NIKA2 run #53 (Jul 9, 2021)
Motivation
On past May 25th morning, we started the NIKA2 acquisition in order to start the observations of the DDT project E03-20 in the framework of a heterodyne Pool #5 (week 21). After that, everything went fine with pretty noise traces in all detectors, and we prepared to observe in the evening. But after lunch, we noticed that a sort of strongly correlated noise (or what was interpreted as such) between 0.05 and ~10 Hz affecting array 2. Noise traces of arrays 1 and 3 looked reasonably fine. Santiago checked that no high-frequency oscillator had been left running in the cabin (using the spectra analyser + a mobile phone), and restarted the crate, synthesizer, and preamplifier on array 2 without success. Pneumatic actuators were checked too (only for not leaving hypotheses out) and thought about possible maintenance jobs or interferences that inadvertently could have affected the instrument. On May 27th the problem still persisted. A sample of time series and noise spectra of array 2 is shown below:
After sharing some screenshots of noise traces with Alessandro Monfardini (cc. Martino Calvo) describing the situation, we agreed with him to a light intervention of the pulse tubes (PT) on the same day. With Santiago Navarro and Dave John, we spent most of yesterday getting rid of the correlated noise on this array by adjusting the PTs (finding a point where the temperature of the 4K-plate is the highest and the noise marginally better), but the traces of array 2 did not improve significantly. Previously, Santiago and Dave replaced the corresponding synthesizer without an improvement of noise traces.
Despite this situation, weather conditions improved (i.e. more or less well suited for NIKA2 observations) on May 28th and 29th for some tests on the sky. We obtained calibrations on NGC7027 and MWC349. The statistics of the ql-reduced maps compared with those obtained in January and February at the same opacity level (tau between 0.4 and 0.5) are nicely consistent for all arrays. According to this, such a noise source did not seem to be affecting the 2 mm maps in terms of sensitivity (rms). No notable distinctive features or strange things are visible on the current 2 mm maps of both sources either. Some science observations (E03-20) were tried (see the report here).
At the beginning of June, Samuel Leclerc and Alessandro suggested stopping the pulse tubes and compressors and immediately (~30 s or so) check the effect on the noise. The anomalous noise pattern in array 2 remained with this test. The problem described disappeared spontaneously -to all appearances- as revealed by the noise traces monitoring carried out since July 5th.
Goals of the run
Thanks to Dave and Santiago we used the T07 technical time of July 9th for (i) checking noise traces of all arrays exhaustively and (ii) test the behaviour of the instrument on the sky after the remission of the correlated noise pattern in array 2. In addition, we take advantage of these observations to test the configuration and performance of the new NIKA2 Pool account (nikas-21).
Observing report and results
Observing in remote mode (Angel Bongiovanni). Times referred to LT.
17h00m - 21h00; Opacity (tau[225 GHz]) above 0.7 at the beginning, going down to ~0.4 at the end of the shift. Unstable in general. First pointing and focus on Mars. Broad/distorted beams (>21" at 2mm) until ~19h. More decent focus corrections and even a bit broadened beams widths after that. Several pointing/bright sources of NIKA2 Pool catalogues observed in previous semesters under opacity conditions quite similar to the predicted ones for that date were chosen, in order to compare rms statistics. In particular, PCCS2G208.07 and M82 fields observed in W2020 and S2019, respectively, were observed during this technical time slot. PCCS2G208.07 is a Planck compact source of the newIRAM_NIKA2_1mm catalogue prepared by Bilal long ago, addressed to densify the pointing (and possibly, flux calibration) source catalogue for the instrument Users community. It’s an AGN classified as a blazar, with reported variability at least since 2016 (it can be noticed in the comparison frames too). The second one is a known nearby starburst galaxy. The rms estimation given at the top of the plots below was obtained from the statistics on a pseudo-annular background polygon limited by 60% and 80% KIDs coverage isocontours in each case, for comparison purposes. For a given source, exposure time and map geometry are the same for each epoch. It seems clear that array 2 is in good shape (as with maps at 1.15 mm):
Just at the end of this observing shift, we obtained some timelines and noise traces for each array with the window open and after close it. As shown below, traces seem to be again as expected:
