images (R and V bands); acquired with IDOT and ARIES large field camera.
8. REFERENCES
[1] Ninane N., Flebus C. and Kumar B., "The 3.6 m Indo-Belgian Devasthal Optical Telescope: general description",
Proc. SPIE 8444-67 (2012)
[2] Ninane N., Bastin C., Deville J., Michel F., Pierard M., Gabriel E., Flebus C. and Omar A., "The 3.6 m Indo-Belgian
Devasthal Optical Telescope: assembly, integration, and tests at AMOS", Proc. SPIE 8444-102 (2012).
[3] Deville J., Bastin C. and Pierard M., "The 3.6 m Indo-Belgian Devasthal Optical Telescope: the hydrostatic azimuth
bearing", Proc. SPIE 8444-150 (2012)
[4] Pierard M., Schumacher J.M., Flebus C. and Ninane N., "The 3.6 m Indo-Belgian Devasthal Optical Telescope: the
active M1 mirror support", Proc. SPIE 8444-186 (2012)
[5] Gabriel E., Bastin C., Piérard M., "The 3.6 m Indo-Belgian Devasthal Optical Telescope: the control system", Proc.
SPIE 8451-82 (2012)
[6] Gregory P. Lousberg, Vincent Moreau, Jean-Marc Schumacher, Maxime Piérard, Aude Somja, Pierre Gloesener,
Carlo Flebus, “Design and analysis of an active optics system for a 4-m telescope mirror combining hydraulic and
pneumatic supports”, Proc. SPIE 9626-24 (2015)
[7] Grégory P. Lousberg, Emeric Mudry, Christian Bastin, Jean-Marc Schumacher, Eric Gabriel, Olivier Pirnay, Carlo
Flebus, “Active optics system for the 4m telescope of the Eastern Anatolia Observatory (DAG)”, Proc. SPIE 9912223 (2016)
Proc. of SPIE Vol. 9906 99064E-11
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 08/08/2016 Terms of Use: http://spiedigitallibrary.org/ss/termsofuse.aspx