Size: 4823
Comment:
|
Size: 14136
Comment:
|
Deletions are marked like this. | Additions are marked like this. |
Line 4: | Line 4: |
== December 12, (Monday) == Nico and Pablo left the telescope. We started observations at 5pm. Everything is going very fine and smoothly tonight (what a difference with yesterday's hectic night !).<<BR>> Our observing plan (Andrew and Samuel start the shift from 5pm [Nico on remote till 9pm] to 3am, then Jean-François and Frederic take over): 1. Dark test 1. Cryostat tilt angle/height vs pointing test 1. Skydips (regular + slew) 1. Nominal focus beam maps 1. @fov_focus (otf zig zag maps at different foci) 1. Tests of injection attenuation on NIKEL boards A and C (see if going from 5 dB down to 10 dB calms down instabilities at 2mm) on secondary calibrators. 1. Deep field == December 11, (Sunday) == Nicolas, Andrew, Pablo, and Samuel started the session. Not long after doing dark scans, it turned quickly to a nightmare. The feet were behaving wrong, I (Samuel) went up touching the air pressure pipes connexions without finding a solution, eventually we were able to talk to Alain to learn how to control the release valve of the feet, this solved the problem. But in the meantime we had lost the interface allowing to monitor the cryostat on the PC which is in the control room. As soon as Alain left we lost the ability to connect to the other one. We absolutely don't know what happened. While we were talking with Andrea and Juan MP to figure it out, we lost the detectors. I went in the computer room and connected a screen to the control PC which is now rejecting VNC connection. The compressor and the turbo pump were switched off! None of us did it, we don't understand ; we nearly lost the instrument, everything was warming up violently (detectors up to ~800mK). Hopefully we reacted fast enough to start again pump and compressor. It took about 1h before we were back to normal. But then we couldn't see any source and the detectors were showing new additional noise features at lower frequency. It was a problem of mirror M5 that we didn't put back in place because we were too busy with recovering the cryostat. This additional noise features, with a strong 1.3Hz peak comes clearly form the cabin.<<BR>> We finally were back to normal on all aspects, and start real observations on sky only near 3 am.<<BR>> We have at least one answer concerning the 8Hz parasite signal: the noise peak is still there, so it's not the EMIR and HERA synthesizers, which were totally switched off. During the afternoon I went in the cabin and monitored traces during EMIR observations; there was a clear correlation with telescope speed: the 8Hz was particularly strong during slewing, almost non existent otherwise. Once on sky we tried tests with various types and speeds of scans to see if we could find a correlation, but we couldn't find anything significant.<<BR>> Jean-François and Frederic took other near 3am. They continued investigation of the parasite 8Hz signal trying different scan speeds, without finding a correlation. Then secondary calibrators and 2 asteroids, and to finish a deep field. == December 10, (Saturday) == Last night we followed the plan (beam maps, pointing session, secondary calibrator...), but forgot about the dark session.<<BR>> The good news is that the cryostat seems to behave fine in terms of cryogenics and pressure in the feet. The bad news is that we discovered 3 problems: 1. The alignment that we did with Alain using the external calibrator was based on a wrong assumption: the calibrator signal does not allow to align the internal optics on the sky, because its light illuminate only a portion of the pupil. I updated the [[EffectOfExternalCalibratorBar|wiki page showing simulations of the external calibrator]], with a simulation of the light emitted by the calibrator. So we will start again tonight using the light from the sky (30K vs 300K for ecosorb). 1. We have a noise at ~ 4Hz on all data from last night. This is clearly visible at 2mm, not at 1mm. We don't know where it comes from... too bad we forgot about the dark session, we might have known if the origin is internal or external. We will look at it tonight. 1. Robert noticed that the pixel data associated with NIKEL boards 0 and 3 (2mm matrix) are particularly unstable (not only jumps but uncorrelated drifts with the sky). Does it come from the params of the tones (see file run20.ini), or is it the electronics (we could exchange cards 0 and 3 with the other 2 of the 2mm or other 1mm boxes), or the cold amps, or something else? Tonight we did: * Looking at traces and, with and without the window cap. The 4Hz parasitic signal seems to be attenuated when the cap is on. This 4Hz parasite exists only on the 2mm signals, for all 4 NIKEL lines, but it's not present at 1mm. The only thing new and common to all 2mm KID is the synthesizer, but we can't play with it when we have telescope time; we'll see tomorrow... Keeping the cap on the window we did a short dark session. * We did the pupil alignment using the ecosorb slab. We found the same result as 2 days ago. We check moving M6 gave the same result, indicating we had an over illumination of the window by the telescope pupil as expected and that the contour of rays making their path up to the array was defined by the internal pupil as expected. I was puzzled by this result, then realized that the measurement we did with Alain was the same: the spot from the calibrator on the window is much smaller than the illumination from the sky background and than the cold pupil contour, so with Alain we never the calibrator light, but only the background from the sky. * Trying things with the synthesizer, swapping NIKEL boards and other tests with electronics seemed way to risky while we have such a good sky. We will see tomorrow if we can do such test when we don't have the telescope. * We do beam maps for now... to be continued... == December 9, (Friday) == The cryostat was still not totally healthy this morning. Alain had the idea to recover some 3He from the external cold trap and re-inject it in the system. His action (plus the remote but meticulous care provided by Alessandro and Martino) finished to heal the cryostat. Since the middle of the day its behave fine and smoothly at last. Alain left the telescope. Our main priorities for tonight observations are beam maps and a real good pointing session. Also some secondary calibrators, skydip, dark session near sunrise.<<BR>> Note we did a trade with the EMIR observer (project 119-16) who has sources in the evening: for the following days we will observe roughly from 22h to 12h. This 2 hours shift gives us a bit less of Uranus, a bit more of secondary sources, in particular MWC349, and we will gain a bit of extra time around these hours. (nb: I updated some more information on yesterday report about the calibrator) |
|
Line 22: | Line 74: |
<<BR>> As foreseen we got the telescope around 15h. The time to put the calibrator bar in place the KID were ready and cold enough to start checking the cold pupil alignment. Johannes left the telescope.<<BR>> Alain and Samuel did the characterization of the cold pupil misalignment, using a slab ecosorb tapped on a plexiglass plate, sliding it from teh edge of the window toward the center and marking the position when it had an effect on the detectors response. Doing it from 4 sides allowed to determine that the cold pupil was misaligned on the window by 2mm horizontally and 10mm vertically.<<BR>> Removing the M3 mirror and using the telescope laser on the elevation axis, we tilted the M6 mirror so that the laser spot on the window rose by 10mm, in order to match the telescope pupil with the cold aperture stop footprint on the window, hence aligned the external optics with the internal optics.<<BR>> I did a Zemax [[CompensateColdOpticTiltWithM6|simulation of the compensation of the cold optics misalignment by tilting the M6 mirror]], so that you understand better what I describe here. Once this was done we did observations using the external calibrator. I did a Zemax [[EffectOfExternalCalibratorBar|simulation of the effect of the external calibrator bar on the beam and on the pupil]] (illumination). As ecpected it has an effect that we can see on the maps we take on the sky. This bar is used just for the test, if one day we implement it for real, we won't use that bar.<<BR>> We noticed during the observations with the calibrator an unexpected oscillation on the time lines. It took us a while to investigate the origin of this problem. Doing tests directly in the receiver cabin we deduced that the oscillation was of mechanical origin. It seemed to be telescope oscillations, unless it was the pulse tube since the frequency matched their frequency, but without knowing how the coupling with the calibrator could exist.<<BR>> At 2h in the morning we completely removed the calibrator. We hope we have enough data to characterize it, although the mechanical oscillations plus instabilities of the cryostat might degrade the data.<<BR>> For the rest of the session we did more classical observations: small pointing session, beam maps, CWLEO, K3-50A, and NGC6946 (face on galaxy). See day to day logbook for more information. Weather conditions remain excellent: clear sky and 225GHz opacities 0.1-0.2. |
Back to the NIKA2 run 7 main page
Contents
December 12, (Monday)
Nico and Pablo left the telescope.
We started observations at 5pm. Everything is going very fine and smoothly tonight (what a difference with yesterday's hectic night !).
Our observing plan (Andrew and Samuel start the shift from 5pm [Nico on remote till 9pm] to 3am, then Jean-François and Frederic take over):
- Dark test
- Cryostat tilt angle/height vs pointing test
- Skydips (regular + slew)
- Nominal focus beam maps
- @fov_focus (otf zig zag maps at different foci)
- Tests of injection attenuation on NIKEL boards A and C (see if going from 5 dB down to 10 dB calms down instabilities at 2mm) on secondary calibrators.
- Deep field
December 11, (Sunday)
Nicolas, Andrew, Pablo, and Samuel started the session. Not long after doing dark scans, it turned quickly to a nightmare. The feet were behaving wrong, I (Samuel) went up touching the air pressure pipes connexions without finding a solution, eventually we were able to talk to Alain to learn how to control the release valve of the feet, this solved the problem. But in the meantime we had lost the interface allowing to monitor the cryostat on the PC which is in the control room. As soon as Alain left we lost the ability to connect to the other one. We absolutely don't know what happened. While we were talking with Andrea and Juan MP to figure it out, we lost the detectors. I went in the computer room and connected a screen to the control PC which is now rejecting VNC connection. The compressor and the turbo pump were switched off! None of us did it, we don't understand ; we nearly lost the instrument, everything was warming up violently (detectors up to ~800mK). Hopefully we reacted fast enough to start again pump and compressor. It took about 1h before we were back to normal. But then we couldn't see any source and the detectors were showing new additional noise features at lower frequency. It was a problem of mirror M5 that we didn't put back in place because we were too busy with recovering the cryostat. This additional noise features, with a strong 1.3Hz peak comes clearly form the cabin.
We finally were back to normal on all aspects, and start real observations on sky only near 3 am.
We have at least one answer concerning the 8Hz parasite signal: the noise peak is still there, so it's not the EMIR and HERA synthesizers, which were totally switched off. During the afternoon I went in the cabin and monitored traces during EMIR observations; there was a clear correlation with telescope speed: the 8Hz was particularly strong during slewing, almost non existent otherwise. Once on sky we tried tests with various types and speeds of scans to see if we could find a correlation, but we couldn't find anything significant.
Jean-François and Frederic took other near 3am. They continued investigation of the parasite 8Hz signal trying different scan speeds, without finding a correlation. Then secondary calibrators and 2 asteroids, and to finish a deep field.
December 10, (Saturday)
Last night we followed the plan (beam maps, pointing session, secondary calibrator...), but forgot about the dark session.
The good news is that the cryostat seems to behave fine in terms of cryogenics and pressure in the feet. The bad news is that we discovered 3 problems:
The alignment that we did with Alain using the external calibrator was based on a wrong assumption: the calibrator signal does not allow to align the internal optics on the sky, because its light illuminate only a portion of the pupil. I updated the wiki page showing simulations of the external calibrator, with a simulation of the light emitted by the calibrator. So we will start again tonight using the light from the sky (30K vs 300K for ecosorb).
- We have a noise at ~ 4Hz on all data from last night. This is clearly visible at 2mm, not at 1mm. We don't know where it comes from... too bad we forgot about the dark session, we might have known if the origin is internal or external. We will look at it tonight.
- Robert noticed that the pixel data associated with NIKEL boards 0 and 3 (2mm matrix) are particularly unstable (not only jumps but uncorrelated drifts with the sky). Does it come from the params of the tones (see file run20.ini), or is it the electronics (we could exchange cards 0 and 3 with the other 2 of the 2mm or other 1mm boxes), or the cold amps, or something else?
Tonight we did:
- Looking at traces and, with and without the window cap. The 4Hz parasitic signal seems to be attenuated when the cap is on. This 4Hz parasite exists only on the 2mm signals, for all 4 NIKEL lines, but it's not present at 1mm. The only thing new and common to all 2mm KID is the synthesizer, but we can't play with it when we have telescope time; we'll see tomorrow... Keeping the cap on the window we did a short dark session.
- We did the pupil alignment using the ecosorb slab. We found the same result as 2 days ago. We check moving M6 gave the same result, indicating we had an over illumination of the window by the telescope pupil as expected and that the contour of rays making their path up to the array was defined by the internal pupil as expected. I was puzzled by this result, then realized that the measurement we did with Alain was the same: the spot from the calibrator on the window is much smaller than the illumination from the sky background and than the cold pupil contour, so with Alain we never the calibrator light, but only the background from the sky.
- Trying things with the synthesizer, swapping NIKEL boards and other tests with electronics seemed way to risky while we have such a good sky. We will see tomorrow if we can do such test when we don't have the telescope.
- We do beam maps for now... to be continued...
December 9, (Friday)
The cryostat was still not totally healthy this morning. Alain had the idea to recover some 3He from the external cold trap and re-inject it in the system. His action (plus the remote but meticulous care provided by Alessandro and Martino) finished to heal the cryostat. Since the middle of the day its behave fine and smoothly at last.
Alain left the telescope.
Our main priorities for tonight observations are beam maps and a real good pointing session. Also some secondary calibrators, skydip, dark session near sunrise.
Note we did a trade with the EMIR observer (project 119-16) who has sources in the evening: for the following days we will observe roughly from 22h to 12h. This 2 hours shift gives us a bit less of Uranus, a bit more of secondary sources, in particular MWC349, and we will gain a bit of extra time around these hours.
(nb: I updated some more information on yesterday report about the calibrator)
December 8, (Thursday)
The cryostat recovery operation by AM MC AB seems to have worked, and we foreseen to reach base temperature in the early afternoon. Carsten agreed to give back to NIKA2 a bit of the time we let to EMIR last night by allowing us to start as soon as the cryostat cold.
The cryostat still show a non optimal behaviour. Nothing too serious hopefully. We added some 4He gas in the dilution mixture, to improve the cooling and calm down the strange temperature behaviours.
The plan for today once we have the telescope and till tomorrow morning is:
- Check the alignment of the pupil.
- Adjust mirrors and cryostat position to have the best alignment. Use possibly the laser of the telescope, removing the M3 mirror.
- Optimize detectors parameters if needed.
- Go on a strong source to do pointing-focus, then a series of few pointings to check the offsets.
- Install the external calibrator and characterize it, by doing several types of scans (pointing, focus, beam maps).
- Find the best XYZ focus.
- Make a pointing session.
- Do beam maps.
- Auxiliary but important scans: dark test, skydip, to be inserted anywhere between items 3 and 8.
As foreseen we got the telescope around 15h. The time to put the calibrator bar in place the KID were ready and cold enough to start checking the cold pupil alignment. Johannes left the telescope.
Alain and Samuel did the characterization of the cold pupil misalignment, using a slab ecosorb tapped on a plexiglass plate, sliding it from teh edge of the window toward the center and marking the position when it had an effect on the detectors response. Doing it from 4 sides allowed to determine that the cold pupil was misaligned on the window by 2mm horizontally and 10mm vertically.
Removing the M3 mirror and using the telescope laser on the elevation axis, we tilted the M6 mirror so that the laser spot on the window rose by 10mm, in order to match the telescope pupil with the cold aperture stop footprint on the window, hence aligned the external optics with the internal optics.
I did a Zemax simulation of the compensation of the cold optics misalignment by tilting the M6 mirror, so that you understand better what I describe here.
Once this was done we did observations using the external calibrator. I did a Zemax simulation of the effect of the external calibrator bar on the beam and on the pupil (illumination). As ecpected it has an effect that we can see on the maps we take on the sky. This bar is used just for the test, if one day we implement it for real, we won't use that bar.
We noticed during the observations with the calibrator an unexpected oscillation on the time lines. It took us a while to investigate the origin of this problem. Doing tests directly in the receiver cabin we deduced that the oscillation was of mechanical origin. It seemed to be telescope oscillations, unless it was the pulse tube since the frequency matched their frequency, but without knowing how the coupling with the calibrator could exist.
At 2h in the morning we completely removed the calibrator. We hope we have enough data to characterize it, although the mechanical oscillations plus instabilities of the cryostat might degrade the data.
For the rest of the session we did more classical observations: small pointing session, beam maps, CWLEO, K3-50A, and NGC6946 (face on galaxy). See day to day logbook for more information.
Weather conditions remain excellent: clear sky and 225GHz opacities 0.1-0.2.
December 7, (Wednesday)
Alain fixed a few issues during the night. I'll let him explain in details, I think he already did it by emails to the relevant people. He also performed an alignment test by moving a sheet of ecosorb in front of the entrance window and found that there indeed was a problem of alignment. by a few millimeters. This will probably be the first thing we address this afternoon when we take the antenna.
Short summary of Alain's mail about the events of the night (Tuesday to Wednesday):
Cryostat cooling too slow: we miss He mixture, this can be deduced from the fact that the still got liquid when the pressure dropped below 2 bars.
NIKEL boards didn't accept to do tuning while the external calibrator is running. → We found out latter during the day that it was due to a "funnel" in the network structure: the 3 crates were plugged to a small switch with a too low bit rate, plugging them directly on the big switch which has the optical link to the computer room solved the issue.
One of the 3 crates crashed, only a power reboot allowed to recover it.
The new synthesizer for the 2mm band showed the same problems as the previous one. → It turned out that the problem was not the synthesizer but the board which allow controlling it.
Quick test on the pupil at the entrance of the cryostat show indication of a misalignment. To be repeated more properly next day.
Around 16h the cryostat had a new blockage, most probably a follow up from the blockage wee had the day before: the dirt that did the 1st blockage must have migrated a bit and blocked again the circuitry yesterday. Alessandro, Martino, and Alain warmed up and cooled down again the cryostat in a special way to trap the dirt in a harmless section of the circuitry.
We had a number of other events. In particular a loss of network that nearly killed the run, because it put the "automate" (electronic box which contain the devices controlling the cryostat) in a wrong state with bad consequences on the cooling. Hopefully we recovered from this. More info probably a bit latter.
December 6, (Tuesday)
Alain, Johannes, Nicolas, Frédéric, Jean-François, Andrew, and Samuel arrived at the telescope.
Alain, Johannes Samuel with the help of Dave and Gregorio spent most of the day in the cabin, installing and testing the new external calibrator. Installation on a horizontal bar fixed in the vertex cabin (between M2 and M3 after passing the M1 vertex). We tested it using the EMIR receiver, plugged on a spectrum analyser. After solving some minor problem it worked. We also changed the synthesizer of the 2mm band (a new generation with an option that caused us some trouble in the previous run) with an older one that Alain brought in his baggages. Nico worked on the Pipeline. Alessandro and Martino took care of the cryostat remotely (need to warm up a bit to take care of a blockage in the dilution unit, then cool down again). We foresee to reach the base temperature near 2am. In the mean time we make a better characterization of the calibrator using the EMIR receiver with its FTS backend: we take the spectra for each of the line emitted by the calibrator (139.5 GHz, 155 GHz, 170.5 GHz, etc.).