CHIMERA: a wide-field, multi-color, high-speed photometer at the prime focus of the Hale telescope

1.8 MB·

by  Leon K. Harding

Mon.Not.R.Astron.Soc.000,1–15(0000) Printed14January2016 (MNLATEXstylefilev2.2)
CHIMERA: a wide-field, multi-color, high-speed
photometer at the prime focus of the Hale telescope
L. K. Harding,1,2(cid:63) G. Hallinan,1 J. Milburn,1 P. Gardner,1 N. Konidaris,1
6 N. Singh,1,2 M. Shao,2 J. Sandhu†,2 G. Kyne,1 and H. E. Schlichting3
1
0 1Cahill Center for Astronomy & Astrophysics, California Institute of Technology, Pasadena CA 91125, USA
2 2Jet Propulsion Laboratory, California Institute of Technology, Pasadena CA 91109, USA
3Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, MA 02139-4307, USA
n
a
J
Accepted2016Jan11.Received2016Jan06;inoriginalform2015Jun09
3
1

ABSTRACT
]
M The Caltech HIgh-speed Multi-color camERA (CHIMERA) is a new instrument that
hasbeendevelopedforuseattheprimefocusoftheHale200-inchtelescope.Simulta-
I
. neous optical imaging in two bands is enabled by a dichroic beam splitter centered at
h 567 nm, with Sloan u(cid:48) and g(cid:48) bands available on the blue arm and Sloan r(cid:48), i(cid:48) and z s(cid:48)
p
bandsavailableontheredarm.Additionalnarrow-bandfilterswillalsobecomeavail-
-
o able as required. An Electron Multiplying CCD (EMCCD) detector is employed for
r both optical channels, each capable of simultaneously delivering sub-electron effective
t
s readnoiseundermultiplicationgainandframeratesofupto26fpsfullframe(several
a 1000 fps windowed), over a fully corrected 5 5 arcmin field of view. CHIMERA was
[ ×
primarily developed to enable the characterization of the size distribution of sub-km

1 Kuiper Belt Objects via stellar occultation, a science case that motivates the frame-
v rate,thesimultaneousmulti-colorimagingandthewidefieldofviewoftheinstrument.
4 Inaddition,italsohasuniquecapabilityinthedetectionoffaintnear-Earthasteroids
0 and will be used for the monitoring of short duration transient and periodic sources,
1 particularly those discovered by the intermediate Palomar Transient Factory (iPTF),
3 and the upcoming Zwicky Transient Facility (ZTF).
0
. Key words: instrumentation: detectors – instrumentation: photometers – methods:
1
0 observational – techniques: photometric – occultations
6
1
:
v
i 1 INTRODUCTION starwillpartiallyobscurethestar.Forsub-km-sizedKBOs,
X
diffractioneffectsbecomeimportantandthedurationofthe
r The Kuiper Belt consists of a disk of icy bodies located occultation is approximately given by the ratio of the Fres-
a beyond the orbit of Neptune. Determining the abundance,
nel scale to the relative velocity perpendicular to the line
materialproperties,andcollisionalprocessesofsub-km-sized
of sight between the observer and the KBO. The Fresnel
KuiperBelt Objects (KBOs) is important, since these bod- (cid:112)
scale is given by λ/2·a ∼ 1.3 km, where a ∼ 40AU is
iesprovidethelinkbetweenthelargestKBOsandthedust-
the distance to the Kuiper Belt, and λ ∼ 600 nm is the
producing debris disks observed around other stars. Obser-
wavelength of the observation. Since the relative velocity is
vations and theory suggest the existence of a break in the
usually dominated by the Earth’s velocity around the Sun,
power-law size distribution at smaller KBO radii (Schlicht-
whichis30 km sec−1,typicaloccultationsonlylastoforder
ing et al. (2012), and references therein). The break in the
of a tenth of a second.
size distribution is generally attributed to collisions that
break-upsmallKBOs(radiusr(cid:54)10km)andthattherefore
modifiestheirsizedistribution.Importantly,thisinterpreta-
To date, only a very small number of high significance
tion has not been tested observationally, as KBOs <10 km
detections of KBO occultations have been reported. Most
in radius are too faint to be in detected in reflected light.
notably, 31,500 star hours of archival data from the Hub-
However,suchobjectscanbedetectedindirectlybystel-
ble Space Telescope (HST) Fine Guidance Sensors (FGS)
laroccultations.AsmallKBOcrossingthelineofsightofa
recordedwith40Hzsamplingfrequency,wasusedtosearch
for stellar occultations by small KBOs (Schlichting et al.
(cid:63) InstrumentScientist;E-mail:[email protected] 2009, 2012). In order to increase the number of detec-
† PrincipalInvestigator;E-mail:[email protected] tionssignificantly,ground-basedsurveysforsimilarocculta-
2 L. K. Harding et al.
tion events are required. However, such a program presents ditionally, photometers and other similar instruments have
unique observational challenges, specifically: used CCD1 or SCD1 sensors, or in some cases CMOS1
or MCP1 architectures (e.g. Dhillon et al. (2007); Law et
(i) Statistics from HST data suggest that a large num-
al. (2006); Matuszewski et al. (2010); O’Donoghue (1995);
ber of stars need to be monitored simultaneously in order
Stover & Allen (1987); Wilson et al. (2003), and references
to achieve a sufficient number of star hours to significantly
therein).CCDshavebeenthepreferreddetectorforthefocal
improve on the existing sample.
planes of telescopes or indeed for photometric instruments;
(ii) Unlike the KBO occultations found in HST FGS
however,theycansufferfromlargeamountsofoutputampli-
data, robust means to rule out false positives caused by at-
fier noise, especially at higher read out rates. CMOS noise
mospheric scintillation noise is essential for ground-based
can be more difficult to characterize than CCD noise be-
searching for KBO occultations.
causeofadditionalpixelandcolumnamplifiernoise,aswell
(iii) Ahighsamplingfrequencyof40Hzisrequiredsuch asnon-linearitiesinthecharge-to-voltageconversions(Holst
that one can adequately resolve the diffraction pattern of & Lomheim 2011). With the advent of the EMCCD at the
the occultation event. beginningofthelastdecade(Jerrametal.2001),extremely
lownoise(<1e−rms),high-speedclocking,hasbecomepos-
CHIMERAwasconceived,designedandbuilttosearch
sible.ThearchitectureofanEMCCDisverysimilartothat
for sub-km sized KBOs in the outer solar system, by detec-
ofaCCD.ThedifferenceliesintheEMCCD’sso-calledhigh
tion of stellar occultations over a wide field of view (FOV).
gain, or electron multiplication (EM) register, which is an
Ourdesignwasoptimizedtoaddressthespecificchallenges
additional stage containing a large responsivity output. In
outlined above, in that:
thisregion,electronscanbeamplifiedbyaprocessknownas
(i) Thenumberofstarsthatcanbesimultaneouslymon- avalanchemultiplication.Theresultisamuchhighersignal
itored for KBO occultations scales with the areal FOV. to noise ratio (S/N), albeit at the proportional reduction
We use the largest format Electron Multiplying CCDs in pixel charge capacity. We have taken advantage of the
(EMCCDs) currently available (1024 × 1024 pix; see Sec- progress in EMCCD detector development to allow us to
tion 4), which, together with the additional requirement deliver 40Hz imaging across our full field with an effective
of adequate sampling of the seeing-limited point spread read noise of <1 e− rms under EM gain.
function (PSF) for the best seeing observed at Palomar
Adetaileddiscussionoftheopticalandmechanicalde-
(∼0.7 arcsec), limits our FOV to 5 × 5 arcmin. A future
signandconstruction,aswellasthedetectorsandsoftware
generationofCHIMERAisplannedtobedevelopedinpar-
development, and instrument performance, involved in de-
allel with larger format (>4K × 4K) EMCCD and sCMOS
livering the CHIMERA instrument to the specifications de-
sensors and will thus increase the areal FOV substantially.
scribed above are presented in Sections 2 – 6. Examples of
Nonetheless, even with the existing FOV, CHIMERA can
firstlightdataandabriefdescriptionofCHIMERA’sfuture
monitor thousands of stars simultaneously with sufficient
plans are described in Section 7.
signal to noise in a 25 ms integration to search for the sig-
natureofaKBOoccultation(seefirstlightimage,Figure9,
right). This is achieved by targeting dense star fields in re- 1.1 Additional Key Science
gionsoftheskywheretheeclipticandgalacticplanesover-
The unique capabilities of CHIMERA, as designed to tar-
lap.
get KBO occultations, are also well suited for a wide swath
(ii) Accounting for false positives due to extreme scintil-
of additional high-time resolution astrophysics. Indeed, the
lationisperhapsthemostdifficultchallengefacingground-
utilityofmulti-color,high-speedphotometershasbeenam-
based KBO occultation detection efforts. The Transneptu-
ply demonstrated by the ULTRACAM instrument (Dhillon
nian Automated Occultation Survey (TAOS II) will make
et al. 2007). CHIMERA can be used to monitor short du-
use of three 1.3m telescopes, requiring simultaneous detec-
ration transient and periodic sources, such as aurorae on
tion in each telescope to rule out scintillation (Lehner et
browndwarfs,eclipsingbinaries,flaringstars,pulsingwhite
al.2014).Byavailingofthe5.1mapertureoftheHaletele-
dwarfs, and transiting planets, and will be well placed for
scope,CHIMERAwillexhibitlowerscintillationnoise(func-
follow-up observations of short-duration transients discov-
tion of D−2/3) than searches on small telescopes. Further-
eredbytheintermediatePalomarTransientFactory(iPTF),
more, the dense fields chosen for KBO occultation searches
and the upcoming Zwicky Transient Facility (ZTF).
will help us further in this regard, in that a large reduction
Additionalkeysciencecaseswillalsotakeadvantageof
inthescintillationnoisepowercanbeachievedthroughdif-
CHIMERA’swidefield,particularlysearchesfornearEarth
ferentialphotometryusinggroupsofnearbystars(Kornilov
asteroids (NEAs). To date, over 10,000 NEAs have been
2012). More important, however, is CHIMERA’s ability to
discovered where 1,000 were >1 km in size (see Kaiser et
conduct simultaneous imaging in two photometric bands.
al. (2002); Larson et al. (2006)). Much like the detection
The diffraction pattern signature caused by the occultation
of small mass KBOs, the detection rates for small NEAs
of a star by a KBO has a specific wavelength dependence
(<50 m) have been largely unsuccessful, with as much as
that will allow us to differentiate between true occultation
98%oftheestimatedhalfamillion50m-classNEAsnotyet
eventsandfalsepositivesduetoextremescintillationevents,
discovered(Shaoetal.2014).Findingandtracking5−10 m
thelatterofwhichwillnotdisplaythesamecolor-dependent
effect (Kornilov 2011).
(iii) The readout speed and noise characteristics of a de- 1 Charge-Coupled Devices (CCD); Segmented Charge-Coupled
tectorcapableofdeliveringfull-frameimagingat40Hzplays Devices (SCD); Complementary metaloxidesemiconductor
a key role in sensitivity to KBO occultation events. Tra- (CMOS);Microchannelplate(MCP)
CHIMERA: Caltech HIgh-speed Multi-color camERA 3
Field lens
r'/i'/z' filter Dichroic Baffle + field mask
Exit pupil ~location
f/3.5 beam
Vixen
380-mm f/3.8
D Canon
C
C2200-mm f/2
d 23
eF u'/g'
RE Collimator
filter
(f=380 mm)
200-inch field corrector
Canon
200-mm f/2
EF232
Blue CCD
Figure 1.SectionviewandraytraceoftheCHIMERAopticallayout.ThebeamentersfromtheHaletelescope(right),passesthrough
the200-inchfieldcorrector,thefieldlens,fieldbaffle(andstop),entrancebaffle,Vixencollimator(mountedbackwards),and45◦dichroic
beamsplitter.Atthedichroic,thebeamissenttoeithertheblueorredside.ThesepathsarebothservicedbyaCanon200-mmf/2.0
lens and EMCCD. Custom filter exchangers are placed in the collimated beam before the Canon lens, where the blue side currently
housesSloanu(cid:48) org(cid:48) bands,andtheredsideeitheroftheSloanr(cid:48),i(cid:48) orz s(cid:48) bands.
NEAs presents a major challenge due to how dim they are quality (typically ∼1.2 arcsec median seeing at Palomar,
(H=28−30mag),andalsobecausetheyalsoexhibithigh yielding a pixel scale of ∼0.3 arcsec pix−1).
apparentpropermotion(0.14arcsecsec−1at0.1AU).Tradi- To meet these requirements, we needed to consider the
tionalsurveysoperatingat>30secweremuchlesssensitive challenges associated with placing an instrument at the
to fast moving NEAs. The smallest NEAs discovered have primefocusoftheHale200-inch.Wesoughtcommercialoff-
typicallybeenlostthereafter,andifsufficientlyaccurateas- the-shelf (COTS) optics that were well-matched to deliver-
trometric data are not collected, future positions (4−6 yr) ing a wide FOV at the 200-inch prime focus. Consequently,
when passing close to Earth cannot be calculated. thedecisionwasmadetodesignacollimator-camerasystem
AnewtechniquehasbeenrecentlydevelopedattheJet and place CHIMERA behind the Wynne corrector (Wynne
PropulsionLaboratory(JPL),synthetictracking,whichwas 1967), which provides a 25 arcmin optically flat field. This
designedtofindverysmallNEAs(H=30mag)–seeShaoet design is illustrated in Figure 1, and described herein.
al.(2014)forfulldetails.AlthoughtheS/Nofasingleshort Although the collimated beam size of 100 mm was
exposureisinsufficienttodetecttheseobjectsinoneframe, too large for COTS dichroic and filter components, the
byshiftingsuccessiveframeswithrespecttoeachotherand CHIMERA collimating and re-imaging optics were success-
subsequently coadding in post-processing, a long-exposure fullydesignedusingCOTSelements.Thecollimatorconsists
image is synthetically created. The resulting frame appears of a Vixen 380 mm focal length astrograph telescope from
as if the telescope were tracking the NEA with >>S/N. Vixen Optics,2 which is mounted backwards in the instru-
CHIMERA’s performance, in conjunction with the ef- ment. Vixen provided an encrypted optical prescription via
fectivenessofthesynthetictracking technique,isanefficient the Zemax3 “black box” model system. We designed a cus-
NEAdetectionmachine.Automatedditheringroutines,dis- tom dichroic beam splitter which was procured from Cus-
cussed later in Section 5, have been specially developed to tom Scientific, Inc4. The re-imaging camera was selected to
conduct blind searches of NEAs over a large fraction of the be the Canon 200 mm lens, based on its large f/2.0 aper-
sky each night. The limiting magnitude for CHIMERA is ture,andwassimulatedusingaprescriptionintheJapanese
∼27 mag (see Section 6), and thus the expected yield for patent literature (JP 2011, 253050, A). The Wynne correc-
NEA observations with CHIMERA is predicted to be 8 torwasmodeledfromtheprescriptionofKellsetal.(1998).
NEAs(<10m,H=28mag)and20(H>25mag)pernight, Thesemodelsallowedustosimulate,andaccuratelymodel,
with sizes down to 7 m, which dwarfs the total global yield the entire optical system in Zemax and predict optical per-
of30NEAsyr−1 forH>28mag.Thispredictionhasalready formance,aswellaspupilplacement.Weshowthetheoret-
been demonstrated with the detection of one new low mass icalthroughputfromthesemodelsinTable1,andcompare
NEAwithadiameterof8mandH=29magwithanearly theinstrumentthroughputasmeasuredon-sky,inSection6.
prototypeversionoftheCHIMERAinstrument(Shaoetal.
2014; Zhai et al. 2014).
2.1 Collimator, field lens & baffle system
The Vixen telescope is mounted backwards in order to act
as a collimator, and is designed with a stop at the first el-
2 OPTICAL DESIGN
The CHIMERA science objectives imposed strict require-
2 http://www.vixenoptics.com/
ments on the optical design in order to provide: i) the ca- 3 Zemax is an optical design program;
pability to image simultaneously in two optical bands, ii) a https://www.zemax.com/home
large FOV ((cid:62)5 × 5 arcmin) and iii) seeing-limited image 4 www.CustomScientific.com
4 L. K. Harding et al.
mm unvignetted beam produced by the collimator. Due to
CHIMERA’s internal mechanical constraints, the dichroic
Red EMCCD waspositioned95.25mmfromtheVixenobjective,seeFig-
ure 2. The surface quality and specifications of the dichroic
were constrained based on simulations of the telescope,
Re-imaging EF-232
camera Wynnecorrector,collimatorandcamerasystem.Thisisdis-
Canon EF 200 cussed in more detail in Section 4.
lens
ThedichroicelementconsistsofafusedSilicasubstrate
withopticalinterferencecoatingsonbothsidesoftheglass.
Filters & filter
exchanger Inordertoavoidvignetting,wespecifiedtheunmounteddi-
Blue EMCCD mensions to be 111.0 × 157.0 ± 0.5 mm, with a minimum
Dichroic beam splitter
& exchanger clear aperture of 100.0 × 142.0 ± 0.5 mm. The longer di-
mension was important since the dichroic was designed to
accept and split an incoming beam at a 45◦ angle of inci-
Vixen
collimator dence (AOI), where the central wavelength was at 567 nm.
Collimator, field lens The transmitted beam has a range of 570 – 850 nm with a
& baffle Baffle throughput of 90%, and the reflected beam has a range of
300 – 540 nm, with an identical throughput to its counter-
Field lens
& field stop part. Careful consideration was given to the dichroic beam
splitter’s central wavelengthand cross-overrange of 30nm,
Figure 2. 3D CAD Solidworks render of the CHIMERA in order to ensure >90% dichroic transmission for each of
collimator-camerasystemwithitsouterstructureremoved.This the adjacent g(cid:48) and r(cid:48) filters.
figure illustrates where the optical elements in Figure 1 were in- Asaresultofthelargedimensionsinlengthandwidth,
tegrated in to the instrument’s internal space. The total field of a thickness of 14 mm was specified in order to maintain
the Vixen collimator is significantly larger than what is used in glass stiffness to withstand coating and mounting stresses.
CHIMERA,thereforethedichroicisheldbyablackframecovered
Thiswasalsoimportantfortransmittedwavefrontaccuracy
infelttopreventscatteredlightenteringtheCanonlens.
to achieve λ/2 peak-to-valley across any 100 mm diameter
circle,withintheclearaperture.Surfaceroughnessis<2nm
ement, such that the entrance pupil sits on the apex of the rms,wheretypicalvalueswerereportedtobe0.5nmrmsby
objective, and where the exit pupil is located ∼360 mm in thevendor.ImagequalitywassimulatedinZemaxbyadding
front of the telescope’s focal plane. Since the 200-inch and both spherical, and astigmatism, to the dichroic surfaces.
theVixendonothavethesamepupilposition,weincluded We predicted that even with three waves of spherical plus
a500mmfocallengthf/10 Plano-Convexpositivefieldlens astigmatism,thesystemwouldhostmostlightinatwo-by-
(Edmund Optics #47-397-INK) to rectify this mismatch. twopixelregion(twopixels∼0.58arcsec).Weover-specified
The field lens places the exit pupil at ∼25 mm past the to one wave to meet this performance.
stop of the Vixen collimator. A second baffle is placed on
the optical path 101.6 mm after the field mask, oversized
by 15% to 66.93 mm to avoid vignetting. This 2-element 2.3 Sloan broadband filters
baffle configuration acts to reduce any scattered light from
We designed a custom linear filter exchanger to house
the Wynne corrector, where the reduction from element
CHIMERA’sfilters,whereeachopticalarmprovided4avail-
one was estimated to be 98%. Although an ideal field lens
able slots, see Figure 2. Sloan Digital Sky Survey (SDSS)
wouldplacetheexitpupildirectlyatthestop,thecost/lead
filters (Fukugita et al. 1996) were selected as the primary
time-to-throughput gain was not significant enough to con-
photometric filter set for CHIMERA, where the blue chan-
sideracustomdesign.Thefieldmaskwasoversizedby15% nelhousestheSloanu(cid:48) andg(cid:48) broadbandfilters,andthered
to 42.75 mm with respect to CHIMERA’s full field (corre- channel, the Sloan r(cid:48), i(cid:48) and z s(cid:48) filters. This range matches
sponding to a converging beam size of 37.16 mm), to avoid
the QE response of the CHIMERA detectors well, where
vignetting and to minimize alignment error.
the SDSS atmospheric cut-off at 300 nm lies approximately
Based on the CHIMERA optical requirements, the
where the response of CHIMERA’s silicon-based detectors
images delivered by the field corrector, field lens, colli- approach zero. This cut-off also applies to the z s(cid:48) filter at
mator and telescope, when imaged with an ideal camera
∼920 nm (Figure 3).
are 0.3 arcsec in rms diameter over 95% of the field, av-
CHIMERA’sSDSSfiltersystemwasobtainedfromAs-
eraged over four equal area fields and wavelengths from trodon5,whereweselectedtheirSDSS‘Gen2’filtervariant.
400 – 900 nm. The geometric throughput of the collimator
The Gen2 offers additional spectral separation between the
is reduced by the f/# mismatch between the 200-inch tele- g(cid:48) and r(cid:48) filters, to better avoid atmospheric air glow where
scope (f/3.5) and collimator (f/3.8). The internal fresnel the (OI) 5577 ˚A line occurs. In addition, the z s(cid:48) filter con-
losses were calculated to be ∼15% end-to-end, yielding a
trols the high-wavelength cut-off, rather than the detector.
total throughput of the collimator to be of order 77%. Outofbandblockingisoftheorder(cid:54)0.03%inGen25.Each
filterwasspecifiedat110.0×110.0±0.1mmofstriae-free
2.2 Dichroic beam splitter fused silica, with a thickness of 3.0 ± 0.025 mm. In order
We designed a dichroic beam splitter since an COTS so-
lution was not available that could accommodate the 100 5 http://www.astrodon.com/
CHIMERA: Caltech HIgh-speed Multi-color camERA 5
100 Sloang0
Sloanu0
BlueEMCCDQE
80
Sloani0
Dichroic
reflectance Sloanz0
%) Sloanr0
( 60 RedEMCCDQE
n
o
si
s
mi
s
n
a
Tr
40
Dichroic
transmission
20
0
300 400 500 600 700 800 900 1000
Wavelength(nm)
Figure 3. CHIMERA filter and dichroic transmission curves, as well as the QE response of each detector. The plane sensor window
throughput is also integrated in to the QE curve calculation. Filters curves reflect the Sloan u(cid:48) or g(cid:48) (blue camera), and Sloan r(cid:48), i(cid:48) or
z s(cid:48) (redcamera).Thedichroiciscenteredat567nm,withatransmission-reflectancecross-overof30nmfor>90%response.Theblue
camerasensorwascoatedwiththee2vUV-viscoating,andtheredcamerasensorwiththee2vvis-NIRcoating.
toallowsufficientmountingareainthefilterexchangers,we The Canon lens was simulated using Japanese patent
specifiedaminimumclearapertureof108.0±0.1mm,with ‘JP 2011, 253050, A’. Model glasses that match the refrac-
blackened edges to minimize reflections. Peak transmission tive index, Abbe number and partial dispersions, are used
was measured to be >95% for all filters, and >90% for u(cid:48), in place of glass names. The simulated image quality of the
with1/4-wavepropagatedwavefrontpriortocoatinginad- Canon lens shows image diameters that are nearly 15 µm,
dition to <0.5 arcmin of substrate parallelism with surface which far exceed the image quality of the Vixen and 200-
roughnesses of 1−1.5 nm. inch optical systems. The Canon lens thusly adds only a
The filter exchanger was designed such that the filter fraction of degraded image quality when compared to prior
was not orthogonal to the optical beam, but tilted by 10◦ optics in the system. The re-imaging camera has 15 sur-
topreventstrongghosting.Narrowbandfiltersarecurrently faces with a total throughput of ∼85% when averaged over
beingconsideredtoaidadditionalfocusedscienceobjectives, 400−800nm.TheEF200expectsanentrancepupilonits
e.g. Hα or Na, with BW∼3 nm. internal stop, whereas the 200-inch and Vixen collimator’s
exit pupil is presented ∼25 mm further downstream in the
Vixen. The vignetting is ∼10% on-axis, to ∼40% off-axis.
2.4 Re-imaging lenses Since the pupils on the camera will move as a function of
field position, where a pupil is located at the most extreme
The Canon EF 200 f/2.0 “double Gauss” style lens was position,thefieldsaregeometricallyvignettedbythecamera
used as CHIMERA’s re-imaging camera. Like all double entrance aperture. Therefore, the total geometric through-
Gauss lens systems, the Canon EF 200 f/2.0 employs a putcanvarybybetween90%(on-axis),to60%attheedge
virtual pupil on its stop located inside the lens itself. In of field. The f/2.0 Canon lens provides a 1.9x demagnifica-
CHIMERA’s design, the Canon EF 200 is driven by a real tion, such that final plate scale is ∼3.4 pix arcsec−1.
pupil located ∼100 mm in front of its entrance aperture,
closetothedichroic.Basedonmarketingmaterialprovided
byCanon(2015),thepredictedimagediametersaresmaller
2.5 EF232 focus/aperture control & window
than CHIMERA’s single 13 µm pix; however, this material
assumes that the lens will be driven as designed. We simu- Althoughtheprimecageofthe200-inchoffersawidefocus
lated the performance of the Canon lens with the external rangeforaninstrument(75mmviatelescopecontrolandan
(real) pupil and found that the Canon lens delivers most additional 125 mm mechanically), CHIMERA requires ad-
light into 5 µm over the entire FOV at visual wavelengths. ditionalfocuscontroloneachchanneltoaccountforthedis-
6 L. K. Harding et al.
Table 1. Theoretical throughput of the full optical system, in-
cluding the 200-inch mirror and Wynne corrector, based on Ze-
80/20 extruded
max modeling. The instrumental throughput alone is discussed Plates for
Al 6061 profiles
in Section 6. These models used the optical prescriptions taken X-Y
from Kells et al. (1998), or provided by Edmund Optics, Vixen
Optics,CustomScientific,Inc.,Astrodon,CanonCameras.
Motors
Element Geometric Fresnel Total
200-inch+corrector .91 .85 .77
Vixencollimator .85 .88 .75
Dichroic 1 .90 .90
SDSSfilter 1 .99 .99
Canoncamera .90-.60 .85 .77-.51
TotalCHIMERA .76-.51 .67 .51-.34
Total(telescope+inst) .70-.46 .55 .38-.26
tinctwavelength-dependentvariationinfocusineachband.
Therefore,wealsoincludedaremotelycontrolledfocusand
aperturecontroller,theEF-232,oneacharm,thatisplaced
inbetweentheCanonEF200lensandtheEMCCDsensor.
TheEF-232wasobtainedfromBirgerEngineering7 andof-
fersindependentfocuscontrolforeachchannel.Inthisway,
the Canon EF 200 lens behaves as if it were connected to a
Canon DSLR, and thus has full functionality.
Finally, the last optical element of the CHIMERA
collimator-camera system is the CCD window that acts
to maintain the hermetically-sealed vacuum chamber con-
taining the EMCCD sensor. This is a plane, UV-grade Support posts
Stiffening/shear
fused silica window, approaching 99% throughput. See for Z/tip-tilt
plates
www.andor.com for more details.
H
3 MECHANICAL DESIGN Prime focus
Pedestal interface
3.1 Materials, structure & custom opto-mechanics W
L
The support structure for the instrument was constructed
outofprecipitation-hardened80/20extrudedAluminiumal-
loy profiles (Al 6061). The main structure is a box-frame Figure 4.3DCADSolidworksrenderoftheCHIMERAinstru-
with diagonal supports at the base for strength and stiff- ment.Wehaveremovedtheexternalpanelingforaninternalview
ness, as shown in Figure 4, (left), and facilitated easy ac- of the optical design. The main structure consists of 80/20 ex-
truded Al 6061 profiles and 12.7 mm Al 6061 plates. The colli-
cess to optics and other components. In addition to these
mator, dichroic beam splitter and re-imaging cameras, all have
supports, 12.7 mm Al 6061 plates were used to clamp sec-
mountingdesignsthatallowX,Y,Zandtip-tiltalignment.The
tions of the instrument together; these also acted as stiff-
axes indicator on the bottom left indicates the H × L × W di-
ening plates. We used 3.175 mm anodized Al 6061 covers
mensionsof1.085m,0.968mand0.713m,respectively.
on the box-frame which provided shear support, and also
aided light-tightness. These components ensured structural
stiffness during lateral loading at low telescope elevations.
EF 200, EF-232 and EMCCD camera as one single part,
We imposed a strict stiffness constraint such that the cen-
hereafterre-imager).Indesigningasinglere-imagermount,
tral beam would maintain its position to <1 pix (<13 µm)
thisgreatlyhelpedopticalalignmentandensuredthatonce
for all gravity vectors, including Palomar’s lowest elevation
angle cut-off of 18◦. theprimefocuscentralbeampassedthroughtheapexofthe
Canon EF 200, it would also be aligned with the geometric
NewportandThorlabsopticalpostsprovidedadditional
centeroftheEF-232andhitthedetectoratthecentralX-Y
stiffness,wheretheirprimaryfunctionwastoprovideX,Y,
pixel position, 512, 512. As a result of the 100 mm filter
Z and tip/tilt fine adjustment, for optical alignment – see
optics, a custom 4-position linear exchanger and lead screw
Figures 2 and 4. Custom mounts were designed and fabri-
was also designed to facilitate movement over 4 positions,
cated for the field lens and baffle, the Vixen collimator, the
where custom mounting surfaces, limit/home switches and
dichroic,thefilters,andthere-imagingcameras(incl.Canon
motors were integrated in to the design.
Allmaterialswereorderedasblack,matted,oranodized
7 http://www.birger.com/ where necessary. All internal walls of the instrument were
CHIMERA: Caltech HIgh-speed Multi-color camERA 7
covered with black felt to minimize internal reflections. We
Table 2. Specifications of the CCD201-20 EMCCD, from e2v,
referthereadertoMarshalletal.(2014)foranalysisonthe asdeployedintheAndoriXonUltra888camera.(cid:63)FT=Frame
characteristicsofvariousmaterialsthatcanreducespectral Transfer; ∗BI = Back-Illuminated; ‡IMO = Inverted Mode Op-
reflectance, where they show that black felt is an effective eration.Wealsoshowsomeexamplesofthenativereadnoiseof
material at doing so. The total mass of the instrument is somehorizontalreadoutrates.
∼115 kg, including wiring and power supplies, 0.968 m in
length and 0.713 m in width, see Figure 4. Parameter Specification
Sensortype EMCCD
Variant FT(cid:63),BI∗,2-phase
3.2 Alignment
Activepixels(image) 1024×1024
We aligned CHIMERA by constructing an ∼ f/3.6 optical Pixelpitch 13µm
system,whichconsistedofacollimatedLED(470–850nm, Digitization 16-bit
Amplifiers Conv.orEM
10dBBW)andalens(D=25mm,FL=90mm).AnLED
Calibratedgainrange 1-1000
driver with a drive current range of 200−−1200 mA was
MultiplicationStages 604
usedtocontroltheintensityandneutraldensityfilterswere
Readoutrate(Conv.horiz.) 1,0.1MHz
installed where necessary to avoid saturation. We placed a
Readoutrate(EMhoriz.) 30,20,10,1MHz
50µmpinholeatthe200-inchfocuspoint(113.26mmabove Framerates 26fps(full)–1000fps(sub)
theprimefocuspedestalinterface).Allopto-mechanicalele- Darkcurrent(IMO‡,-85◦ C) 5×10−4 e− pix−1sec−1
mentswereinitiallyalignedtotheirZemaxmodelpositions
by using a FaroArm8, a portable coordinate mapping arm Readnoise(Conv.,1MHz) ∼6e− rms
Readnoise(EM,1MHz) ∼20e− rms
capableofhighprecisionsin3Dmetrology.Theprimefocus
Readnoise(EM,10MHz) ∼53e− rms
pedestal interface was used as a zeroth reference point. Readnoise(EM,20MHz) ∼120e− rms
Once the system was collimated, we assessed the Eff.readnoise(w/EMgain) <1e− rms
FWHMofthespotgeneratedontheEMCCDafterpassing
throughthere-imagingoptics.Thesensor’s13µmpixpitch
coupled with 50 µm pinhole yielded a FWHM of ∼3.85 pix
voltages (>40 V) than the conventional register (∼11 V),
when in focus, which was aligned to the sensor’s central
and can deliver an effective read noise of <1 e− rms with
pixel. Optical performance was compared to Zemax model-
EMgain,whereastheconventionalamplifierprovidesanoise
ing thereafter. Finally, the dichroic element was positioned
of∼6 e− rms.Crucially,asaresultofmuchhigherbiasvolt-
in the central beam which we define as its zeroth position.
agescontrollingthegainviathehighvoltageclock,relatively
Themotor’shighresolution(4000stepsmm−1)wasencoded
low amounts of raw signal can quickly lead to saturation of
to ensure that the dichroic would always return to this po-
pixels in the high gain register. Therefore, there is always a
sition after being moved by the linear exchanger.
trade-off between the desired reduction in read noise and a
reduction in the effective pixel charge capacity. The Andor
iXon Ultra 888 offers a calibrated gain range of 1−1000,
4 DETECTORS: EMCCDS
where the CCD201-20 can demonstrate a maximum well
CHIMERA uses two Andor iXon Ultra 888 EMCCD cam- depth of 80,000 e− in the image section, and 730,000 e−
eras.ThesedetectorshousetheCCD201-20sensorfrome2v, in the gain register. This is necessary to avoid saturation
seeTable2.Thisisaframetransfer(FT)EMCCD,withan duringtheamplificationprocess.Itisimportantnottocon-
active image section of 1024 × 1024 pix and a 13 µm pixel tinuously saturate the device under high amplification as
pitch.Thestore,orFTsection,has1056×1037pix(includ- this can cause irreparable damage to the EM register.
ing dark reference rows) that allows high-duty cycle image Charge is clocked out through the parallel (image and
acquisition,thusmakingtheCCD201-20thelargest-format store section) and horizontal chains, where the iXon Ultra
EMCCD device currently available from e2v, and ideal for 888 offers variable parallel read out speeds of 0.6 – 4.33 µs,
CHIMERA’s applications. and horizontal read out rates of 0.1, 1, 10, 20 and 30 MHz,
TheCCD201-20isa2-phase,thinned,back-illuminated allowing 1000s of fps. The blue camera is coated with the
deviceandhasdualoutputamplifiers:theconventionalam- e2v BUV anti-reflective coating, whereas the red camera
plifierforhighdynamicrangeandtheEMamplifierforhigh- is coated with e2v’s vis-NIR anti-reflective coating, where
sensitivity.Thinningandback-illuminationhelpstoimprove both of the devices are science-grade sensors (e2v’s highest
QE, which peaks at >90% at 550 nm. The power of the cosmetic and performance yield).
EMCCD is in the EM process; however, this is inherently Thedevicesarerunininvertedmodeoperation(IMO),
stochastic, where there is ∼ 1 − 1.5% probability (α) of thus greatly suppressing generation of minority carriers -
an extra electron getting generated per given amplification namely surface dark current. Further dark current suppres-
stage. Since a device can contain hundreds of multiplica- sion is achieved by cooling the device to a stable -85◦ C
tion stages, the probability of amplification becomes signif- by the Peltier effect via three-stage thermoelectric cool-
icant9. The high gain amplifier is driven by much higher ing (TEC). However, we note that at higher frame rates
(>50 fps), the TEC can only maintain stable temperatures
8 Seehttp://www.faro.com/homeformoreinformation
√
9 EM = (1+α)N, where N is the number of stages. We note “ExcessNoiseFactor”(ENF),andasymptoticallyapproaches 2
thatthereisanaddedvarianceintheEMoutputwhicheffectively forgains(cid:62)10.Post-readouttechniquescanreclaimthisreduction
reducesthesensor’sQEbyupto∼50%.Thisisreferredtoasthe inQE,seeDaigleetal.(2008).
8 L. K. Harding et al.
CHIMERA RED CHIMERA Dichroic CHIMERA BLUE
Filter mechanism mechanism Filter mechanism P200
Telescope
LexiumMdrive LexiumMdrive LexiumMdrive via TCP/IP
Ethernet Motor Ethernet Motor Ethernet Motor
CHIMERA Simultaneous CHIMERA
Filter
component RED exposure trigger BLUE
P200
P200 Dichroic Filter P200
Telescope
component component component component
via TCP/IP
Remote access
EF232 Andorcamera Andorcamera EF232 By VNC Server
Port forwarding
controller controller controller controller
JAVA process JAVA process
C Library C Library Control
Computer
Custom JNI int. Custom JNI int. -CameraLinkx2
All arrows denote libAndorCamera2.so libAndorCamera2.so -USB 3.0 (x6)
a TCP/IP socket -IRIG-B
AndorSDK 2.98.3 AndorSDK 2.98.3
LantronicsEDS 4100 LantronicsEDS 4100
Terminal Server Terminal Server
USB 3.0/Frame USB 3.0/Frame
Grabber PCIeDriver Grabber PCIeDriver
RS232 COM RS232 COM
CameraLink USB 3.0 USB 3.0 CameraLink
Birger Canon Birger Canon
AndoriXon888 AndoriXon888
Focus/Aperture Focus/Aperture
EMCCD EMCCD
Controller Controller
Figure 5. Block diagram illustrating the CHIMERA data system and software communication logic. The arrows represent a TCP/IP
socket connection, where the large rectangular box on the left represents the full system which is controlled by the CHIMERA control
computer,alsolocatedintheprimefocuscage,showninthesmallrectangularboxtotheright.
at∼-65◦C,yieldinga∼3foldincreaseindarkcurrentfrom software is compatible with any Andor EMCCD camera
-85◦ C. This dark current generation is still sufficiently low variant, the SDK can use either USB communication or a
for most ground-based applications (Dhillon et al. 2007). CameraLinkframegrabberPCIecardforimageacquisition.
Sincethegraphicaluserinterface(GUI)iswritteninJava,a
customJavaNativeInterface(JNI)isrequiredtoaccessthe
AndorSDK.Twoframegrabbercards,eachcommunicating
5 CHIMERA SOFTWARE
withasinglecamerainthecaseoftheiXon888,areinstalled
5.1 Control Software in a single computer. Each instance of the software acts as
aTCP/IPsocketserverandallowsexposecommandstobe
5.1.1 Design & integration with Andor
synchronized between the two cameras at ∼ms timescales.
CHIMERA’s control software uses a combination of Java
(JDK 1.7) and C (gcc, nvcc) and relies on the ‘Netbeans’
5.1.2 Data acquisition & optics software control
integrated development environment (IDE). The software
systemmakesextensiveuseoftheJSkyarchive10 forimage The exposure process operates in a ‘run till abort’ mode
display and image quality analysis. Ephemeris calculations wherethecamerastoresthelatestimageinaninternalbuffer
areobtainedfromtheJSkyCalcprograms11 wherethesoft- (∼90fullframeimages),whichisthenretrievedbythesoft-
ware runs Red Hat Enterprise Linux (version 6.5). ware.ThecamerainternalbufferistreatedasaFIFO(‘first
Thecamerasarecontrolledbyusingthesoftwaredevel- in,firstout’)bufferbycheckingthenumberofavailableim-
opment kit (SDK, v2.98.3) provided by Andor. Since the ages and retrieving the oldest image first.
The dichroic and filter linear exchangers are driven by
LexiumMDriveEthernetNEMA17motors13.Motioncom-
10 http://archive.eso.org/cms/tools-documentation/jsky.html.
11 Provided by John Thorstensen of Dartmouth College,
http://www.dartmouth.edu/∼physics/labs/skycalc/flyer.htm.l 13 SchneiderElectricMotion,USA.
CHIMERA: Caltech HIgh-speed Multi-color camERA 9
mandsandrequestsformotorstatusaresenttothemotors 5.2 The CHIMERA Pipeline
using MCode ASCII strings over the TCP/IP port for each
A data processing pipeline for the purpose of the reduc-
mechanismandacustomcontrolwithinthesoftwareallows
tion and processing of images from CHIMERA has been
eachmechanismtobecontrolledfromwithincameracontrol
developed. The three main functions of the pipeline are as
system.
follows: 1) to calibrate raw data, 2) to generate ‘first look’
The EF232 controller (aperture control of the Canon
data products and 3) to provide detailed analyses tools for
lens)isconnectedtoaLantronicsEDS4100terminalserver.
post-processing.Imagecalibrationinvolvesbiassubtraction,
A command string (ASCII) is sent to the EF232 to query
flat field correction, the flagging of bad pixels or columns
the mechanism state and to command focus and aperture
and field distortion correction. The pipeline offers routines
position. All camera and telescope parameters, the current
togeneratecalibratedphotometriclightcurvesimmediately
state of the filter and dichroic mechanisms, as well as the
after data acquisition is complete. This serves as a useful
focus and aperture state are automatically written into the
tool for observers who wish to assess the behavior of an as-
FITS header.
tronomical target during an observation window. Since we
consider this ‘first look’ result as preliminary, the pipeline
offers additional post-processing packages allowing a more
5.1.3 The control GUI & 200-inch comms detailed analysis of the raw data set.
ThepipelinewasdevelopedinthePythonenvironment
Thesoftwarerunsineither‘TargetingMode’,whereallcam-
usingtheNumpy,Scipy andMatplotlib packagesinorderto
eraparametersmaybemodifiedwhilethesystemisacquir-
provide cross-platform support. Although the routines will
ing images, or in ‘Science Mode’, where the parameters are
not require the PyRAF environment to run, it can easily
fixed at the beginning of a set of exposures. Complete con-
be run in conjunction with IRAF/PyRAF tasks. This work
trol of binning and region of interest allow the observer to
made use of Astropy, a community-developed core Python
operateinfullframe,asetofpre-configuredbinningmodes,
package for Astronomy (Astropy Collaboration 2013), for
and/orsub-framedmodes.Italsosupportscustombinning.
source detection and aperture photometry. Among other
The feature set support by the CHIMERA software is
utility routines, the pipeline also includes routines to gen-
extremely extensive and therefore we only cover the main
erateanimationfromdatacubes,andphasefoldingoflight
features here, but refer the reader to the ‘CHIMERA Users
curves. An initial version of the pipeline has been released
Manual’ that can be found on the CHIMERA website for
on the CHIMERA website (see footnote 14) as well as on
moreinformation14.Thesoftwareisfullyintegratedintothe the github15 version control system.
200-inch telescope control system. The telescope telemetry
ispolledat∼10Hzandasbefore,allrelevanttelescopeposi-
tioninformationiswrittenintotheFITSheaders.Telescope
controlisintegratedintotheimagedisplaysystemallowing
6 INSTRUMENT PERFORMANCE
observerstore-pointthetelescopebyselectingpositionson
the image display, or by using simulated paddles. A local 6.1 Noise Characterization
copy of the UCAC3 catalog (Zacharias et al. 2010) is inte-
Evaluating the noise in an EMCCD is similar to that of a
grated into the software and may be queried and displayed
conventional CCD; however, as noted in Section 4, if the
both in tabular form, and as an image overlay.
highgainregisterunderEMamplificationisused,thereisa
gain factor applied which results in a reduction of the read
noise, σ , by an amount R/G, where R is the number of
RN
5.1.4 Dithering & guiding electrons and G is the amplification gain. Additionally, the
ENFproducesanaddedvarianceintheoutputsignaldueto
Asophisticatedditherandmappingcontrolisincludedthat
the stochastic nature of the EM process. As a result, when
allowsobservingpatternstobeexecutedeasily,e.g.10×10
calculating the S/N for an EMCCD, σ becomes negliga-
RN
grid of the 5 × 5 FOV on sky, fully automated, including
blebuttheENFandmultiplicationgainmustbeconsidered
exposure triggering at each settle point and a halt at each
in addition to other paramters such as the clock induced
slew.Theditherandmappingcontrolcontainsanintegrated
charge (CIC) which can dominate under high gain condi-
visualization system that overlays the observing pattern on
tions. CIC is another source of spurious noise caused when
anSDSSimage,andupdatesthestate(bycolor)ofthepat-
a CCD is clocked into inversion, and since the Andor iXon
tern during execution of the observation. The CHIMERA
888EMCCDcameraisruninIMOtosuppresssurfacedark
software also supports high precision guiding using multi-
current, CIC is greater than a device run in non-inverted
ple guide objects based upon the code developed for plane-
mode (NIMO) and must be included.
tary transit photometry with the WIRC instrument (Zhao
In order to characterize CHIMERA’s noise perfor-
et al. 2012). The guide system uses multiple objects to sep-
mance, which includes photon, sky and detector noise, we
arate motion due to seeing fluctuation from the mechanical
calculated the total theoretical noise of the instrument and
driftofthetelescope.Thesystemachievesguidestabilityof
comparedthistothemeasurednoise,whichweshowinFig-
±0.25 arcsec (<1 pix) using this system. We show a block
ure6asanσ vs.magnitudeplot.Wenotethatthetheo-
mag
diagram of the CHIMERA software in Figure 5.
reticalmodelandmeasuredresultswerecalculated/obtained
14 http://www.tauceti.caltech.edu/chimera/ 15 https://github.com/caltech-chimera/pychimera
10 L. K. Harding et al.
Figure 6.σmag vs.magnitudeplotofthetheoreticalvs.measurednoiseperformanceoftheCHIMERAinstrument.Theinstrumental
magnitude of each star in the plot is −2.5·log10(flux). The measured data (black data points) were obtained using the EM amplifier
with an amplification gain of 100. We used a KBO field at α = 18 : 40 : 49.92 and δ = −06 : 46 : 59.8 since it provided hundreds of
stars for analysis. The following parameters were included in the calculation: photon noise, sky noise, read noise (with EM gain), dark
current,clockinducedcharge(CIC),andtheENF.SeeEquation1andEquation2,whichshowthemethodsusedforthiscalculation.
using the EM amplifier, and included the following param- The measured instrumental magnitude of each star in
eters: photon noise, sky noise, read noise (with EM gain), Figure6is−2.5·log (flux)(x-axis)andthestandarddevia-
10
dark current, CIC, and the ENF. We calculated the total tionfromthemeanmagnitude(y-axis)wasfoundasfollows:
theoretical S/N as follows:
σ =1.0857·N/S (2)
mag
N ·k
S/N = star (1)
(cid:112)
σs2ource+σs2ky+σd2ark+σR2N +σC2IC We obtained data from a moderately dense KBO field
at α = 18 : 40 : 49.92 and δ = −06 : 46 : 59.8 in order to
where:
show noise performance in the KBO high cadence regime.
These data were taken at 33 Hz in 2 × 2 binned mode on
– N is the stellar flux in ADU.
star Aug 15, 2015, using the Sloan i(cid:48) band, and were bias sub-
– k is the conversion gain in e− ADU−1.
√ tracted and flat fielded. The data points in Figure 6 show
– σ = ENF2·N ·k, which is the target shot
source star 304starsthatweredetectedat10σ abovebackgroundinan
noise.
(cid:112) averaged image using the IRAF DAOPHOT daofind rou-
– σ = ENF2·N ·k·n , which is the sky noise,
sky sky pix tine (Stetson 1987), where the theoretical performance of
where N is sky flux in photons and n is the number
sky pix the system (calculated using Equation 1) is shown as the
of pixels.
(cid:112) red data points and the measured performance is shown as
– σ = ENF2·q ·t ·n , which is dark cur-
dark dark exp pix the black data points. Dark current was measured to be
rent, where qdark is the dark current in e− pix−1 sec−1 and 3 × 10−4 e−1 pix−1 sec−1 at-85◦ CandCICwasmeasured
texp is the integration time in seconds. to be 8 × 10−3 e−1 pix−1 transfer−1 at a 20 MHz read out
(cid:112)
– σRN = (R/G)2·npix, which is the read noise. rate. Since the read noise was found to be ∼130 e− rms at
(cid:112)
– σ = ENF2·q ·n , which is the CIC, where 20 MHz, we used an EM gain of 100 to achieve an effec-
CIC CIC tr
q istheCICperpixelandn isthenumberoftransfers. tive read noise of ∼1.3 e− rms. In this example, we used
CIC tr
standard aperture photometry; however, for denser fields,