FUTURE INSTRUMENTATION FOR JCMT II Dan Bintley and Per Friberg East Asian Observatory East Asia Sub-millimeter-wave Receiver Technology Workshop 1
ABSTRACT The EAO's James Clerk Maxwell Telescope (JCMT) provides the East Asian Region with a premier facility for deep wide-field mapping of the universe in the sub-millimetre and far-infrared, that includes the capability for simultaneous 2 colour continuum imaging, polarimetry, heterodyne observing at 230 GHz and 345 GHz and VLBI as part of the EHT. We discus plans to upgrade and replace the existing suite of instruments at JCMT, including: A new generation, state of the art multi-receptor heterodyne array, to replace HARP (a 16-receptor 345 GHz array receiver). A new sideband separating dual polarization 230 GHz receiver. New KID arrays for SCUBA-2. A replacement for HARP and an upgraded SCUBA-2 would potentially increase mapping speeds at 345 GHz (850um) by an order of magnitude. Our goal is to enlist the regions' expertise in receiver and instrument design, to solve the many technical challenges required to implement these ambitious plans. East Asia Sub-millimeter-wave Receiver Technology Workshop 2
TALK OUTLINE Focus on SCUBA-2 Describe what makes SCUBA-2 so good Show how the detector arrays were integrated into the instrument and describe aspects of the manufacture process and show some array properties Highlight problems with the SCUBA-2 TES array design The design flaws were things we were aware of we made reasoned decisions at each step However, if we were to make these arrays again there are some different choices that we would make Explain the stage we are at with the plan to upgrade or replace SCUBA-2 East Asia Sub-millimeter-wave Receiver Technology Workshop 3
SCUBA-2: THE GOLD STANDARD FOR LARGE FORMAT SUBMILLIMETRE CAMERAS SCUBA-2 is a 10,000 pixel bolometer camera on the JCMT, operational since October 2011. Two focal planes, each with four 32 by 40 MoCu TES sub-arrays with inline 2-D Time Domain SQUID MUX. Observes simultaneously at 850µm and 450µm, with a 43 sq-arcmin field of view. A survey instrument: - a square degree of sky can be mapped to 6mJy/beam at 850µm in less than 7 hours. East Asia Sub-millimeter-wave Receiver Technology Workshop 4
5 YEARS OF SCUBA-2 OBSERVATIONS Images from the SCUBA-2 Gould Belt survey of star forming regions. (right) Includes Herschel 100 micron, Spitzer 8 micron data. All of this SCUBA-2 data is now freely available;- both as raw data and as reduced maps made with a standard reduction. We have also released a detailed source catalogue that includes all of the JCMT SCUBA-2 observations and data, plus none-propriety EAO observations, data and calibrations. SCUBA-2 850 micron This forms a legacy for JCMT and the East Asian Observatory map of W40 and Serpens South. Credit: JCMT East Asia Sub-millimeter-wave Receiver Technology Workshop Gould Belt Survey 5
SCUBA-2 IN THE ALMA ERA A large field of view coupled with a smaller beam size than Herschel gives SCUBA-2 huge advantage for wide field mapping East Asia Sub-millimeter-wave Receiver Technology Workshop 6
POL-2 AND FTS-2 ENHANCE SCUBA-2 Published data from BISTRO large program FTS-2 not yet commissioned East Asia Sub-millimeter-wave Receiver Technology Workshop 7
DEMAND AND SCUBA-2 PUBLICATIONS Citations of one SCUBA- 2 instrument paper SCUBA-2 publications are increasing year on year East Asia Sub-millimeter-wave Receiver Technology Workshop 8
SCUBA-2 TES ARRAY DESIGN AND PACKAGING Designed in the early 2000s 15 years old MUX wafer TES wafer Not shown are oxide and moly shield layers over the TES East Asia Sub-millimeter-wave Receiver Technology Workshop 9
NIST DESIGN 2D TD MUX Frame/Sample rate ~ 12.7KHz Data out at 200 Hz East Asia Sub-millimeter-wave Receiver Technology Workshop 10
DEEP ETCHING USING DRIE PROCESS Issues of uniformity, depth, undercuts, grass East Asia Sub-millimeter-wave Receiver Technology Workshop 11
PIXEL HEATER AND THERMAL DESIGN SCUBA-2 TES pixel design includes a heater that:- Reduces the required bias current giving lower detector noise Adjusted to compensate for changes in sky power Used to flatfield the array and calculate the per pixel responsivity at the start of each observation Heater surrounds the TES Wired in series, bias common to all detectors on sub-array East Asia Sub-millimeter-wave Receiver Technology Workshop 12
SUB-1K THERMAL AND MECHANICAL DESIGN o BeCu Hairbrush o Batwing o Batwing to cable joint o Array to 1K pcb interconnects o Sapphire Thermal isolation joints East Asia Sub-millimeter-wave Receiver Technology Workshop 13
East Asia Sub-millimeter-wave Receiver Technology Workshop 14
East Asia Sub-millimeter-wave Receiver Technology Workshop 15
East Asia Sub-millimeter-wave Receiver Technology Workshop 16
East Asia Sub-millimeter-wave Receiver Technology Workshop 17
Assembled SCUBA-2 focal plane Each sub-array is 32 x 40 40 x 40 better fills the field of view East Asia Sub-millimeter-wave Receiver Technology Workshop 18
ARRAY PROPERTIES 850 TC East Asia Sub-millimeter-wave Receiver Technology Workshop 19
INITIAL ARRAY PROPERTIES: NEP AND POWER HANDLING We were able to improve on these initial average dark NEP results, by optimising the sub-arrays TES bias & heater setups East Asia Sub-millimeter-wave Receiver Technology Workshop 20
IMPROVING THE ON-SKY PERFORMANCE Effective number of contributing detectors Yield of detectors and MUX is 95% yet only 60% of detectors contribute to mapping at each wavelength This is a consequence of the variation of detector properties such as Tc and G across the wafer and the compromise in setting bias and heater to achieve the best effective NEP per sub-array 450 focal plane responsivity Before and after optimizing the effective NEP In general we found lower TES bias improves both dark NEPs and array mapping speed but fewer pixels in transition East Asia Sub-millimeter-wave Receiver Technology Workshop 21
SCUBA-2 ON-SKY PERFORMANCE Superb facility instrument Reliable few faults Consistent performance Easy to operate NEFD for each waveband as a function of fractional sky transmission East Asia Sub-millimeter-wave Receiver Technology Workshop 22
Higher then expected NEFD Measured NEFD NEFD based on dark NEP Expected NEFD Figure 1: The measured NEFD as a function of normalised sky transmission at 450µm (left: blue points) and 850µm (right: red points). The black lines are fits to the measured data. The purple and green lines represent the calculated NEFD using the measured dark detector noise at 450 and 850µm, respectively. The blue and red curves are the calculated NEFD using a detector noise assuming theoretical performance, again at 450 and 850µm, respectively. Mapping speed is reduced by a factor of 10 at 850um The team reported their initial findings and recommendations in May 2013 (the interim report see reference 3). At that time a number of tests and analyses were still outstanding and no firm conclusions had been reached for either of the main issues. The following two sections discuss the progress made in terms of 3.2 Excess power loading. Figure 1 shows a plot of the measured NEFD as a function of sky transmission at 450 and 850µm (from Holland et al. 2013). Superimposed on the plot are two further curves derived from an optical model of the system (including sky, telescope and instrument): (1) the calculated NEFD using the East Asia Sub-millimeter-wave Receiver Technology Workshop 23
ISSUES WITH SCUBA-2 DETECTORS Thermal conductance G higher than design and is not uniform over sub-array or between sub-arrays. Phonon noise is therefore higher. Tc is not uniform across the sub-arrays (variation greater than 10mk) There was a deliberate choice made to error on too high total power handling to ensure the arrays would not saturate at expense of noise performance Hybridization process led to potential parasitic resistance on some sub-arrays or etched some TES on others East Asia Sub-millimeter-wave Receiver Technology Workshop 24
ISSUES WITH SCUBA-2 DETECTORS One TES bias common to all pixels on a sub-array One pixel heater bias common to all pixels on a sub-array Limited multiplexing rate Some nonlinear detector response, particularly with large signals. (A problem for FTS-2). Mapping of weak extended sources difficult East Asia Sub-millimeter-wave Receiver Technology Workshop 25
PLANS FOR AN UPGRADED SCUBA-2 Reuse the SCUBA-2 cryostat, cold optics and cryogenics Reduce stray light / excess optical power on the arrays Replace the detector arrays? The alternative to upgrades is a new instrument to replace SCUBA-2 East Asia Sub-millimeter-wave Receiver Technology Workshop 26
REDUCE STRAY LIGHT/EXCESS POWER There have been significant improvements in the Cardiff filter design and manufacture, since the SCUBA-2 filters were made and installed. The Cardiff filter group have redesigned the 6 filters and thermal blockers from the window to the 4K optics box. East Asia Sub-millimeter-wave Receiver Technology Workshop 27
RESULTS OF THE THERMAL FILTER CHANGE Dark NEPs reduced on all sub-arrays largest gain at 450um Small mapping speed gain Temperatures reduced inside 4K box (4K box warmer due to remote motor PTC cold heads) Still measure excess power therefore possible stray light from the complex optical path remains a potential issue and requires further work to fully understand East Asia Sub-millimeter-wave Receiver Technology Workshop 28
SCUBA-2 OPTICAL PATH East Asia Sub-millimeter-wave Receiver Technology Workshop 29
NEW DETECTOR ARRAYS: - TES OR KID Both can achieve the required specs to upgrade SCUBA-2 Neither are off the shelf solutions TES are more mature and have advanced since SCUBA-2 SQUID MUX has improved lower power more uniformity less magnetic pickup higher multiplexing factors Know how to make good absorber and/or couple to TES KID arrays are potentially simpler to fabricate KID arrays don t require a SQUID MUX The cold electronics for KIDs is an amplifier per 1000/2000 pixels For TES arrays, the cold electronics is a SSA per 40 pixels KID arrays are far less sensitive to thermal fluctuations East Asia Sub-millimeter-wave Receiver Technology Workshop 30
KID ARRAYS FOR SCUBA-2 o Choice of Pixel design o Choice of materials [Al, TiN] o Choice of frequency of operation Detectors o Coupling to radiation MKID: feedhorn/antenna superconducting pair breaking detector antenna determines RF-band Cooper Pairs: h resonator determines KID-readout frequency Inductance iωl(p o detector Testing + multiplexing and filter in one optimisation sky ) structure Resistance R(P sky ) Antenna Coupled KID Feedline Connects all KIDs to readout Coupler Si micro lenses 1 mm 2Δ Quasiparticles: Z s = R + iωl Next gen BLASTpol KID resonator NbTiN 2 /h=1.4 THz lossless 0.1 mm Al 2 /h=80 GHz radiation absorption AMKID Antenna In focus of lens Light Dark LEKID design East Asia Sub-millimeter-wave Receiver Technology Workshop 31
ENGINEERING REQUIREMENTS FOR 5 feedlines INSTALLING KID ARRAYS IN SCUBA-2 5400 pixels Si lens array 2 cm East Asia Sub-millimeter-wave Receiver Technology Workshop 32
BASELINE FOR NEW ARRAYS TES 32 x 40 sub-arrays KID 40 x 40 sub-arrays Number of pixels Pixel type Current TES (850um) New TES (850um) NEW KID (850um) 4 x 32 x 40 4 x 32 x 40 4 x 40 x 40 Filled array Fl/2 Filled array Fl/2 Filled array Fl/2 Dark NEP 1x10-16 W/Hz 0.5 5x10-17 W/Hz 0.5 5x10-17 W/Hz 0.5 Power Handling Mapping speed 160 pw TBD 1 8 12 East Asia Sub-millimeter-wave Receiver Technology Workshop 33
THE NEXT STEP: A NEW 850 MICRON KID ARRAY INSTRUMENT FOR JCMT RX CABIN? A good next step would be to design and build a KID camera that can be installed in the JCMT Rx cabin:- Cardiff have a submitted a grant proposal in to build such an instrument. They make the science case for an instrument that will map at 100 SQ- Degree cosmological field to a depth of 2mJy. Such a survey would be important for several main goals: (a) To trace SMGs out to very high-z in statistically significant numbers to sample the general SMG population, rather than just the bright lensed sources (b) To use these samples to investigate the role of dust-obscured star-formation at redshifts greater than 4 and therefore the rate of enrichment of the interstellar medium by metals at these very early epochs (c) To investigate how SMGs trace large-scale structure and what environments they mostly are sited in East Asia Sub-millimeter-wave Receiver Technology Workshop 34
The factor of at least x10 gain compared to SCUBA-2 will be achieved by a combination of the following factors: (1) A field-of-view of 15 arcminute diameter, compared to roughly 8 arcminutes for SCUBA-2 (2) Improve the usable pixel yield from the current 60% with SCUBA-2 to 95%. (3) Achieving a sensitivity per pixel much closer to the theoretical limit about 30 mjyhz -1/2 These potentially would lead to x30 gain in mapping speed Whether or not Cardiff gets the grant collaboration on this project or a more modest plan for a initial smaller field of view KID camera that can be installed into the JCMT cabin would proved JCMT with a state of the art camera that would initially supplement SCUBA-2 and then lead to a replacement or act as a pathfinder for upgrading the SCUBA-2 arrays in the near future. East Asia Sub-millimeter-wave Receiver Technology Workshop 35
CONCLUSIONS SCUBA-2 is working extremely well. Mapping speeds can be further increased by a factor of 10 by upgrading the detector arrays. We started the process by working to improve the control of excess optical power on the focal planes Looking for partners to work with EAO institutions to make new detector arrays. Chosen KID arrays as the best replacement. Next step is a KID camera for the JCMT Rx cabin, that would help develop KID array technology to install in SCUBA-2. East Asia Sub-millimeter-wave Receiver Technology Workshop 36