Nel IRAM KIDs Array Grenoble Nel IRAM LPSC
Néel IRAM KIDs Array Grenoble (Néel, IRAM, LPSC, LAOG) Cardiff SRON Roma-Trento
NIKA timeline November 2008 : Kick-off 2009 : First antenna-doupled arrays in February First demonstration of LEKIDs arrays in May Final arrays made in September 30 -40 pixels, mono-band 2 mm, run in October Pico Veleta Sensitivity: 1. 2· 10 -15 W/Hz 0. 5 Reference: 2010 : 112 x 2 pixels Sensitivity (in progress) 3 6· 10 -16 W/Hz 0. 5 Sensitive to only one polarisation ! Dual-band: 125 170 GHz et 180 270 GHz 2011 (GOALS) : to be discussed (many options)
Electronics PAST / PRESENT NIKA 2009: Bonn electronics NIKA 2009 400 MHz (ADC 8 bits), FFT, max. rate 5 Hz NIKA 2009: Néel box ( « Loren » ) NIKA 2009 50 MHz (ADC 12 bits), 32 ch. , max. rate 50 Hz NIKA 2010: ROACH board (Néel box with up-down converter and USB attenuators) NIKA 2010 230 MHz (ADC 12 bits), 112 channels, rate 22 Hz FUTURE LPSC 2011 GOAL : > 256 channels modules over > 400 MHz band Fast rate (at least 20 Hz are needed) First prototype funded by CNES and IN 2 P 3 SRON 2011 see Andrey, but GOAL is >1 GHz band, FFT (rate ? )
Noise spectrum on (good) sky Average LEKIDs detectors noise : 2 Hz / Hz 0. 5 at 1 Hz (stable during the run
Misterious 50 s-period noise 100 Hz steps !!! Detectots noise is less than 10 Hz p-p !! Appearing from time to time, lasting for a few hours and then disappearing.
50 s noise again
A HUGE insect in the cabin
EMIR chopper ON/OFF
B-field jumps on antenna-coupled array
B-field jumps on LEKIDs
Laboratory (sky-simulator) performances Average 4 m. K/Hz 0. 5 4· 10 -16 W/Hz 0. 5 per pixel Problem confirmed on the telescope. Can be explained by the electronics : sensitivity depends critically on power per KID. Taken into account for LPSC 2011 electronics. Best pixel < 2 m. K/Hz 0. 5 ( 1. 1 m. K/Hz 0. 5 if « speed read-out » )
Very rough estimation of S/N The average « very raw » S/N on Mars is: Signal : Noise : 2 mm 2 -4 k. Hz 2 Hz/Hz 0. 5 1. 25 mm 10 k. Hz 16 -20 Hz/Hz 0. 5 Mars flux : 40 Jy 107 Jy S/N 1000 -2000 Hz 0. 5 Sensitivity (1 Hz) 20 -40 m. Jy/Hz 0. 5 (per pixel) NEP (1 Hz) estimated by sky simulator !! 500 -625 Hz 0. 5 170 -210 m. Jy/Hz 0. 5 Please take it as very raw. Then other effects might limit the noise on real maps. IN ANY CASE IT’S MORE OR LESS WHAT WE EXPECTED BASED ON THE SKY SIMULATOR RESULTS.
Exemple résultat : Cyg. A
Exemple résultat : DR 21(OH)
Spectral response (Grenoble MPI, 11/2010)
Conclusions DETECTORS Clearly better than 2009 at 2 mm. A factor of 4 very raw S/N estimation. Still to be confirmed on real maps (B fields ? ? ). Best LEKIDs pixels tested on sky simulator already almost OK. Average performances of the array to be improved: - dual-polarisation meanders (LEKIDs). Better for cross-talk too. - AR coating (LEKIDs and antennas) - electronics (same power on all pixels) - read-out on « maximum speed » (but less dynamics) - if all that not enough: Ti. N (or similar) for higher response - …. Andrey ? Jochem ? PIXELS COUNT Much better than in 2009 (factor 4 -6). Limited by the electronics. Arrays sizes will now grow according to electronics developments. OPTICS Same cryostat/filters is OK up to 3 arc-min field-of-view. Use a real dichroic for splitting (factor of 2 in power).
Plans for 2011 A FURTHER TECHNICAL RUN ? TWO SEPARATE RUNS ? A MORE « USABLE » INSTRUMENT ? WORKING ON THE BIG CAMERA ? HOW BIG ? IN CASE WE GO FOR SPRING 2011 : CRYOSTAT PIXELS ELECTRONICS FILTERS SPLITTER DETECTORS 2 mm DETECTORS 1 mm PULSE-TUBE (if possible) 224 per array, over a 400 MHz band Baseline: LPSC 2011. ROACH is backup if LPSC is still under development. from NIKA 2010 DICHROIC (if possible) Same as NIKA 2010; dual-polar if dichroic Antennas or LEKIDs (the best, taking into account the number of pixels that can be actually read-out by existing electronics at that time, and sensitivity)
- Slides: 17