1. RUN HISTORY Preparation: 17/10Cryostat, pumps and electronics mounted in the cabin (total time 2h) 18/10Cooling down to 80mK. Resonances OK (SRON array) 19/10Laser alignment and test on the sky. Seen Venus. Thick clouds. (tot. Time 2h) First slot (SRON array): 20/10 Snow 21/10Morning: seen again Venus and then 3C273. 21/10Morning: seen again Venus and then 3C273. Afternoon: wind and rain. 22/10 Afternoon: good weather. Seen 3C345, MWC349 (6Jy) 22/10Morning: rain. Afternoon: good weather. Seen 3C345, MWC349 (6Jy) Preparation second slot: 22/10 20h-24h. Heating up the cryostat 23/100h-5h. Open/close, fix a leak on the compressor. 7-8h. Back in place. Second slot (LEKIDs): 24/10Bonn electronics. Total power scans. Planets, quasars, sub-Jy sources, GRB. 25/10FPGA electronics. Total power scans. A number of sources. 26/10FPGA electronics. Try wobbler mode. Various problems. 27/10FPGA electronics. OK with wobbler and total power. Extended sources.

2. Tool for finding the resonances - BEFORE

3. Tool for finding the resonances - AFTER

4. SRON array - 42 pixels + 2 blinds - total bandwidth 200MHz - Bonn electronics (5Hz)

5. SRON array – First Light Time Domain Trace Working in the cabin Mirror No beam (300K) Sky Venus Telescope crazy (horizon) Frequency scan Clouds Detectors dynamics 19/10. Technical time for alignment.

6. SRON array – Noise spectrum Taken with the mirror on the cryostat input  Detectors noise average 12 mdeg/Hz 0.5 @ 1Hz 45 mdeg/Hz 0.5 @ 0.1Hz

7. SRON array – Venus response and S/N S = 6 deg(TBC !!!) The response to Venus is, on average, S = 6 deg in phase (TBC !!!) Venus was 10.7 arc-sec in diameter, for a temperature of 232K (http://nssdc.gsfc.nasa.gov/planetary/factsheet) The beams FWHM is 24 arc-sec. So, the effective temperature of Venus is (order of magnitude):  T = 232K  (10.7/24) 2 = 46K (Dilution of 232K on the beam) N = 12 mdeg/Hz 0.5 So since the noise at 1Hz (previous slide) is N = 12 mdeg/Hz 0.5 We have S/N = 500 Hz 0.5 @ 1Hz The NET of the single pixel is thus: NET pix =  T / (S/N) = 46 / 500 = 92 mK / Hz 0.5 Since the beam signal is split over X  4 pixels on average (0.5  F ) NET beam = NET pix / X 0.5 = 46 mK / Hz 0.5

8. SRON array – 3C273

9. SRON array – Improvements using radius Should be a factor of 3 in S/N using the radius read-out (Andrey). …….. Applicable to the LEKIDs too.

10. SRON array – Skydip (1.1 to 1.8 airmasses) Sky was really bad. Barely seeing the EL effect in the clouds.

11. LEKIDs - 30 pixels - total bandwidth 45MHz - Bonn electronics (5Hz) or FPGA (48Hz)

12. LEKIDs – Noise spectra (Grenoble) Detectors noise: Detectors noise: 4 mdeg/Hz 0.5 @ 1Hz 10 mdeg/Hz 0.5 @ 0.1Hz

13. LEKIDs – Noise spectra (on sky, during total power scan) Sky noise (correlated) dominates below 0.4 Hz Detectors noise: Detectors noise: 5 mdeg/Hz 0.5 @ 1Hz 12 mdeg/Hz 0.5 @ 0.1Hz Well comparable with that measured in Grenoble. SKY noise Detectors noise

14. LEKIDs – Noise spectra (on sky, during wobbler scan) The continuous is still comparable. Lines are the wobbler and the harmonics of course (signal).

15. LEKIDs – Mars signal The PSF “halo” is clearly seen already in the time-domain raw data. We have 5 deg PHASE signal for Mars on the average pixel.

16. LEKIDs array – Mars response and S/N S = 5 deg The response to Mars is, on average, S = 5 deg in phase Mars was 8 arc-sec in diameter, for a temperature of 210K (http://nssdc.gsfc.nasa.gov/planetary/factsheet) The beams FWHM is 24 arc-sec. So, the effective temperature of Mars is (order of magnitude):  T = 210K  (8/24) 2 = 23 K (Dilution of 210K on the beam) N = 5 mdeg/Hz 0.5 So since the noise at 1Hz (previous slide) is N = 5 mdeg/Hz 0.5 We have S/N = 5000/5 = 1000 Hz 0.5 @ 1Hz The NET of the single pixel is thus: NET pix =  T / (S/N) = 23 / 1000 = 23 mK / Hz 0.5 Since the beam signal is split over X  2 pixels on average (  0.7  F ) NET beam = NET pix / X 0.5 = 17 mK / Hz 0.5

17. LEKIDs array - BL1418+546 (estimated 500mJy) Visible in the first scan. Faintest source detected 200mJy (WR147).. Good S/N Quick look not adapted for longer integrations.

18. LEKIDs array - G34.3

19. LEKIDs array – M87 with wobbler

20. LEKIDs array – GRB091024

21. LEKIDs array – Skydip (1.1 to 1.8 airmasses)

22. LEKIDs array – Skydip 2 Detectors dynamics The dynamics is OK to include the whole EL range. but TBC Sky was not exceptionally good (   0.3 but TBC)  large signal In case of strong clouds it might be needed to re-center the resonances.

23. LEKIDs array – B fields during slew Strange behaviour during large telescope slews.. Jumps. Superconducting box ? LEKIDs more sensitive to B fields. Not seeing it during observations.

24. CONCLUSIONS Great experience; same performances measured in lab, or a bit better. OK with cryogenics, alignment, interfacing and so on… TO BE DONE for the FUTURE (factor of 10 missing on sensitivity, dual band, pixels)INSTRUMENT/OPTICS: - Design/fabricate the alternate optics for dual band 1.25 and 2mm - Pulse tube cryostat (easier for IRAM) - AR on the lenses/windows DETECTORS: - Reducing the phase noise by changing the C geometry - Improving the sensitivity by reducing the volume of the resonators and using optimal Q - Optical coupling (e.g. thickness, backshort) - Films quality (for LEKIDs) - Start 1.25mm designs - Improve the homogeneity of the pixels (EM cross-talk, other effects ??) for larger arrays SOFTWARE/ELECTRONICS: - Optimise the electronics in general (starting from the cold amplifier) - Radius/amplitude implementation (a factor of 3 better S/N according to Andrey’s results) - Pixels de-correlation - Off-line pipeline - Open Source and LPSC electronics in case Bonn no longer available