System AIV at La Palma

• On spectral flat-field calibrations we discovered modulations of the continuum of almost 30% of the flux and periodicity of 65 pixels. Since they were not present in Geneva, we deduced that it had to be produced by parallel plates interference in the FEU light path of the calibration light. One component in the parallel beam is the ND-wheel. After removing the ND-wheels, the most of these modulations disappeared. A residual pattern of about the same periodicity but 5% modulation was still visible. we deduced that it must be due to the glass plate holding the pupil mask in the calibration-light path. Today, the FEU was modified to remove the mask and change the location of the ND-wheel in the convergent light beam. After installation of the modified calibration-light system, the interferences have completely disappeared. 
• The fibers from the calibration unit to the FEU were inverted and caused the wrong fiber to be illuminated. We exchanged the fibers and corrected the labels.

• Exposure meter was installed and aligned. Values are of the order of 100'000 cps for the ThAr (140'000 cps for fiber A and 80'000 cps for fiber B). The difference is probably due to the fact that there is not ND-filter on fiber A and to some efficiency of the spectrograph fibers themselves. Some cross-talk of about 1% between the channels was observed.
• The spectrograph was prepared for vacuum and is now to be closed and evacuated. This will be done on April 11th, however, in order to avoid interference with on-going tests, since the focus will have to be adapted when going in vacuum.
• Investigating the diffusion effect on the CCD in the red orders: The suspicion that the fiber cladding of the octagonal fibers may transmit light could be discarded by making tests using lab fibers (exactly the same results). The fact that we have cross-talk on the exposure meter of 1% (HARPS-S 0.1%) may indicate to the possibility that the 'diffusion'  comes from the parabola or the multilayer coating. This possibility and possible solutions are being investigated. Another cause could be the gradient coating on the chip.

• Naidu and Charles have tested and characterized the CCD. The programmables gains were set to 1, 1.7, 2.66 and 3.33. The measured gains for the two available modes have been measured to been 1.73 and 1.0 for the 'fast' and the 'slow' read-out modes, respectively. The measured RON is of about 4.5 e- and 2.5 e-, achieved after some shielding optimization.

• By today, the sequencer is running and delivering the correct and complete fits headers.
• Christophe and Nicolas are debugging the templates to make all of them pass and produce the correct frames.
• Basic functionalities of the ICS (LCU and AG) are running correctly. Some stability issues noticed. Other functionalities are implemented but have not yet been tested.
• Pointing of the telescope by the sequencer working correctly and precisely. Note that the pointing of the telescope can be as off as 30 arc sec if not 'calibrated'.
• Through-focus template needs to be developed.
• AG guider guiding correctly, but the table defining filter as a function of magnitude needs to be reviewed.
• ADC not yet tested
• Exposure meter and monitoring functions not yet tested

• Demonstration of the LCU and AG interfaces to TNG staff.
• Acquisition of calibration frames and checks of the behaviour of the FEU and CU.

• CCD tests and characterization continued. Grounding further improved, vacuum pump identified as a significant noise contributor, especially at low readout speed. It will have to be unplugged during normal calibration and science activities.
• Measurement of readout noise, gain, photon transfer curve and linearity for slow readout mode. Results are nominal.

• Communication tests between telescope, sequencer (centering) and autoguider. No more conflicts.
• Further work on the creation of FITS frames with correct filename and hierarchical FITS header gathering information from the various subsystems. To be continued.
• Update of user instructions and demonstration of the LCU and AG interfaces. Gao leaving on March 30.

• Acquisition and guiding tests performed during the first half of the night.
• Further debugging of calibration templates. Some issues with the FEU/CU configuration identified: calibration neutral density filters still to be properly adjusted, selection and on-time of ThAr lamps to be checked, "dark" CU carriage position still sending Fabry-Perot light...
• Further guiding and offload tests performed. Good long-term stability was achieved on a V=8.6 star. Tip-tilt mirror position sometimes driven to large values, but mostly stays close to the middle. Further checks of the offload commands needed. Integrated guiding image looks somewhat asymmetric, centroiding algorithm to be checked.

• CCD tests and characterization started. Careful grounding of all components.
• Measurement of readout noise, gain, photon transfer curve and linearity for fast readout mode. Results are nominal.
• Examination of LED frame shows that something is masking one of the LEDs. Visual inspection (opening a side flange on the vacuum vessel) has revealed that the culprit is the "giraffe" that was left over in front of the camera. To be removed when Michel returns.

• Implementation of a target centering procedure in the Sequencer. An external python module is called to compute the offsets between target and fiber hole and to send these offsets to the telescope.
• Creation of FITS frames with correct filename and hierarchical FITS header gathering information from the various subsystems.
• Update of the table relating stellar magnitude, neutral density filter position and fiber hole coordinates.

• Acquisition and guiding tests performed during the first half of the night.
• Tests of the target centering procedure. After some initial debugging the offsets sent to the telescope are correct and the target can now be nicely centered on the fiber hole before starting the guiding. This is an essential step towards smooth operations of the instrument!
• Some conflicts appeared between processes when sending pointing and offset commands to the telescope (several servers running simultaneously). To be solved with TNG staff today.
• Further guiding and offload tests performed. Thanks to proper centering, guiding can now start properly, quickly finds the fiber hole and stabilizes. After some debugging the offset computation formula seems to send correct offloads to the telescope. The offloads correctly bring the tip-tilt mirror back towards "zero" position. Exact movements of the telescope to be further checked, but overall stability and robustness are very encouraging.

• Final connexions and adjustments (e.g. LEDs available for flat-fielding)
• Closing of the vacuum vessel (no vacuum operation yet)
• Spectrograph fully available for tests and characterization
• Cabling work in the detector/spectrograph cabinet

• Tests and debugging of the execution sequence for templates tun_A, tun_AB, obj_A and obj_AB. The automatic configuration of the FEU and CU seems to be working well. The neutral density filter position for the calibration fibers should be recalibrated. Further tests to be done with the other templates.
• Progress on the assembly of the final FITS frames with complete FITS header gathering information from the various subsystems.
• Implementation of initialization and startup templates for the whole instrument. These can now be executed from the NSTS.

• Acquisition and guiding tests performed during the first half of the night.
• The focus was on the computation of the guiding offsets and tests of the offload commands to the telescope. The sending of the offsets to the TCS is working properly. Checks are ongoing on the computed offsets to verify that the tip-tilt mirror position is properly taken into account in the computation. To be further tested tonight.
• Tests of the ADC and neutral density filter positions are ongoing. The table relating stellar magnitude to neutral density position needs to be revised.

• Final cabling of pumping system
• 3T window adaptation
• Optical table removed from VV1 for final adjustments, gluing and screwing
• Further adjustments and checks of detector vacuum

• Tests and debugging of the execution sequence for templates tun_AB, obj_A and obj_AB. The automatic configuration of the FEU and CU seems to be working well. Further tests to be done.
• Acquisition procedure now working as foreseen (Sequencer receives acquisition image and the user clicks on the target).

• Guiding tests performed during the first half of the night.
• The main difficulty is the lack of a centering procedure once the target has been selected. At the moment guiding can only start if the target is within a 100x100 (~17x17 arcsec) pixel window around the fiber hole. For some unknown reason the pointing was repeatedly off by ~1 arcmin and there is presently no way to automatically compute and send offsets to the telescope to properly center the target. Centering the target with the Dolores acquisition GUI does also center it properly for HARPS-N. The telescope pointing model will be checked. The implementation of a centering procedure is mandatory.
• Tests of the ADC and neutral density filter positions have started. To be continued. Proper configuration of the folding mirror in the FEU for all acquisition templates should also be checked.

• Various activities around the spectrograph
• CCD tilt further adjusted to optimize focus in all parts of the spectrum
• Problem of light "diffusion" around some of the spectral orders further investigated. The problem always affects the same orders, all on the red part of the chip, and is present whatever the light source (tungsten, ThAr, FP). It does not seem to depend on flux level (no diffusion-flux correlation when varying flux level or from order to order). No periodic pattern is seen, affected orders seem to be distributed at random. The problem persists whatever readout mode is chosen (slow, fast, 1 channel, 2 channels). Investigations ongoing to determine whether the problem comes from the CCD or the optics.

• Detector cold and running
• All detector subsystems connected (controller, shutters, etc.)
• Spectrograph aligned:

1. Blaze angle adjusted such that to the right and left of each order there is equal intensity
2. Rotation of the detector around optical axis verified. Orders in the middle of CCD about parallel to border between two read-out zones. No adjustment was necessary.
3. Fiber rotated such that the two fibers are vertical on the chip. The guiding screw in the exit had to be removed and the fiber rotated by about 5 degrees clock-wise with respect to vertical axis as seen from the parabola.
4. The height of the detector was adjusted (150 µm lower) such that the reddest order is safely at 13 pixels from edge and no order is cut in the center of the CCD.
5. Tip and tilt are adjusted to achieve best image quality all over the CCD.

• Some strange diffusion effect detected on the CCD. The orders and the spectral lines seems to be broader than others at the basis of the line. Some orders in the red part of the CCD seem to be affected much more than the others. To be investigated.

• Guiding started and tested on an 8th magnitude star. First working, then a problem occurred and it was not possible to make a startguiding anymore. After some attempts, we came back to nominal situation.
• Acquisition of a V = 6.5 mag start was done. Guiding was started and a spectrum recorded: FIRST LIGHT with HARPS-N! Beautiful spectrum of expected SNR: SNR = 156 in 2 minutes with seeing of 0.7 arcsec.
• Spectrum with simultaneous Fabry-Pérot recorded
• GJ1214 (V = 14.7 mag) acquired and guiding started. Guiding working well at about 10 Hz. After introduction of a delay of 500 ms guiding images looks even better and very stable!

Thursday 15th of March

by Francesco Pepe

FEU:

• L1 fitted in the derotator
• Fiber source installed in the derotator
• Fiber head illuminated from the fiber-exit end and watched into the guiding camera
• Guiding camera focused on the fiber head using torch illuminating the fiber-head mirror
• Derotator fiber source focused on the fiber head and used to adjust mechanically the tip tilt mirror (set to middle range) to center the light with the fiber-head hole
• To be checked tomorrow: the pupil alignment (using the 'Christmas' light M1) and the ADC orientation

Fibers:

• Fabry-Perot fibers pulled from CU to spectrograph

Control SW:

• TCS-ICS communication tested for ADC and parallactic angle
• Started to integrate into one software system
• Network and hardware configuration working now
• Netgear Switch still not working (compatibility with HP?). Will be replaced by a spare TNG HP switch.

Spectrograph:

• Grism mounted in its support and cleaned
• Fixed-wall of the 3T mechanically installed and electrically connected
• Pumping system installed and pipes connected to a large extent
• Preparing the Fabry-Perot location (to be installed above the ventilation channel) to the left of the vacuum vessel
• Heating foils for the vacuum vessel feet glued in their supports

Hardware:

• Static addressing fixed on the CU and SP cabinets

Wednesday 14th of March

by Francesco Pepe

General: Again a very good day. The HARPS-N Team is great, and the TNG support fantastic. Thanks to everybody!

FEU:

• All cables between cabinet and FEU attached and mechanisms tested. Cables, mechanisms and controllers are all ok.
• Fiber head brought to the FEU and connected to it. Fiber fed with laser light from behind. Alignment looks already great.
• Tomorrow optical alignment to be completed

Control HW and SW:

• Network now working and IP mapping going on
• Requested to OG to switch all components to static IP
• Rooting for the GPS checked. Longer cable to be procured by Carlos
• TNG configured the network switches for the DU and SP cabinets
• All temperature sensors of the spectrograph wired and connected to controllers. All fine.
• N-TCS integration carrying on
• Autoguider running for the optical alignments of the FEU; therefore it will be difficult to make remote tests tomorrow.

Fiber run:

• Fiber run performance checked: efficiency, FRD, etc. Everything is as in Geneva.
• Fiber pulled from 1T to the FEU. Fiber length is perfect, no issue.

Spectrograph:

• Final alignment of the optical bench with respect to the detector and distance between detector head and cryostat adjusted
• Mechanical assemblies mounted on the bench (lens mounted, parabola support, etc.)
• Started preparing the boxes for shipping back some of the OG material
• Small mirror for the exposure meter being prepared. Mask glued on the sides
• Preparation of the grism support started

DRS and scientific Software:

• Telecon to discuss archiving and installation concept for DRS and NSTS
• ICS machine installed and ready to receive control software

Tuesday 13th of March

by Francesco Pepe
General:

• Arrival of Andy Born, Andy Vick, Francesco Pepe in the early morning

FEU:

• All tilt and y-z alignments done. All ok, apart from a decentering in z on the alignment target on the FEU (about 2.5 mm lateral misalignment). May be adjusted with M4, to be checked whether critical. Impact to be estimated by ray tracing.
• Even without any internal alignment and the tip-tilt switched off, the reference beam falls at 1 mm from reference hole in dummy fiber head! 
• Tomorrow moving and connecting electronics before final alignment
• Derotation tested using a laser beam while pointing the telescope. Beam stayed absolute stable.
• Ethernet cables pulled to the derotator and tomorrow sockets will be mounted by TNG at the FEU cabinet

Electronics hardware and software:

• Bad switch exchanged. It is possible to remote log-in on the ICS machines now.
• Parallactic angle directly provided by TNG system
• Offset tests to be done to verify 'pointing model' (direction for off-loading as a function of pointing coordinates). Will save on sky time.

Fiber Run:

• Alignment of the fiber head verified and still perfect, as in Geneva.
• A rope was pulled through the building to pull back the fiber run. 

Vacuum System:

• Mechanical axis of the bench aligned with respect to the vacuum vessel motion axis.

1T:

• Heating system installed
• Room cleaned
• Wiring of the spectrograph cabinet done

Monday 12th of March

by Ian Hughes 

• Alignment of the front end unit in rotation has been done. Rx and Rz are now ok.
• The rotation of Ry still have to be corrected of 0,3°.
• Facility to redefine fiber hose on guiding camera has been done.
• Fibers for network are all connected except for the switch.
• Electric installation is now completed in 2T and nearly finished in 1T.
• 3T has been unloaded for the container and been placed in 2T at the right position to determine the correct position of the fixed wall. This for drilling the floor before cleaning.
• 3T is now stored under the Gazebo. The fans installation had started.
• The optical table has been placed in the vacuum vessel for a first alignment.
• 2T and the Gazebo has been completely cleaned and we are now in "clean room" configuration.

Sunday 11th of March

by Ian Hughes

• The fixed wall of 3T is in 2T and ready to be fixed.
• The vacuum vessel has been moved in place and been adjusted regarding the THK rails.
• The vacuum vessel has been put back at atmospheric pressure and opened.
• Ones the vacuum vessel was opened, an other check between the inside rails and the THK rails has been done.
• The optical table has been brought inside 2T, but still in the Climafol envelope to protect from dust. For that, the centre of the vacuum vessel has been removed outside 2T and brought back in. (the hoist has been used intensely today !)
• The roof of 2T is now closed.
• The cover of the vacuum vessel (VV3) is stored under the "Gazebo".
• The electric installation in 2T and the network is finish. Thank you to all the TNG team for they great job and they support during the week-end !!!
• The heating feet system is cabled.

Saturday the 10th of March

by Ian Hughes

Friday the 9th of March

by Ian Hughes 

Telescope side:

• The focus of the auto-guiding camera has been checked.
• The Front End Unit has been closed and all the checks have been done previously. No issues.
• The Front End Unit has been placed from horizontal to vertical position and has been installed on the frame. No issues. (see pictures form David)
• The rotation has been done to checked the clearance. No issues.
• The mounting brackets have been removed.
• Some measurements have been done at the end of the day. The XYZ position is in 1 or 2 mm. The tilt is in 0.7° for Rz and 1.4° for Rx.
• David and Angus are back on Monday.

Spectrograph side:

• OG team arrived at 10:30 am at TNG, just after the container CT2.
• Nothing inside the container has moving during the transportation. Everything is in good condition.
• The container CT2 is empty except for the vacuum vessel. (This will be done tomorrow)
• The carriage of the hoist has been installed on the beam. (The hoist will be hanged tomorrow)
• A few boxes has been unpacked and the equipment stored in different places.
• The detector unit has been unpacked and placed on the table in the rest room, ready to be connected tomorrow.
• The second and last coating of paint on the floor of 2T has been done by Carlos from TNG.