3kk – The Optical-NIR Multi-Channel Nasmyth Imager forthe Wendelstein Fraunhofer Telescope

Florian Lang-Bardla, Klaus Hodappb, Shane Jacobsonb,Ralf Bendera,c, Claus Gossla, Maximilian Fabriciusa,c, Frank Gruppc, Ulrich Hoppa,c,

Wolfgang Mitscha

aUniversitats-Sternwarte Munchen, Scheinerstr. 1, D81579 Munchen, Germany;bUniversity of Hawaii, Institute for Astronomy, 640 N. Aohoku Place, Hilo, HI 96720, USA;cMax Planck Institut fur Extraterrestrische Physik, Gießenbachstrasse, D85748 Garching,

Germany

ABSTRACT

The Ludwig-Maximilians-Universitat Munchen operates an astrophysical observatory on the summit of Mt.Wendelstein1 which will be equipped with a modern 2m-class, robotic telescope.2 One Nasmyth port of the newFraunhofer telescope is designed to deliver the excellent (< 0.8” median) seeing of the site [1, Fig. 1] for a smallerFoV of 60 arcmin2 without any corrector optics at optical and NIR wavebands. Thus, it will be optimized forfast multi-wavelength follow-up observations of targets of opportunities (e.g. Gamma-Ray-bursts) or e!cientphotometric redshift determinations of huge numbers of galaxy clusters identified in optical (PanSTARRS), SZ(Planck) or X-ray (eROSITA) surveys. We present the design of a compact 3 channel camera which serves thesescience requirements, built partly from commercially available Fairchild-2k optical CCD3 cameras (Apogee),coupled with small Bonn Shutters,4 and mounted on commercial high precision linear stages for di"erentialfocusing. A specially designed beam-splitter system maintains the high optical quality. The NIR camera is builtin cooperation with the Institute for Astronomy in Hawaii. The combined operation of this camera with twospectrographs at the same telescope port has already been presented at SPIE 2008.5

Keywords: CCD camera, Beamsplitter, Near Infrared

1. INTRODUCTION

In the era of 8m class telescopes like the VLT’s, or upcoming survey telescopes like PanStarrs, LSST and VISTAthat will produce huge amounts of data, telescopes with apertures from 2-4 m and state of the art instrumentationwill become more and more important. Possible applications for these smaller telescopes and their instrumentsare follow up observations of interesting objects found in the mentioned wide field surveys, long term studies notpossible to schedule on bigger facilities or testing new ideas in instrumentation projects. We also need facilitiesto educate the next generation of astronomers.

In this paper we present the design and current status of an Optical-NIR Multi-Channel Nasmyth Imager(3kk). It will be optimized for fast multi-wavelength follow-up observations of targets of oppurtunities (e.g.Gamma-Ray-bursts) or e!cient photometric redshift determinations of huge numbers of galaxy clusters identi-fied in optical (PanSTARRS), SZ (Planck), or X-ray (eROSITA) surveys.The telescope we will use for the 3kk is currently built by Kayser-Threde GmbH for the observatory of theUniversitats-Sternwarte on the summit of Mt. Wendelstein in the Bavarian Alps.2 The instruments will bemounted to two Nasmyth ports equipped with derotators. There is one wide field port with a corrector systemto deliver a field of view (FOV) of 0.7! diameter and a smaller port without any corrector optics with a FOVof 0.14!. The plate scale of the small port is 74 µm per arcsec. The camera mounted on the wide field portis a mosaic of four 4kx4k e2v CCD’s built by Scientific Instruments in Tucson.6 The uncorrected port will beequipped with the follwing instruments:

Further author information: (Send correspondence to F.L.-B.)F.L.-B.: E-mail: flang@usm.uni-muenchen.de, Telephone: +49 89 2180 6965

Figure 1. Left: Layout of the optical part of the 3kk. Beamsplitter 1 (BS1) reflects wavelengths shorter than 695 nm,beamsplitter 2 (BS2) all wavelength below 970 nm. The remaining light is transmitted to the near infrared channel. Thefully transparent corrector plate in front of the red optical channels corrects for aberrations introduced by BS1. Right:Design of the near infared part of the 3kk. After the entrance window, light is reimaged using six lenses without changingthe plate scale. A cold stop masking the obstruction of the secondary mirror and filters are placed near the pupil image.The design is folded two times to keep the dimension of the camera small.

WWAVE: A Shack-Hartmann sensor for telescope alignmentVIRUS-W: An IFU spectrograph7

FOCES: A pressure and temperature controlled Echelle spectrograph8

3kk: Optical-NIR Multi-Channel Nasmyth Imager

The instruments and mechanism to change between the di"erent focal stations were already presented at theSPIE conference 2008 in Marseille.5 To calibrate the spectrographs and as reference source for WWAVE, atelescope simulator with di"erent calibration lamps will also be mounted on the port.

2. 3KK OPTICAL LAYOUT

The optical design of the instrument was driven by the following requirements:

Spot quality: 80% encircled energy within two pixels inside a radius of 0.07! FOV (corresponds to the largestused detector in the infrared channel).Field distortion: Geometric field distortion should be less than 0.2 pixels over the whole field in the opticalchannels. (Corresponds to 0.014% maximum relative distortion)Low ghost intensity: The corresponding ghost of a bright point source should be below 1E-4 of the sourcescentral intensity. It should also be well seperated.

The optical layout of the visual channels is shown on the left in figure 1. Light seperation for the di"erentchannels is realized with two beamsplitters. Because the plates placed under 45! in the beam introduce abber-ations to the system, we followed the idea of Woche et. al.9 and use wedge shaped and not coplanar plates.Light comming from the telescope hits beamsplitter 1 (BS1), that reflects wavelength shorter than 695 nm whilebeamsplitter 2 (BS2) reflects wavelength up to 970 nm. Longer wavelengths are transmitted to the near infraredchannel. The corrector plate in front of the red optical detector is anti reflection (AR) coated on both sidesand corrects the aberrations introduced by BS1. The reflection bands, angles of incident (AOI), thicknesses andwedge angels of the plates are given in table 1. After passing the beamsplitters and corrector plate each opticalchannel consists of a filter, a flat entrance window to the camera and the detector.

The design of the infrared part is shown on the right of figure 1. It was also important for us to keep thedistortion in this channel as low as possible over the FOV (see figure 2).

Table 1. Transition Bands, AOI, wedge angle, thicknesses of the beamslitters, and the corrector plate. The corrector plateis AR coated on both sides and corrects the aberrations introduced by BS1 in the red channel.

Plate Reflection range AOI thickness wedge angle(nm)

Beamsplitter 1 340 - 695 45! 9 mm -694""

Beamsplitter 2 695 - 970 !45! 9 mm 1064""

Correctorplate full transparent !35! 9 mm -1064""

After passing the entrance window there is a reimaging optic that consists of six lenses with cold stop andfilter. The cold stop masks only the secondary mirror of the telescope. The spider structure is neglected, becausethis would imply a rotating cold stop mount. Because of space constraints in the dome we folded the beam twotimes to keep the outer dimensions of the camera small. Both fold mirrors are standard gold coated mirrors fromEdmund Optics GmbH to gain high reflectivity.

In the near infrared we use the standard J, H, Ks and two narrowband (H2, Br!) filters, ordered by BARRAssociates Inc.. In the optical we use the Sloan filterset (u’, g’, r’, i’, z’). These filters were ordered from Lasercomponents (german distributor for Omega Optical Inc.). The distribution of the filters over the channels isshown in table 2.

Table 2. Filters used in the di!erent channels. Given are the cut on and cut o! wavelength (50% flux of the average peaktransmission) of the transmission bands

Channel filter cut on (nm) cut o! (nm)1 u’ 327 383

g’ 409 546r’ 555 692

2 i’ 695 831z’ 850 970

3 Y 972 1071J 1166 1332H 1481 1780Ks 1989 2313Br! 2141 2187H2 2113 2139

Figure 2 shows the calculated spot diagrams and distortion of the red visual and the near infrared channel.80% of the encirceld energy are within a two pixel element over the whole FOV. The distortion on the left isscaled by a factor 10000, on the right by a factor 100.

3. 3KK MECHANICAL DESIGN

The mechanical design of the camera is split into two parts. The first part consists of the two optical channels, thesecond of the near infrared camera (WNIR). As described in section 2 we use two beamsplitters for wavelenghthseperation. The whole optical path up to the CCD entrance windows is operated at ambient temperature andpressure. Contrary, the infrared path is evacuated and cooled to 85K starting with a flat entrance window behindthe second beamsplitter. The detectors in the visible spectrum are commercial camera systems from Apogee

Figure 2. Top: Spot diagrams of the red visual and the near infrared channel. The boxes are equivalent to a two pixelelement of the HAWAII-2RG detector (36µm). 80% encircled energy are within a two pixel element over the whole FOV.Bottom: Distortion of the red optical and near infrared channel. The displacement in the red optical channel is scaledby a factor 10000, in the near infrared channel by a factor of 100

Figure 3. Mechanical design of the optical channels. Both channels are constructed in the same way. The filterwheel infront of the UV enhanced camera is equipped with u’, g’ and r’ filters, the one in front of the broadband coated camerawith i’ and z’ filters.

Instruments Inc.. The Near Infrared camera is a customized system built at the Institute for Astronomy inHawaii.

3.1 Optical channelsFor the optical channels we use two Apogee ALTA E3041 cameras equipped with 2Kx2K CCD’s from FairchildImaging (15µm pixels). The cameras are watercooled and the CCD temperature is regulated with a threestage peltier element. Under maximum power consumption the CCD can be cooled to 60!C below ambienttemperature. At a typical operating temperature of !50! C on Mt. Wendelstein this results in a dark currentof 0.13 e!

s·px for the blue channel camera and 0.02 e!

s·px for the one used in the red channel. The detector ofthe blue channel was ordered with an UV enhanced coating, the one for the red channel is broadband coated.A seven position filterwheel, also from Apogee, is mounted in front of each camera. To optimize the accuracyof photometric measurements we ordered two Bonn Shutters10 with 60 mm free aperture from the ArgelanderInstitut fur Astronomie in Bonn. Apogee camera, filterwheel and Bonn shutter are directly attached to eachother and placed on a linearstage from MICOS for di"erential focusing. To mount the cameras on the linearstagewe designed a tip-tilt stage that also allows lateral movements. The camera adjustments compensate for thefixed beamsplitter positions. In figure 3 you see the current mechanical design of one of the optical channels.Both channels are constructed the same way.

3.2 Near Infrared channelThe mechanical design of the WNIR camera is completed and machining of the parts started in March 2010 at theInstitute for Astronomy in Hawaii. The design of the camera is shown on the left of figure 4. The vacuum jacketconsists of three parts. The middle section of the camera is the supporting structure of the instrument, where allinner components are connected to. Also the threads to mount the camera to the telescope are machined to thispart. The other two components only serve to close the vacuum vessel and have no specific sti"ness requirements.They are therefore designed as thin as possible to save weight. There are two electronic boxes attached to thevacuum vessel. One contains the control electronics of the filter wheel, the other one covers the Teledyne Jade-2Card that provides the interface to the SIDECAR ASIC detector control electronics. The feedthrough of cablesfrom the inside is done with vacuum plugs from Souriau Industrial. For the flex cable connecting the JADE-2card to the ASIC we use an established approach developed at the IFA. The cable is sandwiched between twoaluminum pieces and sealed with epoxy. This part is machined to allow an O-ring sealing with the jacket. Aconvection vacuum gauge to measure the pressure inside the camera and a vacuum valve to evacuate the systemare placed on top of the camera.

Figure 4. Left: Final design of the near infrared camera. Right: Mechanical design of the inner parts of the ear infraredcamera. The optic mounts, filterwheel, detectormount and the LN2 vessels used to cool the components to 85K areattached to a central plate. Trusses made of G10 are used to mount the central plate to the vacuum jacket. Bottom:Detailed view of one of the lens mounts.

The right side of figure 4 shows the inside components of the camera. All components are mounted to acentral plate. Both sides of the plate are covered with polished aluminum radiation shields (not shown in figure4). The central plate is connected to the vacuum vessel with trusses made of a fiber glass (G10), combining rigidmechanical support with electrical and thermal isolation. On the lower side you see the big liquid nitrogen can,that is responsible for cooling all optical components, mounts and radiation shields. The hold time of this canis about 24 hours. On the upper side of the central plate the optics, filterwheel, HAWAII-2RG detector, cryoASIC and a small nitrogen can are installed. The small nitrogen can is only connected to the detector, providinga stable temperature for the detector.

The design of the lensmounts follows the concept used before in the HICIAO11 and IRIS instruments12 alsobuilt at the Institute for Astronomy. The lenses are spring loaded radially against two hard support points andaxially against reference surfaces. There is no possibillity to adjust the lenses in their mounts, we rely on thefabrication tolerances achived by machining. All optics and their mounts are designed for 85 K operation. Allmounts are designed to allow for the di"erent expansion coe!cients of the materials (Glas, Aluminum) and fortemperature gradients during cool-down. Figure 4 also shows a detailed view of one of the lens mounts. Thelenses were delivered by ISP Optics.

The filter mounts are directly machined into a spur gear from Berg, Inc. The design is equal to the lensmounts.A second gear on a stepper motor from Phytron GmbH is used to drive the wheel. A Hall sensor gives the referenceposition. Very accurate positioning of the filters is not necessary, because the filters are placed in the pupil image.The HAWAII-2RG detector is directly attached to the last lens mount and is cooled via a copper line to thesmall liquid nitrogen can. The detector sits on a molybdenum plate that matches the thermal expansion of thedetector substrate. The filling ports of the liquid nitrogen cans are mounted with metal bellows to the vacuumvessel.

4. ELECTRONIC & THERMAL DESIGNOn the summit of Mt. Wendelstein the Bayrische Rundfunk is running a broadcasting station right next to theobservatory. The strength of the radiation emitted by this station leads to the problem, that badly shieldedelectronics is almost unusable there. This circumstance forces us to take special attention to the shielding ofthe electronic parts used in our instruments against high frequency radiation. Because a lot of the electroniccomponents installed in the 3kk are commercial products made for lab use, we have to make an extra e"ort toget these products running in a reliable manner under these conditions. Aware of this problem, we designed anEMC (ElectroMagnetic Compatibility) - safe enclosure for the wide field imager6 on the second nasmyth port ofthe telescope. Because all the electronics has to be installed inside this enclosure we need to think about coolingthe inner volume. As soon as the behavior of this enclosure, including cooling of the heat dissipating parts istested on the wide field imager, we will adapt this design to the spatial needs of the 3kk. It is planned to coveronly the optical channels with this enclosure. Because the infrared camera is attached electrically conducting tothe rest of the structure, we consider it shielded good enough to work properly.

5. EFFICIENCY OF THE SYSTEMMost of the transmission and reflection curves we used to calculate the e!ciency of the instrument were deliveredfrom companies fabricating the di"erent parts of the optic. The reflectivity curves of the telescope mirrors wegot from the company were only measured from 400 nm to 1200 nm. For the e!ciency calculations we took thelast value of the reflectivity curve and kept it constant over the whole range of the near infrared channel. Wedid the same for the calculations in the u band. The only measured quantity so far is the transmission of theSloan filterset, done on a testbench built for the OmegaCam Project.13 All surfaces of the optical componentspassed on the way to the di"erent focal stations, the quantum e!ciency of the detectors and the reflectivity ofthe telescope mirrors are convolved and the result is shown in figure 5. The obstruction of M1 by M2 is not takeninto account. Plotted is the transmission in the filterbands before and after passing all optical components. Incomparison to GROND,14 a seven channel imager built for the MPI/ESO 2.2 m telescope at La Silla (Chile) thetotal throughput of the 3kk is about 20-30 % lower in all filterbands. The following reasons are responsible forthe lower total e!ciency:

• The total reflectivity of our telescope is about 10% lower in the optical, and 25% lower in the near infrared.

• Lower peak transmission of the 3kk filters

• In the GROND design the optical CCD’s are more customized to the required detection wavelength. Thequantum e!ciency of the HAWAII-2RG detector in the near Infrared channels is slightly better than thedetectors used in GROND.

AcknowledgementThe multi channel imager was supported by the DFG cluster of excellence Origin and Structure of the Universe.

REFERENCES[1] Hopp, U., Bender, R., Goessl, C., Mitsch, W., Barwig, H., Ri"eser, A., Lang, F., Wilke, S., Ries, C., Grupp,

F., and Relke, H., “Improving the Wendelstein Observatory for a 2m-class telescope,” in [Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series], Presented at the Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference 7016 (July 2008).

[2] Hopp, U., Bender, R., Grupp, F., Barwig, H., Gossl, C., Lang-Bardl, F., Mitsch, W., Thiele, H., Aniol, P.,Schmidt, M., Hartl, M., Kampf, D., and Schoggl, R., “The compact, low scattered-light 2m WendelsteinFraunhofer Telecope,” in [Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series],Presented at the Society of Photo-Optical Instrumentation Engineers (SPIE) Conference 7733 (2010).

[3] Vu, P., Liu, X. ., and Laxson, D., “Large area detectors and new sensor technologies at Fairchild Imaging,”in [Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series], Presented at the Societyof Photo-Optical Instrumentation Engineers (SPIE) Conference 6276 (July 2006).

0

0.2

0.4

0.6

0.8

1

300 400 500 600 700 800 900 1000 1100 1200

Thro

ughp

ut [%

]

Wavelength [nm]

Throughput optical channels

u beforeu after

g beforeg after

r beforer after

i beforei after

z beforez after

0

0.2

0.4

0.6

0.8

1

1000 1200 1400 1600 1800 2000 2200 2400 2600

Thro

ughp

ut [%

]

Wavelength [nm]

Throughput near infrared channel

Y beforeY after

J beforeJ after

H beforeH after

Ks beforeKs after

Figure 5. E"ciency of the u’,g’, r’ i’, z’ (left) and J,H,Ks filterbands. The e"ciency includes reflectivity or transmission,respectively, of all opticals surfaces, filter transmission curves and the quantum e"ciency of the detectors.

[4] Reif, K., Muller, P., Klink, G., Polder, M., and Poschmann, H., “A new design for small aperture ”BonnShutters”,” Astronomische Nachrichten 328, 711–+ (July 2007).

[5] Grupp, F., Lang, F., Bender, R., Gossl, C., and Hopp, U., “A multi-instrument focal station for a 2m-class robotic telescope,” in [Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series],Presented at the Society of Photo-Optical Instrumentation Engineers (SPIE) Conference 7014 (Aug. 2008).

[6] Gossl, C., Bender, R., Grupp, F., Hopp, U., Lang-Bardl, F., Mitsch, W., Altmann, W., Ayres, A., Clark, S.,Hartl, M., Kampf, D., Sims, G., Thiele, H., and Toerne, K., “A 64 Mpixel camera for the WendelsteinFraunhofer telescope Nasmyth wide field port – WWFI,” in [Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series], Presented at the Society of Photo-Optical Instrumentation Engineers(SPIE) Conference 7735 (2010).

[7] Fabricius, M. H., Barnes, S., Bender, R., Drory, N., Grupp, F., Hill, G. J., Hopp, U., and MacQueen,P. J., “VIRUS-W: an integral field unit spectrograph dedicated to the study of spiral galaxy bulges,” in[Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series], Presented at the Society ofPhoto-Optical Instrumentation Engineers (SPIE) Conference 7014 (Aug. 2008).

[8] Pfei"er, M. J., Frank, C., Baumueller, D., Fuhrmann, K., and Gehren, T., “FOCES - a fibre optics CassegrainEchelle spectrograph,” A&AS 130, 381–393 (June 1998).

[9] Woche, M. F., Laux, U., and Papamastorakis, J., “Dichroic beam splitter for convergent beams,” in [Societyof Photo-Optical Instrumentation Engineers (SPIE) Conference Series], M. Iye & A. F. Moorwood, ed.,Presented at the Society of Photo-Optical Instrumentation Engineers (SPIE) Conference 4008, 930–933(Aug. 2000).

[10] Reif, K., Klink, G., Muller, P., and Poschmann, H., “High Precision ”Bonn Shutters” for the largest CCDMosaic Cameras,” Astronomische Nachrichten 326, 666–+ (Aug. 2005).

[11] Hodapp, K. W., Tamura, M., Suzuki, R., Jacobson, S., Stahlberger, V., Yamada, H., Takami, H., Guyon,O., and Abe, L., “Design of the HiCIAO instrument for the Subaru Telescope,” in [Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series], Presented at the Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference 6269 (July 2006).

[12] Hodapp, K. W., Chini, R., Reipurth, B., Lemke, R., Jacobson, S., Chonis, T., Dement, K. D., and Bischo",K., “The Infra-Red Imaging Survey (IRIS) system,” in [Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series], Presented at the Society of Photo-Optical Instrumentation Engineers (SPIE)Conference 7012 (Aug. 2008).

[13] Kuijken, K., Bender, R., Cappellaro, E., Muschielok, B., Baru"olo, A., Cascone, E., Hess, H., Iwert, O.,Nicklas, H., Reif, K., Valentijn, E., Baade, D., Begeman, K. G., Bortolussi, A., Boxhoorn, D., Christen, F.,Deul, E. R., Greggio, L., Harke, R., Haefner, R., Hopp, U., Ilijevski, I., Klink, G., Kravcar, H., Magagna,C. E., Mitsch, W., Mueller, P., Poschmann, H., Rengelink, R., and Wellem, W., “OmegaCAM: wide-field

imaging with fine spatial resolution,” in [Society of Photo-Optical Instrumentation Engineers (SPIE) Con-ference Series ], A. F. M. Moorwood & M. Iye, ed., Presented at the Society of Photo-Optical InstrumentationEngineers (SPIE) Conference 5492, 484–493 (Sept. 2004).

[14] Greiner, J., Bornemann, W., Clemens, C., Deuter, M., Hasinger, G., Honsberg, M., Huber, H., Huber, S.,Krauss, M., Kruhler, T., Kupcu Yolda, A., Mayer-Hasselwander, H., Mican, B., Primak, N., Schrey, F.,Steiner, I., Szokoly, G., Thone, C. C., Yoldas, A., Klose, S., Laux, U., and Winkler, J., “Gronda 7-channelimager,” Publications of the Astronomical Society of the Pacific 120(866), 405–424 (2008).