Paon4 status O. Perdereau on behalf of the PAON4 team (LAL, CEA/IRFU, Paris & Nançay Obs.) LAL, Univ. Paris-Sud, CNRS/IN2P3, Université Paris-Saclay, Orsay, France 11th TianLai meeting Pingtan, Guizhou Sept. 18th, 2018 O. Perdereau Paon 4 18/9/18 1 / 19
PAON4 Characteristics : 4 antennas ( ∼ 3 deg beams) in Nancay ( ∼ 200 km south of Paris) 2 polar./antenna Frequency band 1250 - 1500 MHz ( ∼ 1275 - 1480 MHz usable) ± 20 degrees from zenith transit observations , ∼ 24h scans , ≥ 2015 test bench for electronics, DAQ and on-line computing analysis O. Perdereau Paon 4 18/9/18 2 / 19
PAON4 Characteristics : 4 antennas ( ∼ 3 deg beams) in Nancay ( ∼ 200 km south of Paris) 2 polar./antenna Frequency band 1250 - 1500 MHz ( ∼ 1275 - 1480 MHz usable) ± 20 degrees from zenith transit observations , ∼ 24h scans , ≥ 2015 test bench for electronics, DAQ and on-line computing analysis O. Perdereau Paon 4 18/9/18 2 / 19
Digitization and cabling two independent ADC boards (4 channels/boards) two FPGAs on each (2 channels/board) external trigger ( ∼ 8 kHz) distributed to ADC boards (and internally to the FPGAs) digitized packs of data analyzed (FFT s) by a two layers servers small cluster (4+2) two channels added recently (gain drift calibration & RFI) on a 3rd ADC board O. Perdereau Paon 4 18/9/18 3 / 19
recent works The strange case of the perturbed channel Gain calibration Phase stablity & calibration Recent scans O. Perdereau Paon 4 18/9/18 4 / 19
T ypical time-frequency maps (autocor.) “trend” : temporal gain variations (temperature) correction using blind extra channel O. Perdereau Paon 4 18/9/18 5 / 19
T ypical time-frequency maps (autocor.) SATELLITES RFIs ? 21 cm CygA (+Gal.) “trend” : temporal gain variations (temperature) correction using blind extra channel O. Perdereau Paon 4 18/9/18 5 / 19
T ypical time-frequency maps (autocor.) “trend” : temporal gain variations (temperature) correction using blind extra channel O. Perdereau Paon 4 18/9/18 5 / 19
The “perturbation” Time-frequency map (4H) O. Perdereau Paon 4 18/9/18 6 / 19
The “perturbation” Signal trends additionnal noise ( ?) rather than extra gain change ? O. Perdereau Paon 4 18/9/18 6 / 19
The strange case of the perturbed channel(s) Only one antenna at a time was affected Phenomena observed in 2015, prior to october (stopped after some work on site ?), on antenna 3 Antenna 4 was “preferred” in 2017-8 ∼ 80 % probability of occurence Sometimes ( ∼ 20 % ) phenomenon switched to another antenna fast rise/fall, relative amplitude 40 to 100 % investigations (long) : ◮ permutations in signal transmission & RO : perturbation from “outside” (source seen in some FSL ? radar ? phone ?) ◮ improvements in grounding & filtering unsuccessfull ◮ direct investigations not successfull (at first) O. Perdereau Paon 4 18/9/18 7 / 19
Start and end hours (some coarse measurements on low-res plots) ⇒ Correlation with sunset (start) and rise (end) at Nançay in 2015 et 2017-8 (within 20 mn) O. Perdereau Paon 4 18/9/18 8 / 19
And the Guilty is ... going around the parabollas at sunsets or nights with antennas, scopes,... brought nothing but indirect evidences ... and more carefull direct (unrecorded) observations show it’a little bird spending nights inside the feeds’ envelope installing anti-bird nets cured the problem ! O. Perdereau Paon 4 18/9/18 9 / 19
Gain drift correction “Thermometric” new channel (LNA + 50 Ω end-cap) readout by same electronics than the antennas’ feeds ⇒ template for gain variation with time (temperature) linear fit (one offset per day) ⇒ correction of gain variation better than ∼ ± 2 % O. Perdereau Paon 4 18/9/18 10 / 19
Electronic phase relative calibration Differential phase shifts between signal transmission lines use bright source transits to fit relative phases (and other params) for relative phases using point source signal modelling in frequency bins phase low degree polynomial model → used to correct data H and V polar treated independently at the moment O. Perdereau Paon 4 18/9/18 11 / 19
Electronic phase relative calibration Differential phase shifts between signal transmission lines use bright source transits to fit relative phases (and other params) for relative phases using point source signal modelling in frequency bins phase low degree polynomial model → used to correct data H and V polar treated independently at the moment O. Perdereau Paon 4 18/9/18 11 / 19
Observations - summer 2018 Thanks to C. Pailler for this data-taking (and more !) O. Perdereau Paon 4 18/9/18 12 / 19
Observations - summer 2018 : examples scan at CasA (RA ∼ 23 . 5 h ) elevation O. Perdereau Paon 4 18/9/18 13 / 19
Observations - summer 2018 : examples Standing waves CasA scan at CasA (RA ∼ 23 . 5 h ) elevation O. Perdereau Paon 4 18/9/18 13 / 19
Observations - summer 2018 : examples scan at CasA (RA ∼ 23 . 5 h ) elevation O. Perdereau Paon 4 18/9/18 13 / 19
Observations - summer 2018 : examples Sun through FSL ? CasA Corr. noise? scan at CasA (RA ∼ 23 . 5 h ) elevation O. Perdereau Paon 4 18/9/18 13 / 19
First comparision with expectations PAON4 A04 (elev=11.4) and expected 2H3H @ 1390 MHz PAON4 A04 (elev=11.4) and expected 2H3H @ 1390 MHz PAON4 A04 (elev=11.4) and expected AutoCor @ 1420 MHz PAON4 A04 (elev=11.4) and expected AutoCor @ 1420 MHz 20 20 Signal Level, Kelvin Signal Level, Kelvin Signal Level, Kelvin Signal Level, Kelvin 6 18 18 PAON4 data (X-Cor.real) Expected (X-Cor.real) 16 16 4 14 14 2 12 12 PAON4 data (AutoCor) 10 10 0 Expected (AutoCor) 8 8 6 6 -2 -2 4 4 -4 -4 2 2 0 0 -6 -6 0 5 10 10 15 15 20 20 22 22 22.5 22.5 23 23 23.5 23.5 24 24 RA (hours) RA (hours) RA (hours) RA (hours) PAON4 A04 (elev=11.4) and expected 2H3H @ 1390 MHz PAON4 A04 (elev=11.4) and expected 2H3H @ 1390 MHz PAON4 A10 (elev=1.4) and expected 2H4H @ 1390 MHz PAON4 A10 (elev=1.4) and expected 2H4H @ 1390 MHz Signal Level, Kelvin Signal Level, Kelvin Signal Level, Kelvin Signal Level, Kelvin 6 PAON4 data (X-Cor.imag) PAON4 data (X-Cor.imag) Expected (- X-Cor.imag) Expected (- X-Cor.imag) 0.1 0.1 4 2 0.05 0.05 0 0 0 -2 -2 -0.05 -0.05 -4 -4 -0.1 -0.1 -6 -6 22 22 22.5 22.5 23 23 23.5 23.5 24 24 17 17 17.5 17.5 18 18 18.5 18.5 19 19 19.5 19.5 20 20 RA (hours) RA (hours) RA (hours) RA (hours) O. Perdereau Paon 4 18/9/18 14 / 19
Summary (1) Several improvements in our set-up (e.g. new “systematics”-oriented channels) A long lasting problem solved 2018 summer data looks promising Still to do : RFI filtering and/or frequency selection, Sun contamination, ... Some imperfections e.g. standing waves in cables, correlated noise led to developments of NEBULA/IDROGEN boards (D. Charlet’s talk) O. Perdereau Paon 4 18/9/18 15 / 19
NEBULA Board ADC 081020 2x 1GSPS • MTCA 4.0 standard, Double-width, Input bandwidth : 2Ghz full size AMC. • FPGA : 5AGTMC7G3F31 WR 1Gb • Stand alone mode (12v) Eth • DAC 2 channels 1GSPS. Synch & config 10Gb Data • White Rabbit compliant. config Eth 10Gb • On board configuration (µC) Data/config FPGA ctrl/cde • Very low noise synthesizer PLL & conf synthesizer cleaner (LM04828) SPI IPBus SPI • Front panel : WR SFP+ µC ATMEGA EPLD 2x SFP+ 10GbEth IPMI SPI I2C MAXV • Backplane connectivity : PCIE x4 Gbe IP bus,PCI 4x Gen3, IPMB IPMB, CLK & trigger lane. SSRAM Flash x 2 Power CTRL WhiteRabbit LMK04828 DS1014 512K 1Mb Daniel Charlet Guiyang 09-2018 2 O. Perdereau Paon 4 18/9/18 16 / 19
NEBULA board, PPS test • 400 fs after 1000s over 100 m 3 O. Perdereau Paon 4 18/9/18 17 / 19
IDROGEN board IDROGEN board CLK.Ext LMK04828 SI5338 10 FMC+ 7 4 160 I2C • MTCA 4.0 standard, Double-width, SPI CLK CLK full size AMC. RTM 28 • FPGA : 10GX027H4F34 PPS / trig.Ext • Stand alone mode (power 12v) PCIEx 4x Gen3 • HighPinCount FMC slot. Ethernet 40G ARRIA 10 SX 10AX027H4F34 USB • White Rabbit compliant. White rabbit Ethernet 1G SFP+ Ethernet 1G IP bus • Front panel connectivity : WR SFP+ I2C Ethernet 1G QSFP+ 40G, USB • Backplane connectivity : Serial link 40G QSFP+ 1Gbe IPbus,PCI 4x Gen3, AS prog IPMB, CLK & trigger lane. JTAG I2C ● RTM connector : J30. I2C spi JTAG IPMB ● Low cost MAX10 ATMEGA128 Cypress EPCQ USB FX2LP USB Daniel Charlet Guiyang 09-2018 4 O. Perdereau Paon 4 18/9/18 18 / 19
Summary (2) Stay tuned for more on data analysis (sensitivity, sky maps,...) main improvement : IDROGEN board → will be described by D. Charlet this afternoon will be tested and used soon (early 2019 ?) in PAON4 THANK YOU ! O. Perdereau Paon 4 18/9/18 19 / 19
Recommend
More recommend
