Home

LEONARDO da VINCI

Contents

1. LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 60 44 Template based operation of LdV 2 17 45 Instrument state change 2 18 46 Instrument mode setting 2 18 47 Stellar Interferometer user interface 2 18 1 48 Engineering mode GUI 2 18 6 49 Individual Instrument parameters setting 2 18 7 50 HAWAII window shape 2 4 7 51 Fringes search 2 12 1 52 Mode switching time 2 8 53 Preheating of the thermal source 2 12 2 54 Idle as default setup 2 8 55 LdV division in units 2 3 56 LdV opto mechanical elements 2 4 57 Instrument states 2 6 58 Graphical User Interface 2 18 59 Archived data 2 11 2 60 ICS DCS commands and answers logging 2 6 61 Engineering mode 2 7 62 No OPD offsets when using FSU 2 9 5 VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 61 2 20 SECOND GENERATION UPGRADES In this section are described the functions and capabilities that are not necessary for VINCI to work but which would improve its productivity and ease of use They could be implemented as an improvement to the current concept after the first fringes have been obtained It is important to keep the necessary options opened in the software design 2 20 1 Automated Injection Optimization In this section two possible algorithms for automatically peaking the alignment of the star into the fiber are described 2 20 1 1 Fast image scan algorithm This algorithm will
2. Computes the gain factor and the read out noise for each pixel weights replicate 1 0 n elements means 0 0 results fltarr 64 64 2 gain fltarr 64 64 readout noise fltarr 64 64 print Computing gains and readout noises for each pixel for i 0 63 do begin for j 0 63 do begin results i j regress transpose means i j variances i j weights yfit const RELATIVE WEIGHT gain i j 1 results i j 0 readout noise i j sqrt variances 0 i j gain i j endfor endfor print Gains print gain 1 10 1 10 print RON print readout_noise 1 10 1 10 median RON median readout noise VLT TRE ESO 15810 2330 1 0 5 October 2000 44 LISA Test Report median gain median gain print Median RON print median RON print Median gain print median gain Computes the Power Spectral Density PSD for each pixel and each light intensity print Power Spectral density for each pixel fourier complexarr 6 64 64 256 psd fltarr 6 64 64 129 for n 0 5 do begin for i 0 63 do begin for j 0 63 do begin fourier n i j fft cube n i j 1 for k 0 128 do begin psd n i j k abs fourier n i j k 2 endfor endfor endfor endfor Computes the mean PSD for each pixel over the different light intensities print Mean PSD for each pixel over the different light levels mean psd per pix fltarr 64 64 129 for i 0 63 d
3. 2 6 INSTRUMENT STATES From the user point of view the instrument shall be in one of the four states specified in this paragraph Req 57 VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 16 Number Name Power Moving Functions Reference LISA Status Sources TCCD logging 1 Off Off Motors and encoders off Off Off Off 2 Loaded On Motors and encoders off Off On Active 3 Standby On Motors and encoders on Off On Active 4 Online On Some motors off after Any On Active reaching position encoders on Transition possibilities between instrument states 1 gt 2 2 gt 1 3 3 gt all 4 gt all Power up down from to state 1 or 2 of the whole instrument including CCD will be done manually Transitions between conditions 2 4 will be performed under SW control When LdV is set to state Online the motors whose positions are monitored by differential encoders are initialized by sending them to their references The instrument has achieved the Loaded state only when all the real motor positions are available to the ICS either read from absolute encoders or set to the reference The switching from Loaded to Online can then be done No particular starting position is required by the instrument design in the Online mode The user has the possibility to choose any setup once the instrument is Online A default starting setup for the
4. OBSERVATOIRE EUROPEEN AUSTRAL UNIVERSITE PARIS VII DENIS DIDEROT OBSERVATOIRE DE PARIS MEUDON DESPA THESE pr sent e pour obtenir le dipl me de DOCTEUR DE L UNIVERSIT PARIS VII DENIS DIDEROT SP CIALIT ASTROPHYSIQUE ET TECHNIQUES SPATIALES par PIERRE KERVELLA INTERF ROM TRIE OPTIQUE AVEC LE VLT APPLICATION AUX ETOILES C PH IDES VOLUME II DOCUMENTS Soutenue le 14 Novembre 2001 devant le Jury compos de M Daniel ROUAN Pr sident M Pierre LENA Co Directeur de th se M Andreas GLINDEMANN Co Directeur de th se M Denis MOURARD Rapporteur M Stephen RIDGWAY Rapporteur M Vincent COUDE DU FORESTO Examinateur Photo de couverture Observatoire de Paranal depuis le NTT Peak mars 2001 Table des Mati res 1 Introduction 2 LdV Software User Requirements 3 LISA Test Report 4 LdV Precision and Sensitivity 1 Introduction Ce volume regroupe trois documents en langue anglaise crits lors de mon travail de th se et qui ont t r f renc s dans le volume principal Ils pr sentent une approche plus d taill e et plus technique de l instrument De mani re ne pas alourdir le document principal du m moire de th se ils sont reproduits s par ment dans ce second volume Les deux premiers documents Sections 2 et 3 pr sentent le fonctionnement de l instrument VINCI Le premier VINCI Software User Requirements Section 2 donne les sp cifications utilis
5. 7 BEST PHOTOSITES SELECTION 7 1 PRINCIPLE To select the best possible photosites for the light detection it is necessary in principle to measure the noise level of all the pixels in the frequency range used by the DCS The detector useful readout frequency range is about 2 2000 Hz It is possible to use the special 64x64 readout mode of LISA to have a sampling rate of about 30 Hz on every pixel in the selected window in double correlated mode It is reasonable to assume that the noise level of the LISA camera does not depend on the readout frequency of the detector this assumption was checked a posteriori and is valid The selection can therefore be done in two steps 1 first construct a map of the lower left corner 64x64 region indicating the bad and noisy pixels Using the 64x64 readout mode gives access to the power spectral density up to a frequency of 14 3 Hz 2 do a finer selection of the best pixels in this area by acquiring a series of scans at the highest possible frequency In this case the maximum reachable frequency in the power spectral density can be of more than 1000 Hz For these tests the data was acquired using the control panel of the camera and saved into standard FITS format data files The data processing was done using the IDL software including the ASTRON package 7 2 PIXEL GAINS 7 2 1 Measurement procedure To obtain the values of the gains for all the pixels the detector was exposed to an increa
6. see Table 2 for test 10 the most significant This result is about the same as for the best focus obtained previously tests 6 and 7 It is suspected that the previous tests were done with the detector in the saturated regime on the strongly non linear part of the sensitivity curve but the focusing could fortunately still be achieved correctly in this regime Doc WVLT TRE ESO 15810 2330 Issue 1 0 Date 5 October 2000 Page 13 LISA Test Report Table 2 Result of the in focus test on LISA Energy Comments Infinity in one pixel 545 33 38 50 3 Black 4 904 Collimated beam suspected saturation e EE background 10 545 Black 4 904 Short exposure time 64x64 readout mode p no saturation plus fixed pattern removal BEST EVALUATION 4 5 4 Optical efficiency Due to electronic leaking of the target pixel into its neighbors only 5896 of the energy brought effectively by the photons into this pixel are present in electronic form for readout ref study by Gert Finger on the HAWAII chip The other 4296 are present in the four neighbors of the target pixel in the fast and slow shift register directions horizontal and vertical 1596 of the energy is lost in each neighbor pixel along the fast shift register and 696 in the neighbor pixels along the slow shift register Therefore the 56 measured with the detector in focus test 10 correspond to a very good optical efficiency 4 6 PROBLEMS 4
7. 5 3 o V 1 e T s The resulting instrumental visibility as a function of the exposure time is plotted on Figure 2 The fastest integration time for the whole interferogram in VINCI will be 7 milliseconds effective fringes length 70 microns 10000 microns s fringe speed This gives a maximum visibility measurement precision per scan of 0 0426 The length taken into account is not the total scan length but only the length of the fringes packet only 70 microns including the secondary lobes The piston has no effect on the flat part of the interferograms 1 0000 0 9500 0 9000 0 8500 0 8000 instrumental visdbiliity 0 7500 0 7000 0 50 100 150 200 250 300 350 400 Exposure time milliseconds Figure 2 Instrumental visibility as a function of scan duration 5 1 2 FLUOR IOTA experimental results A piston noise value of 0 2 multiplicative noise is what is generally observed on IOTA on the 40 meters baseline and for bright stars Typically the signal to noise ratio of the squared instrumental visibility Au u 0 2 on the bright sources with a fringe speed of 500um s and a 114 microns scan length This means that the piston noise equivalent power is one fifth of the star signal Therefore the precision obtained on the individual measurements of u is assymptotically 20 assuming no other source of noise This translates into a 1 precision limit on the visibility measurements in 100 scans ar Sens
8. S sqrt total noise fraction thermal n thermal 2 dt df total noise fraction piston n piston 2 total noise fraction detector n detector 2 dt df total noise fraction photon n photon 2 dt df total noise
9. n scans n elements flux 0 0 print Means and variances means fltarr n pix variances fltarr n pix print for n 0 n pix 1 do begin means n mean flux n variances n variance flux n endfor print flu0x 0 75 0 20 2 DH Computes the Fourier Transform for each pixel and each scan print Fourier transform for each pixel and each scan fourier complexarr n pix n frames n scans psd fltarr n pix n frames 2 1 n scans for i 0 n pix 1 do begin for j 0 n scans 1 do begin fourier i j fft flux i j endfor endfor Computes the Power Spectral Density PSD for each pixel and each scan print Power Spectral density for each pixel and each scan psd fltarr n pix n frames 2 1 n scans for i 0 n pix 1 do begin for j 0 n scans 1 do begin for k 0 n frames 2 do begin psd i k j abs fourier i k j 2 endfor endfor endfor Computes the mean Power Spectral Density PSD for each pixel over the scans median psd scans fltarr n pix n frames 2 1 print Mean PSD over the scans for i 0 n pix 1 do begin for j 0 n frames 2 do begin mean psd i j mean psd i j endfor VLT TRE ESO 15810 2330 1 0 5 October 2000 46 LISA Test Report endfor end 11 4 PLOT 64x64 DATA pro data_plot means variances gain readout_noise fourier psd median_psd SYNTAX data plot means variances g
10. when it is at its nominal position on the on axis parabolae INA1 distance from the autocollimator 2000 mm and INB1 distance from the collimator 1500 mm Pupil position Additional lens position 1320 mm to 3900 mm Inserted 3900 mm to infinity Removed Two preset focus positions are selectable Req 6 one for the infinity focus without TLENS PRESET1 this is used in the PupilCheck ArtificialStar setups and the other for the focus in the laboratory with TLENS at the foreseen distance of the pupil image projected by the VCM PRESET2 After the rough focus has been achieved the fine focus procedure can be done if necessary Pupil Check Artificial VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 22 COMU INA1 MOTORS pec PONT emunroeus Di D D COMU INA3 FAST SCAN N A N A N A N A N A su S oec ea EC COMU OUT1 MOTORS A TIP TILT FOCUS ROTATION COMU POLA A POLAB INA N A N A N A N A MOTORS ALIU ALI SLIDE NOMIRROR AL ALI4 ALI3 ALI4 POSITION ALIU TCCD LENS POSITION ALIU TCCD FOCUS N A ADJ ADJ ADJ ADJ PRESET2 PRESET2 PRESET1 PRESET1 ARTU LEONARDO light N A N A N A ES source 2 8 5 Image Check The artificial star is off The setups are the same as the Pupil check except for the additional lens which is not used The pre position PRESET1 for the TCCD focus is the focus at infinity This mode and all the
11. 11 Data content in the flux data mode is not the four windows total fluxes are correct No correct for the two additional channels 12 Presence of a strong frequency peak in Interference with electroluminescent light tubes Yes the noise at a frequency of 100 Hz through power supply TBC 13 Presence of low frequency peaks in the Ground loops producing waves on the Yes camera noise detector TBC Table 8 Points to be clarified Question Suggested solution Solved 1 Structure of the data produced by IRACE The DCS to IRACE SW interface document Yes not clear should be completed 2 Possibility to save a data cube of 64x64 Yes images for photosites selection 3 Effective exposure time and frame time Modify dictionnary Yes are not available in the keywords 4 3x3 windows generation and readout Generate and check clock paterns Yes still pb VLT TRE ESO 15810 2330 1 0 5 October 2000 42 LISA Test Report 11 APPENDIX IDL SIGNAL PROCESSING PROCEDURES 11 1 READ DATA pro read data flux pro read data cube flux Reads the test data from LISA version 13 09 2000 Reads the data cubes into the memory sprint Data Files are being read into memory cube 0 full mrdfits dark 1 DIT fits 0 cube 700 full mrdfits level 700 1 DIT fits 0 cube 1800 full mrdfits level 1800 1 DIT fits 0 cube 4400 full mrdfits level 4400 1 DIT fits 0 cube 9000 full mrdfits level 9
12. 2 dt 4 oe The 1e6 factor is to account for the W m2 MICRON reference flux 3 n oe hermal nois quivalent power squared nu c lambda delta nu c lambda delta_lambda 2 c lambdatdelta_lambda 2 B 2 h nu 3 c 2 1 exp h nu k Temp 1 etendue E lambda 2 epsilon n thermal 2 h c lambda delta nu B etendue 2 Piston nois quivalent power squared piston l exp 1 5 1 fringes n lambda frequency 5 3 n piston S piston Detector nois quivalent power squared n detector 2 frequency RON h c QE lambda 2 Photon nois quivalent power squared n photon 2 h c lambda 2 A delta lambda 1e6 T F0 10 m 2 5 The le6 factor is to account for the W m2 MICRON reference flux Returns MINUS the SNR total noise n thermal 2 dt df n piston 2 n detector 2 dt df n photon 2 dt df snr S sqrt total noise ar Spi VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 34 of 35 11 4 NOISES CONTRIBUTIONS NOISES CONT function fraction piston fraction detector fraction photon noises cont scan_speed magnitude tel escope This function computes the contributions of each source of noise version 17 05 00 oe d oe d i These parameters change with the telescope Siderostat without Beam Compressors if telescope 0 oe A 0 084 Area of the telescope m2 T 0 00612 Photomet
13. VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 41 Observing Night Observation Files Detailed Content Block 2 Block 3 Block 10 min Star parameters Instrument configuration VLTI configuration TCCD images pointers LISA full frame pointer Software parameters Hardware parameters Environmental data Obs parameters t 2 20290 Obs parameters Obs parameters 2 12 DESCRIPTION OF THE OBSERVATION PROCEDURE 2 12 1 Observation Procedure with VINCI This section gives an overview of a typical observing session 1 Observations Scheduling the observations scheduling is conducted off line from the real time operating system The creation of a formatted observing list should be possible before on site arrival through a front page software This system could help the observer determine e which sources are visible e the previous observations of his targets e the magnitudes including the K magnitude if necessary computed from the other bands measurements e the estimated visibility e the calibrators used so far e the baselines used so far on this target The querying of this database is possible through every field and it is possible to use the database while on the observing site The calibrators selected by the observer with the help of dedicated software are integrated in the list of the observations to be conducted Their nat
14. WG2X1 Focus of parabola ART1 WG3F1 WG4F 1 WG5F 1 INA2 FIBER Observation WG1X1 WG2F2 Focus of parabola INA1 INB2 FIBER Observation WG2F6 WG3F6 Focus of parabola INB1 MONA IN 1 FIBER Observation WG1X2 WG2F4 Injection of light from the in A BEAM WG3F2 side of MONA beam A MONA IN 2 FIBER Observation WG1X2 WG2F5 Injection of light from the in B BEAM WG3F3 side of MONA beam B MONA OUT 1 FIBER A Observation WG2X1 WG3F4 Injection of light from the BEAM out side of MONA beam A MONA OUT 2 FIBER B Observation WG2X1 WG3F5 Injection of light from the BEAM out side of MONA beam B OUT1 FIBER Observation WG2F6 WG2F7 Output of the MONA box 4 WG3F6 WG3F7 fibers bundle COMA3 POSITION Night IN OUT Present in normal operations COMB3 POSITION Night IN OUT Present in normal operations VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 64 ALI9 POSITION Night IN OUT Used only for engineering 2 20 5 Photometric Calibrations for the TCCD At the beginning of the night it would be interesting to obtain flat field images of the bright sky with the TCCD in order to calibrate the images obtained during the night Once the object image has been acquired a dark field should be acquired of the same exposure duration as the object exposure to be substracted from the object image Moreover it is desirable to obtain o
15. both at the entrance and output of the fiber The theoretical maximum eficiency is 96 Four fiber interfaces are foreseen on VINCI between the fiber input LdV Precision and Sensitivity VLT TRE ESO 15810 2177 1 0 12 July 2000 8 of 9 and the detector resulting in a transmission of 85 Antireflective coatings may cancel this effect but their implementation is not yet certain on VINCI 3 5 FIBER INJECTION DYNAMIC LOSSES Table 6 Tip tilt residuals and flux losses Telescope UTs without AT AT Beam Siderostats AO Expander Arcsec sky RMS 0 025 0 032 0 103 0 58 of K band diffraction limit 45 13 42 53 Fraction of flux transmitted 0 80 0 95 0 80 0 70 Due to residual turbulent motions of the star image on the fiber head the mean fraction of flux injected is lower than the theoretical 80 value The values of RMS tip tilt errors projected on the sky are listed in Table 6 taken from RD1 including STRAP for the ATs and UTs but not for siderostats The flux loss values are directly the values of the mode of the fiber at the corresponding lateral displacement of the star image RD4 p 66 The UTs with AO tip tilt residuals are included in the Strehl ratio budget Section 5 2 3 6 EFFICIENCY OF THE TRIPLE COUPLER The MONA fibered triple coupler is specified to have the following characteristics m minimum photometric efficiency of 75 m interferometric efficiency normally above 95 3 7 IMAGI
16. fiber The intensity of the background on the detector shows clearly the temperature variations of the soldering iron due to the temperature regulation loop Figure 9 For all the tests except when mentioned otherwise the K band filter green in aspect was put in front of the detector Even after shutting off all the lights in the room the background level on the detector is still very high causing the saturation of the detector even with a very faint additional light source 0 0 20 0 40 0 60 0 80 0 100 0 12000 13000 14000 535 46 m 545 38 15000 Mean over 728 pix 16000 17000 Intensity ADU 18000 19000 EE a 2 e 21000 time s VLT TRE ESO 15810 2330 1 0 5 October 2000 19 LISA Test Report Figure 9 Variation of the background due to stray light from a soldering iron about 25 degrees off axis The curves correspond to three different test sequences Tests have also been conducted with the soldering iron hidden from direct view to the detector by a metal piece but the variations are still very visible several thousand ADUs due to reflections on the laboratory walls and other optical elements 0 0 50 0 100 0 150 0 200 0 250 0 Time s Figure 10 Background with cold shutter closed Mean 19 28 RMS 2 07 ADUs When the cold shutter is closed the variations of the background signal are inexistent as visible on Figure 10 This sho
17. fiber connected to a laser diode located at the Nasmyth focus of the selected telescope on the TCCD located in the interferometric laboratory Then the position of the image of the source can be modified by tilting a mirror in the VLTI optical train TBD preferably located in a pupil plane or close to it to limit cross coupling between tilt and OPD The image alignment is expected to occur once per night or before any observation depending on the stability of the VLTI The procedure to obtain and measure the image of the Nasmyth light source with the TCCD is the following 3 Obtain an image with the TCCD in the Image Check setup with the Nasmyth multimode laser diode switched on VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 48 4 Provide the image to the user together with information on the quantity of light available and the position of the spot In the current design the role of LdV is not to provide the commands to set the mirrors of the optical train to align the pupil Only the display and shape measurements of the image are a requirement for the LdV software But this has TBC when a clear alignment procedure has been defined 2 15 5 Star Image Centering The Technical CCD is used to check the star image quality and position During observations in the stellar interferometer mode the TCCD is used in the following sequence after the telescopes are tracking on the targe
18. unvignetted to the instruments This is done by imaging the pupils of the telescopes on the TCCD chip The pupils are the images of the secondary mirrors of the feeding optical systems either siderostats UTs or ATs The delay lines system is built so as to keep the beam B pupil plane in the laboratory this allows an extended field of view for the imaging applications Normally the image of the pupil is obtained by turning on the light source at the center of the secondary mirror The image obtained on the TCCD is then a point which has to be centered relatively to the axis of the technical CCD optics The refractor in front of the TCCD defines the optical axis of the VLTI The pupil alignment is expected to occur once per night or before any observation depending on the stability of the VLTI The location of the pupil planes for the two beams will be on the injection parabolae INA1 and INB1 This means that for imaging the pupil it is necessary to focus at their distance with the TCCD As it is used also for viewing the stellar images it is normally focused at infinity VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 47 In order to focus at a few meters this requires the addition of a converging lens in front of the TCCD Once this is done the TCCD can focus on the pupil image which is situated on INA1 INB1 The OS should get the expected pupil position from VLTI OS Req 37 and fr
19. 1 0 12 July 2000 26 of 27 error Unit Telescopes with AO Siderostats without BC Siderostats with BC A viliary Telescones Magnitude K Figure 7 Precision on the calibrated visibility for a 40 m baseline VLT TRE ESO 15810 2177 1 0 12 July 2000 27 of 28 a error linit Telescopes with AO Siderostats without BC Siderostats with BC Ay iviliarv Telescones Magnitude K Figure 8 Precision on the calibrated visibility for a 195 m baseline 9 6 ATMOSPHERIC TURBULENCE AND FRINGE SENSOR IMPORTANCE The atmospheric turbulence introduces random corrugation of the incoming wavefront both in shape perpendicular to the propagation direction and longitudinal delay along the direction of propagation also called differential piston effect The adaptive optics allows the flattening of the wavefront in the perpendicular direction Once corrected the wavefront can be injected much more efficiently in single mode optical fibers This results in a higher flux used for beam combination and increased sensitivity The role of the fringe sensor unit FSU is to stabilize the fringes longitudinally to allow for longer integration times with the science instrument Normally the piston effect limits the scan time to a few tens of milliseconds The FSU removes it by quickly adjusting the internal OPD by means of a delay line The FSU could become available for th
20. 4 1 8 Median low frequency PSD over the 64x64 pixels Figure 25 shows the median PSD obtained over the 64x64 pixels area Two peaks are clearly visible but the very low frequency of these peaks make them not very problematic for the detection of the interference fringes which frequency is usually set to about 100Hz VLT TRE ESO 15810 2330 1 0 5 October 2000 34 LISA Test Report When using LdV with a fringe tracker it will be necessary to decrease the fringe frequency down to very low values for faint objects possibly interfering with these peaks In this case they could become an important problem Median Power Spectral Density over the 64x64 array dark 3x104 T T T T T T T T T T T T T T N X a I Power Spectral Density EEE x D peg 3 PUE EE Eo SERE EE EC E DR E 7 o 1 1 il L 1 1 1 l 1 L 1 1 50 Relative Fraquency Maximum pix 128 3 33 Hz 0 026 Hz bin o m a Figure 25 Median Power Spectral Density for the 64x64 pixels from 0 to 3 33 Hz X 128 The Y scale is arbitrary 7 4 2 High frequency 0 828 Hz 7 4 2 1 Data analysed This second type of measurements is done using the windowed readout mode of LISA This is necessary because the readout frequency of the camera in the 64x64 pixels mode is limited to less than 30 Hz which is too low for the study of the behavior of the detector at the frequencies used for the detection of the fringes The windows used w
21. 5 to 10 minutes The injection optimization is a hard point for the control software if it is intended to be done automatically The necessity for an operator to do it manually may not be possible to avoid A possible algorithm for the automatic injection optimization is described in the section 2 20 1 It is foreseen to achieve the complete automatization of the LdV operations and this possibility should be left open in the software design From the user point of view the parameters are the same as for the synchronized mode except that the signals are displayed in a continuous loop The only difference is that the piezo mirror might not be moving effectively if simpler from the SW point of view Only the computed pixel readout frequency also called frame rate is used effectively as a parameter The difference with synchronized mode is that the camera is read continuously without the need to wait for the synchronization of the piezo thus resulting in higher data rate The parameters used are the same as for the synchronized mode in order to check that they are good with respect to the flux coming from the star i e that the star is not saturating the detector for example Parameter Range Values Data Acq Total 1 300 microns OPD Range Data Acq 1 2500 microns s Fringe Velocity Data Acq 0 1 3 microns Sample Interval 1 10 points per fringe Once the parameters are set the user
22. 6 1 Clamping of the camera mount Currently the camera is not attached at all on the VINCI optical table It is important to foresee an addition to the current design to fasten the mount securely to the table In the current status the camera is very unstable and can be moved easily by hand 4 6 2 Sign of the ADU counts All the counts in ADU delivered by the camera when in double correlated mode are negative Higher flux gives lower negative ADU counts This was a misleading behavior and the sign of the subtraction of the two readouts in double correlated windowed readout mode was modified in the last version of the software as of 11 09 00 The 64x64 and 512x512 pixels modes were not corrected 4 6 3 Filters order in the filter wheel Counted counter clockwise seen from the front of the dewar towards the detector the correct order is the following Table 3 Filters positions in the filter wheel Position Filter 1 CLOSED OPEN K BAND green color H BAND orange color OPEN OPEN O 01 B P VLT TRE ESO 15810 2330 1 0 5 October 2000 14 LISA Test Report 4 6 4 Connector surface During these experiments the connector of the fiber was apparent on the detector as a darker spot The explanation seems to be that the detector is seeing itself in the connector polished surface therefore it appears much colder than the surroundings Generally speaking the background level is very high
23. ANN A BAAR RSR ES RCECIE f3 t8 3 3 EJ E o VD04 HIGH4 Lowa 4 60 FRMCHK4 FRMCHK3 39 RESETB4 RESETB3 e 58 READS DRAINT RER ER RS a Blo ISISISISISISISESESES SERIES SEC SSS5ESERSG DE 25 SEEEGREEEEFEEEER gtSxFs Be 8 aoa x KA 2020569 2 gt a 30 age gt 4st a aa 3 Eo ul gt a u ul a 9 o D e o o o5 Figure 6 The HAWAII array LISA detector uses quadrant IV picture from Rockwell The X axis of the LISA camera is along the fast shift register positive direction and the Y axis is along the slow shift register positive direction Vignetted Zone 100 Figure 7 LISA Hawaii detector usable part LISA Test Report b ELE EE 5 October 2000 18 Vignetted Zone 7 LONDRES 1 100 20 45 Figure 8 Zoom on the usable part of the array 5 2 BACKGROUND LEVEL During the tests of the detector the background was very apparent and variable depending on the environment in a wide cone maybe about 50 degrees angle in front of the dewar All the tests presented in this section were done using the full quadrant integration mode giving an exposure time of 2 seconds This is very long and therefore not really significant of the normal acquisition regime A test was conducted by putting a soldering iron 30 cm off the detector axis and 70 cm away from the window angle from the dewar window 25 degrees It was installed as a low intensity light source for the
24. AO Figure 5 Exposure time calculator graphical user interface 9 3 CALIBRATORS BASELINE AND SHAPE FACTOR The calibrators available in the first weeks will not benefit from the accuracy improvement procedure proposed in Section 8 1 3 Therefore the transfer function measurement accuracy will be limited for the longest baselines to about 1 For the short baselines up to 20 meters the uncertainty will be less than 0 05 The shape factor will be derived from the catalogued spectral types and is not expected to be a limiting parameter for the final precision as no very peculiar sources are foreseen during this first period 9 4 EXAMPLE THE CASE OF ZETA GEMINORUM In order to have a realistic case for the observation simulations the case of the Cepheid Zeta Gem with the calibrator star HD 49968 is considered Zeta Gem is a 1 98 magnitude star in K angular diameter of aproximately 1 8 mas HD 49968 is a 2 2 magnitude calibrator from the Cohen list with a uniform disk angular diameter of 1 87 0 030 mas Considering the longest baseline of 195 meters it will be necessary to observe for about 1300 seconds on each Zeta Gem and HD 49968 with the siderostats to get to a 2 accuracy on the visibilities This results from the fact that the correlated visibility of this star is high as it is resolved but the uncertainty on the calibrator also adds to the difficulty of the observation T VLT TRE ESO 15810 2177 LdV Precision an
25. Determine tscan_start tscan_end tDAQ_start tDAQ_end and transfer that information to VINCI ICS and LISA DCS Ti tscan start Begin scan Start generating voltage ramp TDAQ start Begin data Start acquisition recording frames TDAQ end End data Wait for Wait for End recording acquisition Delay Line in tscan end frames TRACK status Reset piezo Perform LISA in READY status to quicklook status READY analysis Piezo in READY Transfer status offset Check that one information to more scan is the delay line requested Transfer scan Determine data to VINCI tscan_start OS tscan_end Reset LISA tDAQ_start LEONARDO da VINCI Software User Requirements VLT SPE ESO 15810 1852 1 1 16 September 1999 28 tDAQ_end and transfer that information to VINCI ICS and LISA DCS Tscan_end End of piezo motion Ti 1 Begin scan Start tscan_start generating voltage ramp The time intervals Tscan start tDAQ start and DAC end Tscan end are used to accelerate and decelerate the piezo mirror These intervals are not compressible as they are defined by the hysteresis curve of the piezo itself 2 9 4 Real time considerations The critical time interval is t1 Leen eng which corresponds to the dead time during which VINCI is not acquiring data The software architecture should be designed to ideally make it zero or at least minimal Req 21 thi
26. Figure 10 Background with cold shutter closed Mean 19 28 RMS 2 07 ADUs 19 Figure 11 LISA cold mechanics and baffling seen from the front photo MPE 20 Figure 12 LISA detector support and baffling seen from the back photo MPE 20 Figure 13 Variance as a function of signal for LISA The slope is 0 1182 giving a gain factor of 2 9 electrons ADU X axis is the mean signal in ADU Y axis is the observed variance of the temporal sequence in ADU 21 Figure 14 LISA windowed mode readout order 23 Figure 15 Gain plot for pixel 31 38 G 5 1096 27 Figure 16 Gain plot for pixel 32 38 G 6 8939 27 Figure 17 Gain plot for pixel 33 38 G 6 9652 28 Figure 18 Gain map over the 64x64 pixels lower left area Pixel 0 0 is in the lower left corner X axis is to directed the right Y to the top 29 Figure 19 Histogram of the 64x64 lower left pixels gains measured at 6 66 Hz frame frequency 29 Figure 20 Readout noise map of the lower left 64x64 area at 6 66 Hz frame frequency The image grey scale is linear between 0 and 30 electrons 30 Figure 21 Histogram of pixel readout noises at 6 66 Hz frame frequency The median value is 20 18 e and the maximum number of pixels is found at 18 e There are still a number of pixels at less than 15 e 31 Figure 22 Power spectral density for pixel 31 38 from 0 to 3 33 Hz X 128 Y scale arbitrary 32 Figure 23 PSD for pixel 32 38 from 0 to 3 33 Hz X 128 Y scale arbitrary 33
27. HAWAII based infrared camera is built by the MPE Garching 1 1 SCOPE This document defines the software user requirements specific to LEONARDO da VINCI 1 2 APPLICABLE DOCUMENTS 1 LdV Technical Specifications VLT SPE MEU 15810 0002 v1 0 11 07 99 2 Interface Control Document between the VLTI and its Instruments VLT ICD ESO 15000 1826 v1 0 23 04 99 1 3 REFERENCE DOCUMENTS LdV Optical Definition VLT SPE MEU 15810 1000 v1 0 11 06 99 LdV Mechanical Design VLT SPE MEU 15810 2000 v1 0 12 07 99 LdV Sources and Guided Optics VLT SPE MEU 15810 1001 v1 0 10 07 99 LdV Electronics Design VLT SPE MEU 15810 3000 v1 0 13 07 99 Reference alignment sources and waveguides in LEONARDO VINCI Vincent Coude du Foresto 27 02 99 8 Data Acquisition in VINCI Terminology and Chronology Vincent Coude du Foresto 28 04 99 VLT Software Management Plan VLT PLA ESO 00000 0006 v2 0 21 05 92 10 Technical report on Image Processing Algorithms for TCCD systems VLT TRE ESO 17240 1689 23 10 98 11 VLTI Software Requirements Specification VLT SPE ESO 15400 0866 18 12 96 N D OR Fa Doc VLT SPE ESO 15810 1852 LEONARDO da VINCI Issue 1 1 Date 16 September 1999 Software User Requirements Page 2 1 4 ABBREVIATIONS ACRONYMS AND TYPOGRAPHIC CONVENTIONS goal min Numerical requirement goal value minimum value Req Numbered requirement see section 2 19 ADJ Adjustable ADU Analog Dig
28. Hz 1 Hz Yes DCS Visibilities plot time visibility 2 Hz 1 Hz No OS Visibilities plot settings axis Observation Yes OS Visibilities plot statistics last mean sigma 2 Hz 1 Hz No DCS Visibilities histogram visibility number of values 2 Hz 1 Hz No OS Visibilities histogram settings axis Observation Yes OS Fringes found warning signal 2 Hz 1 Hz No 2 18 3 List of Parameters displayed by the Pupil Image Check Modes GUI VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 56 Origin Description Typical Possibility Update of GUI Frequency Setting VLTI _ Seismic activity warning symbol Continuous No VLTI Time signals Local UT Sidereal Continuous No ICS Status of LdV hw warning symbol green ok red error Continuous No OS Technical CCD exposure time Observation Yes OS Current beam on the TCCD beam A beam B 1 Hz 0 5 Hz Yes DCS Technical CCD images one beam at a time 2 Hz 1 Hz No OS Computed pupil image center coordinates on the TCCD Observation No 2 18 4 List of Parameters displayed by the Artificial Star Mode GUI Origin Description Typical Possibility Update of GUI Frequency Setting VLTI _ Seismic activity warning symbol Continuous No ICS Status of LEONARDO hw warning symbol green ok red Continuous No
29. OPD OS ROF SNR SW TBC TBD TCCD VINCI WS VLT TRE ESO 15810 2330 1 0 5 October 2000 4 LISA Test Report Graphical User Interface Graphical Engineering Interface Garching Optical Laboratory Hardware Instrument Control Software Infrared Real Time Display Instrument Workstation Local Control Unit LEONARDO da VINCI the whole instrument The artificial star subsystem The HAWAII based infrared camera The LISA LCU workstation The fibered recombiner Not applicable Optical Path Difference Observation Software Raw Observation File Signal to Noise Ratio Software To Be Confirmed To Be Defined ESO Technical CCD The main optical table of LdV Workstation LISA Test Report ee 5 October 2000 5 2 DEWAR GENERAL CHARACTERISTICS 2 1 COLD TEMPERATURE AUTONOMY The measurements presented on Figure 1 have been done with an external temperature of about 20 degrees C The upper tank capacity is about 6 liters and the lower tank capacity is about 3 liters The real autonomy is 24 hours but not more LISA Nitrogen consumption 6 00 Upper tank 5 2 I day Lower tank 0 4 I day 5 00 i 4 00 a 3 00 Cd 2 e Upper tank 200 a Lower tank 1 00 0 00 0 5 10 15 20 25 Time h Figure 1 LISA Nitrogen consumption 2 2 NITROGEN FILLING The filling of the two tanks was achieved without problem following the instructions of the manual provided by MPE It
30. Online state could be chosen a good starting point could be the Stellar Interferometer Idle setup but this is not mandatory Power up following a power failure may leave the instrument in a hazardous condition so there are a number of hardware interlocks for protection of the LISA camera The motion of the motors and other mechanical systems should be stopped after a power failure The commands issued to the ICS and DCS and the corresponding replies from the ICS are logged when the instrument power is on Req 60 The mechanical design of the functions is such that the positions will be kept due to friction To reduce dissipation of motors the mechanical devices will be positioned and then the motors switched off The only remaining dissipation is that of the encoders They will be left on if the position of the related moving device has to be known to a high precision and if it is moved during the observations Which encoders will have to be left on has to be checked In the OFF LOADED and STANDBY states LdV is not directly operational The BSA and BSB cubes of LEONARDO are placed in positions OUT where they do not block the light beams as the other instruments may need to access them The optical elements on the VINCI table do not require particular positioning and should be left in the last on line state setup used VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 17 2 7 LDV ENGINE
31. POLA A POLA B beam 2 3 2 Alignment Toolkit Unit 2 3 2 1 Manually Movable Devices m o c type ALI12 ALI13 ALIS ALIA Technical CCD Autocollimation Manual Mirror ALIS 2 3 2 2 Computer Controled Devices NC IWON type ALI10 short range focusing ALI7 Technical CCD Feed Mirrors Slide 1 large translation ALI ALIS ALIA slide ALI ALI5 Technical CCD Section TCCD ALI6 24 4 TCCD DCS 2 3 3 Artificial Star Unit 2 3 3 1 Manually Movable Devices Element Control Comments type Reference Source Folding Mirrors Manual ARTB1 ARTA1 ARTA2 Stellar Beam Folding Mirrors Manual VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 10 PARTB2 ARTAZ 0 0 0 0 Artificial Star Injection On axis Manual Local Adjustment normally Parabola Motor not necessary ART1 Artificial Star Injection Flat Mirror Manual Local adjustment normally ART2 Motor not necessar Reference Source Glass Cube Manual Transmissive at 2 2 ARTA and 10 microns 2 3 3 2 Computer Controled Devices Control Comments type Artificial Star Light Source On Off Light switches K Laser ART3 Reference Source Beam Splitters Motor 1 large translation 3 positions BSA BSB 2 3 4 Infrared Camera Unit Control Comments type Main Infrared Camera See The commands are sent to the LISA OUT3 Section 2 9 LISA DCS 2 4 MOVABLE HARDWARE DESCRI
32. Q the SNR ar EE VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 31 of 32 11 APPENDIX MATLAB ROUTINES 11 1 EXECUTION BATCH The files are written in wk1 spreadsheet format for use in Excel table siderostats without bc abaque 0 wklwrite table sids witout bc table siderostats without bc able siderostats abaque 1 klwrite table sids table siderostats able ats abaque 2 klwrite table ats table ats able uts without ao abaque 3 klwrite table uts without ao table uts without ao table uts with ao abaque 4 wklwrite table uts with ao table uts with ao zZ ct x cd e Gt 11 2 MAIN PROGRAM ABAQUE function table_results abaque telescope dp table results contains the following data 1 magnitude of the star 2 resulting SNR per channel per scan 3 resulting precision for 1 scan and 1 channel 4 optimal frequency for acquisition 5 required piezo speed 6 precision for 100 scans and the two channels supposed decorrelated 7 fraction of noise due to the piston 8 fraction of noise due to the detector 9 fraction of noise due to the photon oe oe d d d oe oe oe dp dp 1 51 2 2 2e 6 Total length of the scan in m lambda 2 2e 6 Wavelength in m n se 5 Points per fringe max speed 0 01 Maximum scan speed in m s dp dp Siderostat without Beam Compressors if telescope 0 max_loop 85 31 end Siderostat
33. SE E teles 8 4 2 FOCUSING PROCEDURE tnt Tee ea tele Tente cte dete ee se relate ileal ei 9 4 3 EXPERIMENTAL RESULTS itte testae te e DR LER E DEN LEE ee 9 4 4 FOCUS CORRECTION TO APPLY TO THE DOUBLET nennen enne nne ersten 11 4 5 RESULT OF THE FOCUSING venrei on EEEE E bep pug eb ee itp to epe E EE bett euet dee 11 4 5 1 eege eebe dag Zeg gg ge Ae bt e eet bech e geste El ee iltlenitaitel 11 4 5 2 E Ge 12 4 5 3 Energy in the target pixel sense 12 4 5 4 L E 13 4 6 PROBLEMS tette ee e ee ede 13 4 6 1 Clamping of the camera mount 13 4 6 2 Sigmof the ADU COUNTS i i ec ta d tete ed o e 13 4 6 3 Filters order in the filter wheel ss 13 4 6 4 Connector SUFJ CE a n e o rr POR ee EEN 14 4 6 5 Camera mount stability sine 14 5 OPTICAL AND BASIC DETECTOR TESTS sssssesssnsnenneenneneonsenceneeseencesceneeneeeneeeeee 15 5 1 FIELD OF VIEWER HR 15 5 1 1 DeSCHDIOn se ee dod uve ty ir ege e Ans nan denis 15 5 1 2 eebe p atateiitaieiiodtetioded anii Maat ren Cep fo et I LH 15 5 1 3 Diffuse illuminatiOTi 51h iae ele ete Edel en eet Deet gief 15 5 1 4 Focused saturated Spot x5 ie ode iba det eebe ede ede d tee ie desierit 15 5 1 5 To be checked Usable detector area 16 52 BACKGROUND LEVEL srl eneee EAEI eb pego bee ARAA EEEE ETE 18 5 3 DETECTOR READOUT NOISE IN FULL FRAME MODE nee 20 5 3 1 Aproximate gain facto sive atur opes ee i boite nee eios baleitela n ndietiia 21 5 3 2 Readout noise in full quadrant mode 512x512 pi
34. The two LdV optical tables LEONARDO and VINCI ALIU are separated by about 10 to 15 meters LEONARDO is the first optical table after the beam compressors The laboratory layout is not yet frozen and the precise positions of LEONARDO and VINCI ALIU are still TBD 2 2 2 Optical System The most recent version of the optical design of LdV is presented p 5 The beam diameter which will be accepted by LdV is 18 mm corresponding to the diameter produced by the beam compressors LdV is divided in four functional units COMBINER COMU it groups the fiber injection optics INA INB the MONA fiber combiner box the fiber output optics OUT and the filter wheel of LISA FILT ALIGNMENT TOOLKIT ALIU the Technical CCD head and associated optics the beamsplitter cubes ALI1 and ALI5 the ALI slide bearing ALIS and ALIA ARTIFICIAL STAR ARTU the LEONARDO artificial star optics and light sources lasers thermal light INFRARED CAMERA LISA the main HAWAII based infrared camera of LdV including its controler 2 2 3 LISA Camera The infrared detector of LdV is a HAWAII 1024x1024 array enclosed in a liquid nitrogen cryostat with a filter wheel a cold stop and a lens Only a quadrant 512x512 pixels will be used by LaV The four outputs of the optical fibers from the combiner are imaged on four windows of one or a few HAWAII detector pixels each of the detector two for the interferometric outputs and two for the phot
35. User Requirements 46 2 15 TCCD PROCEDURES 2 15 1 TCCD focusing The focusing of the TCCD is done in a standard way The image of a point source is obtained then fitted with a gaussian bidimensional curve and a correction is computed from the width of this gaussian Is is iterated a few times if necessary In Pupil Check mode the focus will be achieved on the red LED which is at the center of mirror M2 or on the artificial star image In Image Check mode it will be done on the stellar image either artificial or from the sky The focusing might be needed from time to time as a maintenance procedure In the following sub sections the TCCD is supposed focused 2 15 2 TCCD Calibrations It is necessary to obtain an image of the fiber inputs on the TCCD This is done by injecting light in the MONA box from one the outputs and illuminating the fiber input head Via the ALI1 and ALI5 beamsplitter cubes the resulting artificial stars are imaged on the TCCD Once the image has been acquired the precise position of the fiber input is computed by a gaussian fit These coordinates are the reference on which all the star light has to be concentrated during the observations 2 15 3 Pupil Check This section refers to the optical configuration described in the reference document 2 section 3 9 2 At the beginning of the night and to avoid any loss of stellar light it is important to check that the light beam arrives unobstructed
36. VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 13 These on axis parabolae are used to inject the light in the input fibers of MONA They are movable in tip tilt and focus translation These motions are used before every fringe acquisition to maximize the flux injected in the optical fibers by placing the star precisely on the fiber head During this operation the flux is monitored on the LISA camera pixels Element Range Values INA1 0 25000 microns Focus INA1 0 12500 microns Tip INA1 0 12500 microns Tilt INB1 0 25000 microns Focus INB1 0 12500 microns Tip INB1 0 12500 microns Tilt 2 4 7 OUT OUT is the fiber bundle output after the beam combination in MONA The four fibers are grouped in a bundle and they are imaged on the LISA HAWAII camera through a fixed lens The position of the bundle head has to be adjusted very precisely in lateral position focus and rotation so as to image the four fiber heads each on a single pixel of the detector Only the focus and rotation around the optical axis are remote controlled to be adjustable during the observations The other adjustments will be done during the day The goal is to put a maximum fraction of the light from the fiber output on a single pixel of the HAWAII detector As this may not be possible due to the small size of the LISA camera pixels is
37. a quick look analysis is performed by LdV DCS at the LCU level Each scan is rearranged in four 1D arrays synchronous time sequences that record the intensity evolution at each of the fiber outputs I1 I2 P1 P2 An algorithm see section 2 9 6 is performed on the arrays I1 and 12 to determine whether fringes have been observed Req 17 and if yes what is the time tzoppops Of the observed location of the center of the fringe packet corresponding to zero total OPD This time is compared to the expected time tzoppexp Of zero OPD occurrence and the quantity OPDorrset tzoppexp tzoppobs X V is sent to the VLTI delay line OPD controler as an OPD offset For the information of the observer it would be very interesting to compute a simple estimation of the instrumental visibility as a complement to the OPD offset and to send it to the LdV WS software The software should display one computed visibility to the user at a frequency of 2 Hz 1 Hz Req 18 Under normal observing conditions the OPD offset should be smaller than 100 microns Note that the quick look analysis requires that also reside in memory A dark scan or a small series of scans acquired off source The quick look data reduction parameters These parameters are updated at the end of each individual observation In the current implementation of VINCI it has only one spectral channel one window for each signal 11 12 P1 P2 With a typical number of frames per sca
38. due to the absence of a real cold stop for the detector see section 5 2 for details Figure 5 Image of the OUT fiber as seen on the detector of LISA The fiber core bright saturated point near center should be dark if not saturated is surrounded by the envelope dark and the surface of the connector brighter halo The scale being inverted the connector appears in reality darker than the background 4 6 5 Camera mount stability After the focusing of the beam on the detector it was possible to do the first stability tests of the mount First the long term stability of the mount if not charge is applied seems satisfactory with no visible motion of the spot over one night A concern was the difference in the position of the camera when it is loaded with nitrogen and when it is empty For this purpose a 3 kg mass was put on top of the dewar to simulate the presence of a load of nitrogen When the load is positioned precisely on the dewar axis of symetry then there is no apparent motion of the spot on the detector no variation in the intensity of the illuminated pixel This is due to the fact that a translation of the dewar in the plane of focus will not cause any shift of the spot When the load is added slightly offset from the center of the dewar upper plate then the load is asymetric and in this case there is a motion of the spot on the detector of one to two pixels This case will normally not happen under normal observati
39. efficiency or FSU residuals calibrator angular size Photometric transmissivity of the optical train Detector read out noise Hawaii array RON Strehl ratio fluctuations photometric noise Knowledge of the target spectrum shape Stability of the optical train vibrations Thermal background noise seen by the detector Internal turbulence non AO corrected Statistical photon shot noise photon noise Stability of the contrast loss polarization Sampling losses non blocking sampler Atmospheric transmission in the K band Wavefront distortion due to the optical train Possible additional detector noise array defects Differential longitudinal dispersion Fiber injection losses coupling efficiency lt 0 8 Transversal dispersion between visible and infrared Photometric efficiency of the fibered triple coupler Non stationarity of the atmospheric contrast loss Interferometric efficiency of the fibered triple coupler Correlation of the two channel visibilities Bad imaging of the fibers on the detector pixels ar EE VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 6 of 7 3 OPTICS 3 1 AREA OF THE COLLECTORS OPTICS The surfaces of the VLTI light collectors is given in Table 2 The siderostats have a limited aperture equivalent to 355mm on the sky that gives an area of 0 099m It is also reduced by proje
40. error ICS Status of BSA and BSB slides BSA1 BSA2 OUT Continuous Yes ICS Status of the artificial light source on off Continuous Yes VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 57 2 18 5 List of Parameters displayed by the Stellar Interferometer Mode GUI Origin Description Typical Possibility Update of GUI Frequency Setting VLTI Telescopes pointing coordinates RA Dec Alt Az Continuous Yes VLTI Telescopes status track slew Continuous No VLTI Wind load warning symbol Continuous No VLTI _ Seeing value arcsec 2 Hz 1 Hz No VLTI _ Seismic activity warning symbol Continuous No VLTI Time signals Local UT Sidereal Continuous No OS Target name target type scientific object calibrator Observation Yes OS Template situation target list Observation Yes OS Target coordinates RA Dec Observation Yes OS Target V and K magnitudes Observation Yes OS Target spectral type Observation Yes OS Target expected diameter mas Observation Yes ICS Status of LdV hw warning symbol green ok red error Continuous No OS Data acquisition OPD range Observation Yes OS Data acquisition fringe velocity Observation Yes OS Data acquisition computed sample interval Observation No OS Data acquisition number of points per fringe Observatio
41. es pour la programmation du logiciel de contr le de VINCI Il s agit du document de r f rence pour comprendre le fonctionnement pratique de l instrument ainsi que ses possibilit s d volution Le second document LISA Test Report Section 3 porte sur les r sultats des tests effectu s sur la cam ra infrarouge de VINCI lors de son int gration Garching Il pr sente les caract ristiques techniques de la cam ra son principe de fonctionnement ainsi que ses performances Le dernier document concerne la pr cision de l instrument VINCI D terminer la pr cision de mesure d un instrument interf rom trique est un exercice particuli rement d licat Je pr sente dans la Section 4 une estimation de la contribution des diff rentes sources de bruit sur les mesures VINCL ainsi que la pr cision r sultante sur la visibilit Ce document ayant t r dig avant les premi res observations de VINCI le lecteur est invit consulter le Volume I pour les r sultats obtenus en conditions r elles 2 LdV Software User Requirements EUROPEAN SOUTHERN OBSERVATORY Organisation Europ enne pour des Recherches Astronomiques dans l H misph re Austral Europ ische Organisation f r astronomische Forschung in der s dlichen Hemisph re VERY LARGE TELESCOPE LEONARDO da VINCI Software User Requirements Doc No VLT SPE ESO 15810 1852 Issue 1 11 Date 16 September 1999 Prepared P Kervella Name Date Signature App
42. for all the pixels in the lower left 64x64 area of the detector Figure 18 shows a graphical plot of the gains over the array linear scale Typical values are between 5 and 10 VLT TRE ESO 15810 2330 1 0 5 October 2000 29 LISA Test Report Figure 18 Gain map over the 64x64 pixels lower left area Pixel 0 0 is in the lower left corner X axis is to directed the right Y to the top 7 2 4 Histogram of pixel gains Pixel Gains Histogram 120 106 30 60 Number of photosites 20 Figure 19 Histogram of the 64x64 lower left pixels gains measured at 6 66 Hz frame frequency 7 3 PIXEL READOUT NOISES 7 3 1 Measurement procedure The readout noise RON is easily computed from the standard deviation of the noise sigma N in ADU at zero light level cold shutter closed and the gain G in electrons ADU through the formula RON sigma N x G sigma N is directly the time sequences standard deviation obtained with the cold shutter closed and G is already known for each pixel from the previous measurement LISA Test Report o no 5 October 2000 30 Figure 20 Readout noise map of the lower left 64x64 area at 6 66 Hz frame frequency The image grey scale is linear between 0 and 30 electrons 7 3 3 Histogram of pixel readout noises VLT TRE ESO 15810 2330 1 0 5 October 2000 31 LISA Test Report Readout Noise Histogram Number of photosites Readout nolse e Figure 21 Histog
43. for pixel 31 38 x fit gain 0 20000 y fit gain fltarr 2 y fit gain x fit gain gain 31 38 2 variances 0 31 38 oplot x fit gain y fit gain write bmp gain plot X31 Y38 bmp tvrd Power Spectral Density window 0 title PSD for pixel 33 38 xsize 800 ysize 600 plot psd 0 33 38 xrange 0 128 yrange 0 50000 xtitle Relative Frequency Maximum pix 128 3 33 Hz 0 026 Hz bin ytitle Power Spectral Density title Power Spectral Density for pixel 33 38 write bmp psd X33 Y38 bmp tvrd Power Spectral Density VLT TRE ESO 15810 2330 1 0 5 October 2000 47 LISA Test Report window 0 title PSD for pixel 32 38 xsize 800 ysize 600 plot psd 0 32 38 xrange 0 128 yrange 0 50000 xtitle Relative Frequency Maximum pix 128 3 33 Hz 0 026 Hz bin ytitle Power Spectral Density title Power Spectral Density for pixel 32 38 write bmp psd X32 Y38 bmp tvrd Power Spectral Density window 0 title PSD for pixel 31 38 xsize 800 ysize 600 plot psd 0 31 38 xrange 0 128 yrange 0 50000 xtitle Relative Frequency Maximum pix 128 3 33 Hz 0 026 Hz bin ytitle Power Spectral Density title Power Spectral Density for pixel 31 38 write bmp psd X31 Y38 bmp tvrd Histogram of the noises of the pixels window 0 title Histogram of the pixel readout noises xsize 800 ysize 600 plot histogram readout noise min 0 max
44. for the output alignment procedure is Req 42 1 Obtain quick exposures from the LISA camera in full frame mode 2 Monitor nearly continuously frequency of a few Hz values for the total flux arriving on the HAWAII detector the parameters of the four images of the fibers outputs position of the maximum light FWHM This is obtained by fitting gaussian curves to the image the percentage of the flux in ADU concentrated in each maximum light pixels the global percentage of the light concentrated in the four maximum light pixels The maximization of this value is the goal of the output alignment procedure 3 Adjust OUT1 through paddle control on the WS z rotation to move the four fiber outputs on the HAWAII detector This adjustement is done by the operator During this motion the positions of the fiber outputs are adjusted so as to put them on the desired columns of the detector The precise adjustment of the positions is done using the computed light concentration percentage The image of the fiber outputs is continuously displayed to the operator The look up table for this image is not dynamic but can be VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 51 adjusted by the user Only focus z direction and rotation can be adjusted remotely while the x and y settings are done by the observer before the observing night and are supposed fixed Once this is done the
45. g mirrors bad positioning or failure whenever power is on The automated operation of LdV via a template sequence is a very demanding goal to achieve Interferometers are very complex technical systems and it is important to realize that no currently working interferometer is completely automated The option to operate the instrument in a manual or semi automated way the first three bullets of the above list should not be discarded 2 18 1 Online Modes Interface The LdV WS sw in Stellar Interferometer Mode should be able to Req 47 e Interact with the VLTI Control System controlling the telescopes and gathering positional and environmental data The pointing instructions are sent to the VLTI Control System e Interact with the LdV Instrument Control System This allows the automatic or semi automatic setting and monitoring of the light switches motor commands encoders readout piezo fast scan parameters e Control the LISA DCS All the acquisition parameters duration frequency are defined by the observer and sent to the camera through the operating system e Control the Technical CCD DCS During the acquisition procedure the WS sw controls the settings of the exposure as well as displays the resulting images and performs the relevant parameters extraction as described in section 2 15 These parameters and data should be displayed in real time to the operator for quality check e Monitor real time information abou
46. in the online modes as no real time interaction between the LdV and VLTI subsystems LISA and DL for example or between LdV LCUs piezo and LISA are activated The only parameter which has to be accessible in the engineering mode and which does not exist in the other modes is the fast scan control signal proportional to the piezo extension coming back from the fast scan piezo INA3 controler Req 31 This signal is produced by the piezo itself and provided as an analog voltage output amplitude TBC by the piezo controler It is thus necessary to foresee an analog digital converter input in the LdV electronics design to access this information This will be used to check the performances of the piezo mirror regarding mechanical hysteresis There is no real time access constraint on this requirement the resulting signal can be displayed after the test cycle is completed These tests will be done at the ICS level and will require a dedicated GUI functionnality VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 34 2 10 DATA FLOW FROM LDV The data quantity coming out of LaV is not very large On a very successful observing night one can expect to observe during 8 hours This means 60 targets observed 10 minutes each plus 2 minutes to switch from one target to the other and acquire fringes Over these 10 minutes of data acquisition about 5000 scans are obtained 4 kbytes each This gives a
47. is bad then the final precision on the target will be substancially better than for the calibrator 8 1 2 Multiple calibration A way to solve this problem is to use several calibrators for the same science star Assuming that the angular diameters are not biased we can reduce the final uncertainty on the transfer function substantially If we consider the standard 2 3 0 035 mas calibrator case observed with the 200 meters baseline it is necessary to reduce the uncertainty from 1 45 to 0 04 to be limited only by the statistical uncertainty of VINCI bright star This means a factor 36 and more than 1300 calibrators to observe Of course this is impossible to achieve By using 10 calibrators which is a reasonable maximum number for a single star we can go to a transfer function precision of 0 4396 on the transfer function for the 200 meters baseline This is still a factor 200 above the VINCI maximum precision but this is probably the best we will be able to achieve on a single observation by using the raw Cohen calibrators catalogue 8 1 3 Refinement of the calibrator catalogue During the course of one night and assuming that we observe a number of targets and calibrators together it will be possible to cross calibrate the calibrators The observations with the longest baselines are able to resolve almost completely the 2 mas calibrators of Cohen This means that we will have the possibility to measure all the angular diamete
48. is difficult to fill the upper and lower tanks to more than 5 and 2 liters respectively 2 3 VACUUM TIGHTNESS The tightness of the dewar is relatively good but condensation was observed occasionally on the metallic surface of the dewar The addition of a vaccuum gauge by ESO to the dewar has made it possible to evaluate precisely the tighness of the camera When the camera is not cooled down the molecular sieve releases the gas it has trapped and the vacuum becomes worse After pumping and cooling down the level of vaccuum is very satisfactory pressure less than 10 6 mbar and stays at low values until the camera is warmed up again VLT TRE ESO 15810 2330 1 0 5 October 2000 6 LISA Test Report 3 MECHANICAL ALIGNMENT 3 1 DESCRIPTION The LISA dewar is attached to its dedicated mechanical support that is installed on the main VINCI table The height of the LISA support alone is not correct for the required height of the beams and therefore it is installed on four aluminium cylinders The whole system is not attached to the VINCI table and is therefore not compliant to the seismic resistance specifications of ESO Clamps have to be installed for the acceptance of the system The support is made of machined aluminum plates The rigidity of this system was tested with respect to the required accuracy on the detector maximum displacement of the fiber image on the detector has to be negligible compared to the pixel size See
49. limiting precision will not be reached for both types of collectors on the brightest stars the photon noise will never be dominating In the present section all the light collector cases including the siderostats with and without beam compressors are examined 9 2 PRECISION CURVES AND EXPOSURE TIME CALCULATOR TOOL An Excel sheet the graphical user interface is presented Figure 5 was created in order to estimate the impact of the variations of the numerous observational parameters on the precision obtainable with VINCI It takes as input parameters e the baseline length all the statistical errors tables computed with Matlab as listed in Section 7 the magnitude of the calibrator the angular diameter uniform disk of the calibrator the uncertainty on the a priori angular diameter of the calibrator and specifically for the exposure time calculator e science target magnitude e science target estimated angular diameter e requested precision A short macro is used to adjust the observation duration to match the requested precision It always considers the optimal scan as the actual acquisition rate The precisions curves are plotted in real time as the user enters the parameters values T VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 24 of 25 Calibrated Visibility Precision with VINCI VLT 4e Zerre A canin o Alf the fallosing computstions are done without Fringe tracker Pass
50. max level 2 max level min level j fix i rise for j fix itrise 2 fix itperiod 2 rise 2 do model j min level for j fix itperiod 2 rise 2 fix itperiod 2 rise 2 do model j min level max level min level j fix itperiod 2 rise 2 rise for j fix itperiod 2 rise 2 fix itperiod rise 2 do model j max level for j fix itperiod rise 2 fix itperiod do model j max level max level min level j fix i period rise 2 rise endfor plot model end VLT TRE ESO 15810 2330 1 0 5 October 2000 50 LISA Test Report 11 8 TRANSFER FUNCTION PLOT pro memory plot transfer function frequency model transfer SYNTAX memory plot transfer function frequency model transfer Plots the processing results for the study of the memory effect version 21 09 2000 for i 1 511 do model transfer i sin pi frequency i 2544 5 pi frequency i 2544 5 set plot X set plot PS device inches xsize 7 0 ysize 7 0 p position 0 15 0 1 0 9 0 9 window 0 title Temporal MTF compared to perfect integrator xsize 800 ysize 600 device file transfer compare ps plot frequency transfer function 0 512 yrange 0 5 1 2 xrange 0 1300 xtitle Frequency Hz ytitle LISA MTF sampled by spikes and perfect integrator line S title Temporal Modulation Transfer Function oplot frequency model transfer write bmp transfer compare bmp tvrd wdelete 0 device clo
51. operator mode Autotest Output Adjust operator mode Autotest Fringe Search mode Autotest Data Acquisition archival of the data 2 17 3 Pupil check Standard Template VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 52 mode Pupil Check Artificial Star A mode Pupil Check Artificial Star B mode Pupil Check telescope A mode Pupil Check telescope B TCCD images are made available to LdV and the external instruments 2 17 4 Image Check Standard Template e mode Image Check telescope A e mode Image Check telescope B e TCCD images are made available to LdV and the external instruments 2 17 5 Stellar Interferometer Mode Observations Standard Template The Pupil Check and Image Check steps require a call to the corresponding standard templates pointing of the telescopes initial positioning of the Delay Lines standard template Pupil Check standard template Image Check mode Stellar Interferometer Injection Adjust operator mode Stellar Interferometer Fringe Search mode Stellar Interferometer Data Acquisition archival of the data 2 18 GRAPHICAL USER INTERFACE The user of LdV shall normally interact with the instrument via the Instrument Workstation and the LdV graphical user interface GUI Req 58 The user will first have to select the way to operate among three options Standard Observing for astronomical targets observations in the stellar interferometer setup In this
52. option the operator will use the GUI to control the observations LdV will also be used as a measurement tool for internal seeing optical alignments internal OPD measurements This will make use of the standard operating modes defined Section 2 8 Maintenance to check periodically the performances of LdV at the OS level The GUI should display the series of tests to be performed and either perform them automatically or under the user s control The results of the maintenance tests are then logged for statistical analysis and preventive maintenance Engineering to carry out detailed tests or settings regarding the instrument hardware devices during improvements of the system or in depth analysis of hardware problems This is done at the ICS level VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 53 In the first two options the possibility shall be given to change the instrument state between on line stand by and off Req 45 Via the GUI it shall then be possible to select the instrument mode either Req 46 Individually via typed commands from the WS sw Via the GUI by form filling Via the GUI using pre stored files Via a template sequence The GUI should also display the status information provided by the LdV and TCCD control and acquisition electronics and other relevant data e g meteo and telescope data The GUI will also generate warnings of hazards or error conditions e
53. raw data files sorted by source The aproximate frequency frame scan observation at which this data has to be acquired is indicated All these parameters are available from the LdV LCUs or WS unless otherwise indicated VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 35 In the future dispersed fringes upgrade the observational data will be vectors instead of single values 1 LISA 11 12 P1 P2 data points once per frame or data vectors intensity wavelength in the future spectrally dispersed mode LISA calibrations once per observation LISA full frame once per observation LISA Offset of the fringe packet center quicklook meters once per scan 2 TCCD TCCD last images telescope 1 telescope 2 possibly with the artificial light source on taken before fringe acquisition once per observation 3 Encoders sensors Opto mechanical computer controled elements positions readings once per observation meters angular degrees see list in the previous sections 4 Hardware parameters once per observation unless otherwise specified Reference UT time at the beginning and end of each scan Reference sidereal time at the beginning and end of each scan LISA number of frames acquired per scan LISA number of frames saved per scan LISA pixels coordinates associated with 11 I2 P1 P2 set of coordinates for windows LISA percentage of the fiber bu
54. should be possible to define a window of several pixels which might not be adjacent instead of a single pixel Req 50 The maximum number of pixels for each of the four windows should be 25 the minimum number 1 See the adjustment procedure section 2 16 2 Element Range Values OUT 0 12500 microns Focus OUT1 0 20000 10 3 degrees Rotation 2 4 8 Polarization Controllers VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 14 In order to compensate for differential polarization states between the two beams the MONA box includes rotative polarization controlers one for each beam They are rotated by two motors whose range has to be defined about 1 turn Element Range Values POLAA 0 360 TBD degrees POLA B 0 360 TBD degrees 2 4 9 LISA Filter Wheel The LISA filter wheel carries four different filters for operation at different wavelengths K band K band H band 2 3 microns narrow band a free position open and an obstructed position closed The filters and the associated software names may change during the life of the instrument The filter names are fixed arbitrarily for the moment but should eventually comply with the ESO filter naming system Element Range Values FILT K KPRIME H NARROW OPEN CLOSED 2 4 10 Piezo Mirror INA3 This is a flat mirror mounted on a piezo stac
55. the camera Figure 28 shows the geometrical configuration of the chopper used to modulate the signal sent to the camera This System is installed at the injection side of the artificial light source This allows to have a sharper cutting of the beam than in the collimated part after the off axis parabola VLT TRE ESO 15810 2330 1 0 5 October 2000 39 LISA Test Report Chopper wheel Fiber acceptance cone Figure 28 Chopper geometrical configuration The chopping frequency as measured on the detector is 22 7 pixels at the highest frequency of 2544 53 Hz This corresponds to a chopping frequency of 2544 53 22 7 112 09 Hz There are 5 holes on the chopping wheel which means that the wheel is rotating physically at f 22 418 Hz The linear speed of the wheel at the radius of 100 mm corresponding to the position of the fiber head is therefore f x 2 x pi x 0 1 14 086 m s The numerical aperture of the fiber is 2 8 meaning that the fiber accepted beam has a diameter of 13 2 8 4 6 mm at a distance of 13 mm The time necessary to shut completely the beam is therefore 0 0046 14 086 0 327 ms At the 2544 53 Hz frequency the single frame time is 0 393 ms This means that the beam is cut by the shutter in 0 832 frame We can therefore consider that the shutoff of the light is instantaneous less than 1 pixel and adopt a square input signal model 9 4 DATA REDUCTION The goal of the data reduction here if to produce a gr
56. the detector 32 data points The waterfall display on the IRTD was correct To ascertain that the correct pixels were listed at the correct positions the fiber spot was moved from one pixel to the other in the same window fourth window The pixel showing the spot was always correct in the Beampix data submode The order in which the data is output is the following using the naming convention of the Figure 14 with A2 2 meaning the second read of the A2 pixel LISA Test Report o eas 5 October 2000 24 A1 1 A1 2 A2 1 A2 2 A3 1 A3 2 A4 1 A4 2 B5 1 B5 2 B6 1 B6 2 On the IRTD these values are displayed as a temporal sequence starting from the bottom of the screen to the top in a kind of reversed waterfall display To obtain the fluxes in ADU positive sign for example on pixel A1 Flux Al A1 1 A1 2 The actual fluxes from the DCS are computed using this formula and therefore have the correct sign 5 4 5 2 Flux data In this data submode the fluxes in each of the four windows first read minus second read to give a positive value is summed to give one value per window The result is a series of four numbers plus two additional numbers containing the total flux in all windows and TBC The display in the RTD shows the following values as displayed on the screen Line Name Order 6 2 O t1 O t2 O t3 5 Total flux F t1 F t2 F t3 4 Flux beam D D t1 D t2 D t3
57. the transfer function estimation 20 8 1 2 M ltiple calibration vy secs bats costa ege degen batiment anale Ten cg ste elei 21 8 1 3 Refinement of the calibrator catalogue 21 8 2 KNOWLEDGE OF THE TARGET SPECTRUM SHARE 21 9 CALIBRATED VISIBILITIES siscssiscssssscsssccssoessesssosssesescosccssssescansessscoesssncaseegssesseseseceosssseaavasscsessocesessesasens 23 9 1 TIGHT e SIE Ee E 23 9 2 PRECISION CURVES AND EXPOSURE TIME CALCULATOR TOOL suisses 23 9 3 CALIBRATORS BASELINE AND SHAPE FACTOR su isesieseesneesneneesneescnsesneeesneeseneeseneesnsesneesneeses 24 9 4 EXAMPLE THE CASE OF ZETA GEMINORUM s sessesseessessreseesstessesseesereseessreseessteseesseessreseesseeseesseesereseeeo 24 9 5 CALIBRATED VISIBILITY PRECISION CURVES FOR 40 M AND 195 M BASELINES eee 25 9 6 ATMOSPHERIC TURBULENCE AND FRINGE SENSOR IMPORTANCE cernere 27 10 APPENDIX SIGNAL AND NOISE FORMULAE i ccsssscssssssseccssssscecssssscccessscsccsesscccessessscescssese 28 10 1 IUE nU NBUIT E 92 28 10 2 Sen gp PEL M n Dre sa 28 10 3 THERMAL NOISE RS Er 29 10 4 PISTONNOISB 22 2 d eee Me sas D a rente enee eegen 29 10 5 BINDER e CEET 20 10 6 PHOTON NOISE SR e ee EE eds au ent es 30 10 7 UNCERTAINTY ON THE VISIBILITY iere rat ENEE qui EEN 30 11 APPENDIX MATLAB ROUTINES ccccssssssssssssscssssccsssscscesssscsccsessnsccsssscccsessececesseccessesescesesenses 31 11 1 EXECUTION BATC
58. total science data quantity over one night of 60x5000x4 kbyte 1 2 Gbyte The technical data instrument parameters and calibration data should a few hundred megabytes at most This gives a maximum data rate of about 1 5 Gbyte per night Most probably the real life data rate will be half of this figure that is 800 Mbyte Req 32 It is important to keep in mind that an upgrade of VINCI to dispersed fringes mode will increase the data flow from VINCI up to 50 times the previous figure with a dispersion on 50 pixels taken as a basis The effective data rate could then theoretically be 60 Gb per night usually 30 Gb night See section 2 20 3 for further details 2 11 LDV DATA STRUCTURE The data from LdV will be in two forms the WS localized data and the archived data 2 11 1 Workstation Localized Data The structure of the data on the Ld V dedicated WS depends mainly on the design of the software itself and thus will be defined later more precisely 2 11 2 Archived Data The general requirement reference for the archived data is Req 59 2 11 2 1 Data Hierarchy The interferometric data can be divided in frame scan batch observation and observation block from the most basic information element to the star and calibrator homogeneous data set which provides the final visibility measurement These terms are defined page 2 2 11 22 Data Sources This section gives a list of the data which has to be stored in the final
59. with BC if telescope 1 max_loop 85 31 end AT if telescope 2 max_loop 110 31 end UT without AO if telescope 3 max_loop 110 31 end UT with AO if telescope 4 max_loop 150 31 end for i 1 max_loop magnitude i 31 10 Scan speed fmins snr vinci 5e 4 foptions magnitude telescope dp If the scan speed is larger than the maximum speed of the piezo then take the maximum value if scan speed gt 0 01 ar T VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 32 of 33 Scan speed max speed freq max speed n lambda end Maximum SNR snr max snr vinci scan speed magnitude telescope Maximum precision in percents per scan and per channel precision max percents 100 snr max Maximum precision in percents for 100 scans and two channels precision max percents 2 channels 100 20 snr max 9 Scan duration in seconds Scan duration 1 scan speed 2 Noises contributions fraction piston fraction detector fraction photon noises cont scan speed magnitude telesco pe Storage of the results table_results i 1 magnitude table_results i 2 snr_max table results i 3 precision max percents table results i 4 scan speed n lambda table results i 5 scan speed table results i 6 precision max percents 2 channels table results i 7 fraction piston table results i 8 fraction detector table results i 9 fraction photon end For the plot
60. 000 1 DIT fits 0 cube 14000 full mrdfits level 14000 1 DIT fits 0 cube 20000 full mrdfits level 20000 1 DIT fits 0 Extracts the lower right corner of the data and keeps only 256 points to compute the Fourier transform faster cube intarr 7 64 64 256 sprint Extracting the useful data 64x64x256 Dry gt Cube 0 cube 0 0 63 0 63 0 255 cube 0 full 64 127 0 63 0 255 prints Cube 1 cube 1 0 63 0 63 0 255 cube 700 full 64 127 0 63 0 255 SEET gt Cube 2 cube 2 0 63 0 63 0 255 cube 1800 full 64 127 0 63 0 255 sprint gt Cube 3 cube 3 0 63 0 63 0 255 cube 4400 full 64 127 0 63 0 255 ett E gt Cube di cube 4 0 63 0 63 0 255 cube 9000 ful1 64 127 0 63 0 255 print 2 Cube Di cube 5 0 63 0 63 0 255 cube 14000 full 64 127 0 63 0 255 Spranty Cube 6 cube 6 0 63 0 63 0 255 cube 20000 full 64 127 0 63 0 255 Reads the scan records The beampix data gives the two values for the two reads of the double correlated readout mode The order of the data in the scan variable is the following first axis readouts for one frame Second axis frame sequence third axis scan sequence scan mrdfits scan mini fiber off 1 BeamPix fits 0 Computes the fluxes for each pixel n pix n elements scan 0 0 2 n frames n elements scan 0 0 n scans n elements s
61. 1 0 5 October 2000 16 LISA Test Report Intensity along the column 36 X 548 2000 4000 6000 8000 10000 12000 intensity ADU 14000 16000 18000 20000 Pixel along the Y axis Intensity along the line 29 Y 29 4000 6000 8000 10000 12000 intensity ADU 14000 16000 18000 Pixel along the X axis The structure of the array is nearly symetric relatively to the diagonal of the quadrant with the first lines and columns not as reactive as the others The explanation for this behavior may be the same as for the presence of dead pixels around the outer edges of the detector indium layer decay The best area for the positioning of the four scientific spots seems therefore to be approximately 20 X 45 25 Y lt 50 intersected with a quarter disc of radius 45 centered on the lower left corner 5 1 5 To be checked Usable detector area In the following figures the usable part of the Hawaii detector is indicated on the quadrant OK square area The size of this area is small but this is the best location with the current optical design This zone has to be verified using a non saturated spot moved across the detector VLT TRE ESO 15810 2330 1 0 5 October 2000 17 LISA Test Report 2 2 20 E a gi D z E CG cC o x pB Osa 9225323E8 22 e5222233 sea5 22223333 TERS Faw ase slseoeKx sztsesaaZgse 8
62. 10 11 Detector tests using the 64x64 pixels and windowed readout modes 5 October 2000 Sections 6 9 The results obtained with the chopping device memory effect are included and the filter wheel order is corrected Readout noise level error corrected VLT TRE ESO 15810 2330 LISA Test Report 10 5 October 2000 ii TABLE OF CONTENTS 1 INTRODUCTION PCR 3 1 1 SCOPE snp RR TO BE n PERSE BI et ee auae 3 1 2 GENERAL PHILOSOPHY veisins norost terne er e a are rp her rer tena 3 1 3 APPLICABLE DOCUMENTS 15 31 net RP e re Eg cet e REPRE ertt t DP ert ri hr eris 3 1 4 REFERENCE DOCUMENTS stees prO tete ie er rede ette tute Ore M erg t e her reper epee 3 1 5 ABBREVIATIONS AND ACRONYMS eene nenne tremit ene tremens tee nenne nenne trennen teen nnenenes 3 2 DEWAR GENERAL CHARACTERISTICS eeeee eee e e eene etna senos ense asta estes ense tas e eae eaa seen sese tassa 5 2 1 COEDTEMPERATURE AUTONOMY ceteras Pepe eG Gr GET Toren a Wim oT ee 5 2 2 NITROGEN FILLING e H 5 2 3 VACUUM TIGHINESS eee ere re REE ee de a ee ne ra aa nie eue 5 3 MECHANICALAATIGNMENT sisssiesesccsscadesadeaevacessietieade eoa saaa Er AE i Et Eae iatea ressos tissai 6 3 1 DESCRIPTION EE RO ORE E EE 6 3 2 ALIGNMENT PROCEDURE Age REN Hae oat iad i ded 6 4 DOUBLET FOCUSING iter eere ene ee Cena aua ee Ux eek eeu E tesa aeS aN e eeatt seo eroatea dee pue pa 8 4 1 ALIGNMENT PROCEDURE teet ae uH RU We Se RUE REC Re PL
63. 2 3 1 Frame 2 11 2 3 2 Scan 2 11 2 3 3 Observation 2 11 2 4 Data Format 2 12 Description of the Observation Procedure 2 12 1 Observation Procedure with VINCI 2 12 2 LEONARDO Interface with the VLTI Instruments 2 12 3 Alignment Toolkit Interface with the VLTI instruments 2 13 Instrument User Manual 2 14 Settings Database 2 15 TCCD Procedures 2 15 1 TCCD focusing 2 15 2 TCCD Calibrations 2 15 3 Pupil Check 2 15 4 Image Check 2 15 5 Star Image Centering 2 15 6 Off line use of TCCD images and LISA full frames 2 16 Injection and Output Optimization Procedures 2 16 1 Refined Injection Optimization 2 16 2 Output Alignment Procedure 2 17 Templates 41 41 44 45 45 2 17 1 2 17 2 2 17 3 2 17 4 2 17 5 2 18 2 18 1 2 18 2 2 18 3 2 18 4 2 18 5 2 18 6 2 18 7 2 19 2 20 2 20 1 2 20 1 1 2 20 1 2 2 20 2 2 20 3 2 20 4 2 20 5 VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements iv Autotest Mode Observations Standard Template Autocollimation Mode Observations Standard Template Pupil check Standard Template Image Check Standard Template Stellar Interferometer Mode Observations Standard Template Graphical User Interface Online Modes Interface List of Parameters displayed by the Autotest Autocollimation Modes GUI List of Parameters displayed by the Pupil Image Check Modes GUI List of Parameters displayed by the Artificial Star Mode GUI List of Parameters displayed
64. 2 4 5 8 Autocollimation OPD offset 2 8 2 9 Pupil and Image Check accessible from other instruments 2 8 4 2 8 5 10 LEONARDO stand alone operation 2 8 6 11 LEONARDO accessible from other instruments 2 8 6 12 LEONARDO high level command 2 8 6 13 Piezo frequency 2 9 2 14 Piezo range 2 9 2 15 Piezo wave shape 2 9 2 16 Piezo LISA synchro precision 2 9 2 17 Quick look algorithm 2 9 6 18 Visibilities display 2 9 6 19 Possibility of dispersed mode 2 10 20 Duty cycle minimum value 2 9 2 21 Data acquisition dead time 2 9 2 22 OPD offset frequency 2 9 2 23 DL Offset time 2 9 2 24 LISA readout rate selected by the user 2 9 7 25 LISA during signal check 2 9 8 26 LISA percentage of light in a single pixel 2 9 8 27 LISA full frame image 2 9 9 28 LISA output alignment checking 2 9 9 29 LISA full frame readout rate 2 9 9 30 LISA full frame storage 2 9 9 31 Fast Scan control signal in Engineering mode 2 9 10 32 LdV data rate 2 10 33 Online data reduction on the LdV WS 2 11 1 34 Observations buffering 2 11 1 35 Control procedure of LEONARDO 2 12 2 36 Control procedure of the Alignment Toolkit 2 12 3 37 Expected pupil position reading 2 15 3 38 Pupil position from the TCCD focus 2 15 3 39 Star image centering procedure 2 15 5 40 TCCD image storage with the LdV data 2 15 6 41 Manual injection optimization 2 16 42 Output alignment procedure 2 16 2 43 Output parameters storage 2 16 2 VLT SPE ESO 15810 1852
65. 3 Flux beam C DELL Le OCS as 2 Flux beam B B t1 B t2 B t3 1 Flux beam A A t1 A t2 A t3 5 4 5 3 Quicklook data A vector containing the processed scans TBC is also produced In the tests we have conducted the quicklook algorithm was not implemented and therefore the data produced was not meaningful VLT TRE ESO 15810 2330 1 0 5 October 2000 25 LISA Test Report 6 READOUT FREQUENCIES 6 1 INTRODUCTION The readout of the camera in double correlated mode is based on five steps Reset of the four windows Optional wait time First readout of the four windows Exposure time Second readout of the four windows Gre The execution of these five steps produces two numbers for each pixel corresponding to the first and second reads Both numbers are accessible in the Beampix data submode for each pixel of each window In the Flux output mode the fluxes are summed over each window The order of the pixels in the data output from the camera is explained in section 5 4 5 The minimum exposure time Min DIT is computed automatically and displayed on the control panel This minimum time corresponds to the time needed to complete a single scan with the specified number of frames It does not include the overhead time necessary to compute the delay line offset and therefore it is strictly the time necessary to acquire one frame multiplied by the number of frames From this number it is easy to compute the effective f
66. 30 bin 1 nan title Readout Noise Histogram xtitle Readout noise e ytitle Number of photosites write bmp Noise histogram bmp tvrd Histogram of the gains of the pixels window 0 title Histogram of the pixel gains xsize 800 ysize 600 plot histogram gain min 0 max 5 bin 0 1 nan title Pixel Gains Histogram xtitle 10 x Gain e ADU ytitle Number of photosites write bmp Gain histogram bmp tvrd Gain factors image window 0 xsize 64 ysize 64 tv gain write bmp gain map bmp tvrd Readout noises image window 0 xsize 64 ysize 64 tv readout noise stretch 0 600 write bmp readout noise map bmp tvrd Median PSD for all pixels window 0 title Median PSD over the 64x64 pixels xsize 800 ysize 600 plot median psd xrange 0 128 yrange 0 30000 xtitle Relative Frequency Maximum pix 128 3 33 Hz 0 026 Hz bin ytitle Power Spectral Density title Median Power Spectral Density over the 64x64 array dark write bmp median psd bmp tvrd end 11 5 PLOT SCAN DATA pro scan data plot frequency mean psd SYNTAX Scan data plot frequency mean psd scans Plots the processing results for the selection of the best pixels for the LISA camera VLT TRE ESO 15810 2330 1 0 5 October 2000 48 LISA Test Report This ends the first step of the best pixels selection as described in the document LISA tests in Garching
67. 6 Siderostat without BC 15 4296 Auxiliary Telescope 24 35 Unit Telescope 24 35 3 2 2 Vibrations and polarization contrast losses Due mainly to residual longitudinal vibrations the contrast of the fringes measured on a point source in the laboratory will not be 100 The VLTI optical train has been specified for a maximum loss in visibility of 1 percent and on site tests on the Unit Telescopes Koehler amp Leveque 1999 have shown that a maximum contrast loss of 5 can be expected due to vibrations in the telescopes only In addition a polarization mismatch is introduced after each reflection A maximum overall loss of 0 4 in K is expected and is neglected in the computations ar Sens VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 7 of 8 The expected interferometric efficiency of the VLTI is therefore 0 95 3 2 3 Static wavefront distortion due to the optical train The static planeity of the wavefront produced by the optical train of the VLTI will impact on the quantity of flux injected into the fibers The quality of the wavefront will be evaluated using the end to end model developed at ESO Detailed beam propagation simulations are currently produced by Rainer Wilhelm 3 2 4 Pupil lateral jitter The pupil lateral positioning effect on the injection in a fiber is given in RD4 For the VLTI the stability of the pupil lateral position depends on the bases used The worst case is for the station loc
68. 6 0 7 5 Auxiliary Telescopes 8 3 9 8 Unit Telescopes without Adaptive Optics 8 3 9 8 Unit Telescopes with Adaptive Optics 12 3 13 8 ar EE VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 20 of 21 8 ASTROPHYSICS 8 1 UNCERTAINTY ON THE CALIBRATOR ANGULAR SIZE 8 1 1 Precision of the transfer function estimation The precision of our a priori estimations of the angular sizes of the calibrators will directly impact the final precision of the science targets visibilities In this section the level of precision expectable from the calibrators is evaluated As a basis we will consider the catalogue of infrared calibrators from Cohen et al AJ 117 1864 1889 1999 Other angular size sources are also accessible but the Cohen catalogue has the advantage of providing a uniform coverage over the whole sky It was created for the calibration of the space observatories in the infrared wavelengths The level of uncertainty of this catalogue is typically o 1 5 of the limb darkened angular diameter of the selected calibrators The conversion of the limb darkened LD values to uniform disk UD angular diameters is relatively easy as the Cohen stars are selected in well known K5 giants region of the HR diagram Therefore it is possible to compute a high precision conversion factor between the LD and UD diameters and to keep approximately the same error bar on the UD value The angular diameters range c
69. 9 Water mm H20 Figure 4 Precipitable water vapor above Paranal A typical value of 90 for the overall transmission assuming 2mm precipitable water vapor in the K band is assumed in the following Table 9 Atmospheric transmissions inside the K band Wavelength range 2 2 1um 2 1 2 2um 2 2 2 3um 2 3 2 4um Transmission 0 72 1 0 96 0 93 2mm water vapor ar Sens VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 15 of 16 5 6 TRANSVERSAL ATMOSPHERIC DISPERSION Between different wavelengths the vector pointing to the star from the telescopes is not exactly the same when the star is not at zenith This results in a loss on the total flux injected into the optical fibers as described in details in RD7 Two kinds of effects will happen on LdV 5 6 1 Between the K band and the visible The guiding of the telescopes is achieved in the visible therefore the lateral shift of the star image due to transversal dispersion The difference between the apparent position of the star in the K band and in the visible will cause some flux loss as the star will move slightly on the fiber head during the exposures typically a few minutes long for LaV At 60 degrees of zenith angle maximum value the differential tilt between the R band visible and the K band can reach 1 41 arcsec In this particular case a correction has to be applied every 10 seconds in order to limit the shift to a maxi
70. CI 1 1 16 September 1999 Software User Requirements 63 A number of optical elements are moved manually If they are forgotten in a wrong position this could cause data to be lost The monitoring of their position with the possibility to send alerts to the user in case of mispositioning might be required in order to ensure the reliable operation of the instrument The particular question of the fiber type sensors is adressed here with respect to the reference document 7 They are not essential for the operation of the instrument but they could make the operation of LdV easier and more straightforward In the following table the name of each fiber connector to be monitored is indicated in the column Measured Physical Parameter for example the ART3 connector on the artificial source The names of the fibers that can be plugged on each connector are listed in the column value for example it is possible to plug WG1X1 WG1X2 WG2X1 WG2F1 WG2F2 on the ART3 connector Each fiber is designated by its coded name for example WG1X2 means WaveGuide 1 connected to the X coupler number 2 on one end Measured Physical Measured Range Values Comments Parameter every ART3 FIBER Observation WG1X1 WG1X2 The main light source fiber WG2X1 WG2F1 feed WG2F2 WG2F3 WG2F4 WG2F5 WG2F6 WG2F7 WG3F1 WG3F2 WG3F3 WG3F4 WG3F5 WG3F6 WG3F7 WG4F1 WG4F2 ART2 FIBER Observation WG1X2
71. EOTSCANIDATA semer e eaae ve b ensue ies EDU EX YR Ede Ite E a e p a eee age epe pina 47 11 6 TRANSFER FUNCTION COMPUTATION eeeeeeeene nennen a trennt a teen nennen 48 11 7 MODEL GENERATION tsp Sege 49 11 8 TRANSFER FUNCTION PLO P re eetretepetebeze pete petere AEE EE tte md te ee Tae ee ns 50 VLT TRE ESO 15810 2330 1 0 5 October 2000 1 LISA Test Report Figures Figure 1 LISA Nitrogen consumption 5 Figure 2 LISA installed on its support 6 Figure 3 Example of a focusing done by moving the fiber test 7 X scale in tens of microns Y in ADUs 10 Figure 4 Direction of the applied focus motion 11 Figure 5 Image of the OUT fiber as seen on the detector of LISA The fiber core bright saturated point near center should be dark if not saturated is surrounded by the envelope dark and the surface of the connector brighter halo The scale being inverted the connector appears in reality darker than the background 14 Figure 6 The HAWAII array LISA detector uses quadrant IV picture from Rockwell The X axis of the LISA camera is along the fast shift register positive direction and the Y axis is along the slow shift register positive direction 17 Figure 7 LISA Hawaii detector usable part 17 Figure 8 Zoom on the usable part of the array 18 Figure 9 Variation of the background due to stray light from a soldering iron about 25 degrees off axis The curves correspond to three different test sequences 19
72. ERING AND MAINTENANCE MODES The need for these modes is only to have direct access to all the hardware parameters of the instrument The normal state for LdV during Engineering and Maintenance operation is Online In the Engineering mode the user can access and monitor individually the hardware devices motors TCCD through the part of the GUI related directly to the ICS and DCS Req 61 It is then possible to adjust all the hardware parameters individually encoders ranges motor speeds voltages 2 8 LDV INSTRUMENT MODES Here are described the different modes and the associated hardware setups while LdV is in Online State Setups are assumed to be particular settings of the movable devices of LdV associated with an instrumental mode The LdV setups allow the user to move the many LdV optical and mechanical elements at the same time as a whole by entering the corresponding request Generally speaking there is no critical time constraint on the motion of the LdV hardware devices except for the piezo mirror see section 2 9 4 No particular hardware incompatibilities are foreseen on the VINCI table The software system should then aim at moving the devices as quickly as possible Req 52 goal 5 seconds minimum 15 seconds in parallel in order to spare observing time But there is no critical time limit driven by the conception of the instrument The following tables give the setups for each LdV mode The setups are nam
73. Figure 24 PSD for pixel 33 38 from 0 to 3 33 Hz X 128 Y scale arbitrary 33 Figure 25 Median Power Spectral Density for the 64x64 pixels from 0 to 3 33 Hz X 128 The Y scale is arbitrary 34 Figure 26 Overview of the PSD for pixel 33 38 read at 1656 Hz frame rate frequencies between 0 and 828 Hz The shutter was open background light present but the fiber spot was not illuminated The large peak on the left is precisely in the bin containing the 100 Hz frequency mean over 20 scans 35 Figure 27 Power spectral density at maximum acquisition frequency for pixel 33 38 with cold shutter closed Notice the overall very low level compared to the previous case 1656 Hz frequency Ths frequencies of the peaks are mostly multiples of 50 Hz therefore suspected to be associated with power supply pickups grounding problems 36 Figure 28 Chopper geometrical configuration 39 Figure 29 Modulation transfer function of the LISA camera spikes compared to a perfect integrator solid line 40 Tables Table 1 Fiber head motion to obtain the focus on the Hawaii array 9 Table 2 Result of the in focus test on LISA 13 Table 3 Filters positions in the filter wheel 13 Table 4 LISA readout modes 22 Table 5 Maximum readout frequencies for LISA 25 VLT TRE ESO 15810 2330 1 0 5 October 2000 2 LISA Test Report Table 6 Modulation transfer function of LISA Table 7 Apparent problems in the LISA system during the tests in
74. Garching Table 8 Points to be clarified VLT TRE ESO 15810 2330 1 0 5 October 2000 3 LISA Test Report 1 INTRODUCTION 1 1 SCOPE This document describes both the procedures and results of the tests conducted on the LISA camera in Garching using the collimated light source provided by the Observatoire de Paris Meudon 1 2 GENERAL PHILOSOPHY VINCI makes a particular use of the HAWAII detector of LISA reading only a few of its pixels at a high frequency The tests described here aim at validating both the optical concept and performances of the camera These tests are preliminary to the acceptance of LISA in Garching and also to the AIV of the whole LEONARDO da VINCI instrument 1 3 APPLICABLE DOCUMENTS 1 LISA Maintenance Manual Draft MPE 06 2000 2 LISA User s Manual Draft MPE 06 2000 3 LISA Acceptance Plan Draft MPE 06 2000 4 LISA Drawings MPE 06 2000 5 LISA Statement of Work VLT SOW ESO 15810 1xxx v 1 0 28 07 1999 6 LISA Technical Specifications VLT SPE ESO 15810 1xxx v 1 0 31 08 1999 7T OUT Alignment Procedure DESPA 17 07 2000 8 ICD between VLTI and Instruments VLT ICD ESO 15000 1826 v 1 0 16 11 1999 1 4 REFERENCE DOCUMENTS 1 5 ABBREVIATIONS AND ACRONYMS ADU Analog Digital Unit AIV Assembly Integration and Verification ALIU The Alignment Unit DCS Detector Control Software DP Data Pipeline GUI GEI GOL HW ICS IRTD IWS LCU LdV LEONARDO LISA LISA WS MONA NA
75. Garching The mention NA not applicable in the Solved column means that the problem was not solved but is not a concern in the actual status of the camera Table 7 Apparent problems in the LISA system during the tests in Garching Problem Suspected Cause Solved 1 Negative values for increasing incoming Wrong sign in the subtraction of the two Yes flux in double correlated mode readouts for the double correlated mode 2 Filters in the filter wheel are not in the Wrong mounting of the filters Yes documented order 3 Blurred image of the collimated light Incorrect focusing of the doublet relatively to the Yes source on the detector HAWAII array 4 Impossible to mount the cryogenic Damage to the threading on the dewar NA security device for the upper nitrogen tank 5 Insecure fastening of the camera support No briddling currently available No on the table 6 Very high level and variable background Improper cold stop NA 7 Three dark quadrants displayed together Data is not discarded early enough in the No with the useful one computer system only for 64x64 and 512x512 8 Dark stripes in the widowed readout mode Program code problem Yes missing data 9 First few pixels of a frame are brighter Reset before the scan removed Yes than the others 10 Generation of the windows for the four Off line generator has bugs see Section 5 4 2 No beams mode is tedious
76. H creuse tete re ee ED ee tete eee ettet dete eee tee eee eet 31 11 2 MAIN PROGRAM ABAQOUB iet degen a ak ole ted eee eee tee dede est 31 11 3 SNR COMPUTATION SNR VINCD sise rennes tente antennes 32 11 4 NOISES CONTRIBUTIONS NOISES CONT enne nnt tnttnt entente nenne 34 ar Sens VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 4 of 5 1 INTRODUCTION 1 4 ScoPE This document gives estimates of the precision and sensitivity that will be achievable with VINCI in its stellar interferometer mode using the different VLTI light collectors It gives the limiting magnitudes in each case the statistical visibility precisions that will be achievable and finally the calibrated visibilities accuracy The impact of the use of a fringe tracker and adaptive optics is also described briefly 1 2 REFERENCE DOCUMENTS 1 Interface control document between VLTI and its Instruments VLT ICD ESO 15000 1826 v 1 0 16 11 1999 2 Functional description of the VLTI VLT ICD ESO 15000 1918 v 1 0 16 11 1999 3 Test Siderostat Optical design Pupil Shape and Sky Coverage VLT TRE ESO 15000 1616 v 1 0 08 06 1998 4 Filtrage modal et recombinaison de grands telescopes Contributions l instrument FLUOR C Ruilier 1999 5 LdV Optical Definition VLT MEM MEU 15810 1000 v 2 0 21 06 1999 6 LdV Sources and Guided Optics VLT SPE MEU 15810 1001 v 1 0 10 07 1999 7 Transversal atmospheric disp
77. NG OF THE FIBER HEADS ON THE DETECTOR The reimaging system is designed to fit all the light coming from the MONA outputs into four pixels The photometric efficiency of this system can be estimated at 3096 Table 7 lists the contributions to this number from the different optical elements values taken from RD11 and RD12 Table 7 Fibers output block transmission Off axis parabola 0 98 Flat mirror 0 98 Camera window CaF2 0 92 K band filter 0 80 Sapphire lens 0 85 ZnSe lens 0 65 Ensquared energy on the detector 0 77 Total transmission 0 30 ar EE VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 9 of 10 4 DETECTION 4 1 DETECTOR QUANTUM EFFICIENCY This curve is taken from the Rockwell web site for the 1024x1024 HAWAII detector The mean quantum efficiency in the K band is 6296 T 78K Area 3 423e 6 Quantum Efficiency 9 34R SWIR1024 Peak 1 970 pm Cutoff 2 579 um 0 8 1 2 2 4 1 6 2 0 Wavelength um 4 2 READOUT NOISE The value given by Rockwell for the Hawaii chip minimum readout noise is lt 10 e However the RON depends on the frequency and the readout electronics and the detector of LdV is read relatively quickly Therefore it seems realistic to take the 10 e value as a basis 4 3 THERMAL BACKGROUND NOISE 4 3 1 Temperature This noise contribution is normally minor compared to the other noise sources in the K band A photon shot noi
78. NSVERSAL ATMOSPHERIC DISPERSION cessscccessscecessecceessecccesseeccesseeecssseeccesseeecseseecsesseseeesseseeesaess 15 5 6 1 Between the K band and the vtstble inner 15 5 6 2 InsidethekK DANS ER 15 5 7 NON STATIONARITY OF THE ATMOSPHERIC CONTRAST Loss 15 5 8 CORRELATION OF THE TWO INTERFEROMETRIC CHANNELS esneesseeseneesrneesesesnnesns 15 5 9 STATISTICAL PHOTON SHOT NOISE cssscessscssecossscesccesesceseccescccsscceseccessccnseccsssessesonsescssecssscessesossesonsnees 16 6 GLOBALE PARAMETERS E 17 6 1 OVERALL PHOTOMETRIC EFFICIENCY ssscssecosssccsssceseccesesossecosscecssecessccesecosseccsssessesossesossecssseessesessesonse 17 6 2 OVERALL INTERFEROMETRIC EFFICIENCY sise esneesensesnns encens screens esnsesnnseeneesnesne 17 7 STATISTICAL VISIBILITY PRECISION CURVE G scccccccssssssssssssssscscscsssssccscsssssessssssecssesseseseees 18 7 1 SIGNAL TO NOISE RATIO sense sensesnsesnnsennesseneesns encens sense 18 7 2 OPTIMAL OPD SCAN SPEED cssccesssessseeesscesseceseecsssecssscessscessecessecesscesuecessecessecessecsscessecesaecessecesseeeateess 18 7 3 CURVES WITH THE DIFFERENT LIGHT COLLECTORS WITHOUT FSU 18 VLT TRE ESO 15810 2177 LdV Precision and Sensitivity o 10 12 July 2000 3 of 4 KS ASTROPHYSICS e x eM 20 8 1 UNCERTAINTY ON THE CALIBRATOR ANGULAR SIZE ccccsssscesssessseessscessecessecsssccsseccseeceseecessecssteesseeesaees 20 8 1 1 Precision of
79. PTION The following sections give the list of the commands related to the opto mechanical elements Req 56 the TCCD and the LISA camera of LdV For every Element the current status as well as the last setup value should be accessible to the user through the GUI in online modes or the GEI in engineering mode Req 1 In general any failure when performing setup actions should be reported to the user immediately through the GUI or GEI Req 2 2 4 1 BSA BSB These beam splitter cubes are movable in and out of the light beams in order to clear the stellar light paths of any obstruction during the interferometric observations The cubes have to be moved in translation to insert or remove them from the beams Three positions are available Beamsplitter cube sending the light directly to the instruments BSA1 BSB1 for autotest mode VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 11 Beamsplitter cube sending the light to the telescopes BSA2 BSB2 for autocollimation mode Without any optical element in the beam OUT for stellar interferometer mode The BSA and BSB slides shall be positioned always in the same direction Req 3 in order to avoid any backlash in the mechanical motion Element Range Values BSA BSA1 BSA2 OUT BSB BSB1 BSB2 OUT 2 4 2 ALI ALIS ALI and ALI5 are beam splitter cubes used to redirect the stellar beams to the te
80. Section 4 6 5 for a detailed report Figure 2 LISA installed on its support 3 2 ALIGNMENT PROCEDURE Check that the VINCI table is horizontal VLT TRE ESO 15810 2330 1 0 5 October 2000 7 LISA Test Report Set all the mechanical adjustments of the LISA camera to central position Check that the LISA support bottom plate is horizontal relatively to the VINCI table For this use a slide caliper to measure the height of the four corners of the support relatively to the VINCI optical table Check that the upper plate of the dewar is horizontal with a bubble level Pay special attention to the setting of the cradle part of the support the direction pointed by the dewar window These two settings directly impact the position of the beam on the detector Measure the height of the entrance window of LISA and of the output mirror of the artificial light source OUT relatively to the VINCI table surface This can be done directly by using a square and measuring the height from the table surface or by using a slide caliper to measure the heights of all the supports The center of the dewar window is located aproximately 149 mm above the bottom surface of the dewar Adjust if necessary of the height of LISA by using the three pads below the dewar Be careful to maintain the same height for all pads The height of the center of the entrance window of the dewar should be 320 mm Check precisely the horizontalit
81. VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 13 of 14 This value is of the same order of magnitude as the 1296 precision of the theoretical computations for the same scan duration i e with a fringe speed of 500um s presented in Section 5 1 1 which is lower because the seeing conditions considered are for Paranal hence with a better seeing than Mount Hopkins 5 1 3 Fringe Sensor Unit The role of the FSU is to reduce the piston sufficiently to have a maximum OPD error of 30 nm up to magH 13 UTs or 10 ATs as specified in RD9 When used with the FSU the theoretical minimum precision with VINCI considering this residual piston only is 0 3796 up to the limiting magnitude of the FSU H 4 with the siderostats For the brightest stars K 2 on the siderostats the theoretical OPD residual is about 1nm taken from RD10 which translates into a theoretical visibility loss of 0 000496 4e 6 Such a number should not be considered reliable as no experiment has ever demonstrated this kind of capabilities Moreover this is notthe final precision as it only takes into account the piston noise The PFSU can theoretically provide the same accuracy but for stars about 2 magnitudes brighter i e up to H 2 The parameters for the upgraded version FINITO should be close to the FSU values 5 2 STREHL RATIO The mean values of the Strehl ratios apart from the quasi static wavefront errors introduced by the optical tr
82. able code taken from the GCVS for example Spectral type s determination s with publications references Period seconds and days epoch julian date wavelength of measurement meters reference publication for periodic variables Expected amplitude of the angular diameter variation if applicable in milliarcsec and radians Comments entered by the observer 6 Software parameters once per observation unless otherwise specified 9 Reference code of the observation can be sequential allows to quickly retrieve the data LdV mode name or code LdV setup name or code Projected baseline length meters once per scan Baseline vector coordinates u v w once per scan OPD velocity meters second once per scan Hour angle seconds once per scan Theoretical RA Dec Theoretical Altitude Azimuth once per scan Wavelength range mini mean maxi coupled to the LISA filter wheel position meters This will be replaced by a vector of wavelength ranges one min mean max triplet per spectral channel for spectrally dispersed data Observed internal OPD meters once per scan 9 VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 37 LdV OPD offset meters Expected visibility on the current baseline mean over the observation with uncertainty dimensionless Technical CCD extracted parameters position brightness of the ob
83. ain readout noise fourier psd median psd Plots the processing results for the selection of the best pixels for the LISA camera This ends the first step of the best pixels selection as described in the document LISA tests in Garching version 12 09 2000 Variance of the pixel as a function of flux gain plot for one good pixel window 0 title Gain plot for pixel 33 38 xsize 800 ysize 600 plot means 33 38 variances 33 38 psym 2 xtitle Mean ADU ytitle Variance title Gain plot for pixel 33 38 x fit gain 0 20000 y fit gain fltarr 2 y fit gain x fit gain gain 33 38 2 variances 0 33 38 oplot x fit gain y fit gain write bmp gain plot X33 Y38 bmp tvrd Variance of the pixel as a function of flux gain plot for one good pixel window 0 title Gain plot for pixel 32 38 xsize 800 ysize 600 plot means 32 38 variances 32 38 psym 2 xtitle Mean ADU ytitle Variance title Gain plot for pixel 32 38 x fit gain 0 20000 y fit gain fltarr 2 y fit gain x fit gain gain 32 38 2 variances 0 32 38 oplot x fit gain y fit gain write bmp gain plot X32 Y38 bmp tvrd Variance of the pixel as a function of flux gain plot for one good pixel window 0 title Gain plot for pixel 31 38 xsize 800 ysize 600 plot means 31 38 variances 31 38 psym 2 xtitle Mean ADU ytitle Variance title Gain plot
84. ain see Section 3 2 3 expected for the different light collectors are listed in Table 8 The values without AO assume an r of 1 meter in the K band at Paranal and do not take into account the residual tip tilt errors see Section 3 5 except fo the UTs with AO Table 8 Strehl ratios for different telescope configurations Telescope type Strehl without AO in K Strehl with AO Siderostat wihout BC 100 Siderostat with BC 100 AT 30 896 UT 1 56 48 K 12 Shaklan Applied Optics 27 2334 1988 has shown that the optimal injection is obtained for a value of Dir of 4 with D the telescope diameter and r0 the Fried parameter It could therefore be interesting to foresee pupil stops for the first observations with the UTs not equipped with AO systems 5 3 PHOTOMETRIC NOISE The adaptive optics and or tip tilt correction systems will leave some residual star motion on the fiber head This will cause a degradation in the quantity of light coupled into the fibers and random variations of the flux measured on the detector From time to time the flux injected in the fibers can drop to zero or small values As the interferometric signals are divided by the photometric signals this causes a decrease in the SNR of the photometrically corrected fringes In the case of VINCI this effect is equivalent to a random photometric loss The single mode fibers filter out completely the tip tilt and high order wavefro
85. am combiner wihout going through any transmissive optical element to reduce the absorption VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 20 Unit Element Stell Interf Stell Interf Stell Interf Stell Interf Stell Interf Idle Injection Fringe Data Polarization Adjust Search Acquisition Adjust OMU INB SLIDE MOTOR OFF OFF OFF OFF OFF pecie dee eli ds ot COMU INB1 MOTORS OFF com BE D I o pn pt H COMU INA1 MOTORS OFF S ss BEE len ek COMU INA3 FAST SCAN ed NI WT Ce TIP TILT FOCUS ROTATION COMU POLA A POLA B P L t ran refer pr por fas ja I E EE ALIU JALI SLIDE N A N A N A N A a omon e e ee ues ALIU TCCD N A N A P ng A MA O O a a i ial ei ua source 2 8 4 Pupil Check The light sources one Light Emitting Diode for each telescope at the centers of the two mirrors M2 are on This mode and all the related data should be accessible from outside of LdV software Req 9 The LdV instrument is exclusively assigned to the instrument requesting the pupil check Only the ALIU and ARTU LEONARDO units are concerned but no other operation can be conducted simultaneously The images and other data position of the center of the pupil obtained in this mode by an external user instrument should be made available to this instrument VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requiremen
86. amera Unit 2 4 Movable Hardware Description 2 4 1 BSA BSB 2 4 2 ALII ALIS 2 4 3 ALI Slide 2 4 4 TCCD Assembly head and lens 2 4 5 INB Slide 2 4 6 INAI INBI 2 4 7 OUTI 2 4 8 Polarization Controllers 2 4 9 LISA Filter Wheel 2 4 10 Piezo Mirror INA3 2 5 Summary of LdV Movable Hardware Positions 2 6 Instrument States 2 7 LdV Engineering and Maintenance Modes c Un oo O WM VM WD 00 GO CO G ee VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements iii 2 8 LdV Instrument Modes 2 8 1 Autotest 2 8 2 Autocollimation 2 8 3 Stellar Interferometer 2 8 4 Pupil Check 2 8 5 Image Check 2 8 6 Artificial Star 2 9 Data Acquisition aspects of LdV 2 9 1 Description 2 9 2 Terminology and typical values 2 9 3 Chronology of data acquisition 2 9 4 Real time considerations 2 9 5 Delay Line Control 2 9 6 Quick Look Fringe Detection Algorithm 2 9 6 1 Construction of the Combined Interferometric Signal 2 9 6 2 Frequency Filtering 2 9 6 3 Fringes Detection and OPD Offset 2 9 6 4 Alternative Algorithm for Fringe Detection and OPD Offset 2 9 7 Synchronized Data Acquisition Parameters SYNC 2 9 8 Signal Check Parameters NOT SYNC 2 9 9 LISA Full Frame Readout FULL FRAME 2 9 10 Engineering Mode Data 2 10 Data flow from LdV 2 11 LdV Data Structure 2 11 1 Workstation Localized Data 2 11 2 Archived Data 2 11 2 1 Data Hierarchy 2 11 2 2 Data Sources 2 11 2 3 Data Time Scales 2 11
87. ample interval it include provision for the acceleration and slow down of the piezo which depends on the wave shape the optimal wave shape is determined based on the OPD range and the frequency through a decision table TBD The integration time for LISA is computed by Sample Interval microns Fringe Velocity in microns s The command to start the synchronized acquisition of data can be sent by the user or the WS sw at any time by switching to the mode Stellar Interferometer Data Acquisition 2 9 8 Signal Check Parameters NOT SYNC In order to precisely center the star image on the fiber head by moving the injection parabolae INA1 and INB1 it is required to have a continuous estimation of the flux on the four used pixels of LISA Req 25 This process is called the injection optimization During this phase the injection mirrors INA1 and INB1 are moved in tip and tilt to position the star image intensity peak exactly on the tiny fiber head a few microns wide The flux observed on the LISA pixels gives the VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 32 information to know if we are getting close to this maximum if it rises we are getting closer if it falls we are going the wrong way Moreover the atmospheric turbulence causes the flux to vary erratically during this process Basically this optimization has to be done before each star observation that is every
88. ample of pixel 33 38 VLT TRE ESO 15810 2330 1 0 5 October 2000 36 LISA Test Report Power Spectral density at f 2544 53 Hz 0 026 0 015 0 010 Powar Spectral density for pixel 33 38 no light 0 005 6 000 1 1 1 1 i 1 1 1 1 il 1 1 1 1 o 506 1600 1500 Frequency Hz Figure 27 Power spectral density at maximum acquisition frequency for pixel 33 38 with cold shutter closed Notice the overall very low level compared to the previous case 1656 Hz frequency Ths frequencies of the peaks are mostly multiples of 50 Hz therefore suspected to be associated with power supply pickups grounding problems 21 7 5 DISCUSSION A strong peak is visible precisely at 100 Hz on the 1656 Hz data and several smaller peaks are visible on the 2544 5 Hz data The presence of this peak is difficult to explain This could be a pickup of the electric power distribution Otherwise the noise is very clean white showing no particular increase with the frequency VLT TRE ESO 15810 2330 1 0 5 October 2000 37 LISA Test Report 8 INDIUM LAYER STABILITY The array used in LISA is an engineering grade detector and suffers from dead pixels at the external edges of the array This is caused by the destruction of the indium layer connecting the photosites The evolution of this defect has to be followed carefully as it may evolve after each cooling of the camera The procedure is simply to save an full frame
89. aph of the modulation transfer function of the camera as a function of frequency at the maximum data acquisition rate It uses a series of IDL procedures that are listed in the appendix The resulting curve is presented on VLT TRE ESO 15810 2330 1 0 5 October 2000 40 LISA Test Report Temporal Modulation Transfer Function 0 8 0 6 LIS amp MTF sampled by spikes and perfect integrator line DA 1 1 1 l 1 1 1 L 1 1 1 o 506 1600 1500 Frequency Hz Figure 29 Modulation transfer function of the LISA camera spikes compared to a perfect integrator solid line The behavior of the camera is close to the perfect integrator curve Table 6 Modulation transfer function of LISA Frequency Hz Absolute MTF value Perfect Integrator Ratio 111 8 1 001 0 997 1 004 336 7 0 963 0 971 0 991 561 6 0 948 0 922 1 028 785 2 0 839 0 851 0 985 1008 9 0 722 0 761 0 949 1232 5 0 592 0 656 0 902 9 5 DISCUSSION The conslusion of this study is that the camera response is not affected by memory effect in a way that could disturb the detection of the interference fringes Especially the Modulation Transfer Function at high frequency is very close to the ideal case LISA Test Report VLT TRE ESO 15810 2330 1 0 5 October 2000 41 10 SUMMARY OF PROBLEMS The following table lists the major problems encountered during the installation and tests of LISA in stand alone mode in
90. ated the farthest from the laboratory e g the J6 station Table 4 lists the pupil lateral jitter as given in RD1 Table 4 VLTI Pupil lateral jitter worst case J6 station Time AT without beam Siderostats UT and AT window expanders with beam expander 0 1s 400 um RMS 90 um RMS 1s 600 um RMS 135 um RMS 10s 730 um RMS 165 um RMS 30 min 790 um RMS 180 um RMS The mean time to acquire a scan in VINCI on a faint star is about 0 2 seconds but the time between two optimizations will be of the order of 5 minutes The figure to take into account is here the latter as the flux loss will not be checked afterwards The resulting flux loss is computed using the following formula taken from RD4 DH g 22561 Pmax with p the lateral misalignment as a fraction of the pupil size In any case the pupil size is 18mm at the entrance of the instruments Table 5 RMS flux loss due to lateral jitterof the pupil Light collector Lateral relative Flux loss RMS jitter AT without beam 750 18000 3 9 expander 4 2 Siderostats UT AT 170 18000 0 2 with beam expander 0 94 3 3 FIBER INJECTION STATIC LOSSES The maximum fraction of the light injected into the optical fibers is 78 This is due to the fact that the mode on which the telescope diffraction pattern is projected has a gaussian shape and not an Airy pattern 3 4 FRESNEL LOSSES Fresnel losses happen at the surface of the fiber
91. by the Stellar Interferometer Mode GUI Interface for the Engineering and Maintenance Modes Setting up the Instrument Parameters List of Numbered Requirements Second Generation Upgrades Automated Injection Optimization Fast image scan algorithm Slow image scan algorithm Automated Output Alignment Spectral Dispersion Sensors Photometric Calibrations for the TCCD VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 1 1 INTRODUCTION The VLT Interferometer Near Infrared Commissioning Instrument LEONARDO da VINCI LdV is composed of three subsystems with separate functions two beams combiner VINCI alignment toolkit ALIU artificial star light source LEONARDO ALIU and LEONARDO are intended to be facilities of the VLTI infrastructure for the other instruments and for alignment VINCI will be used first to debug the VLTI and obtain the first fringes and then as a fiducial point for fringe recovery after changes in the VLTI or its instruments It will also be an important pedagogical tool particularly as it can obtain interference fringes autonomously in Autotest mode The design of VINCI is based on the FLUOR beam recombiner Fiber Linked Unit for Optical Recombination which is currently routinely operated at the Mount Hopkins Observatory Arizona The conception and design of LdV are provided by the Observatoire de Paris Meudon and it is built as an ESO instrument The
92. can 0 0 flux intarr n pix n frames n scans VLT TRE ESO 15810 2330 1 0 5 October 2000 43 LISA Test Report for pix 0 n_pix 1 do begin flux pix scan 2 pix scan 2 pix 1 endfor end 11 2 PROCESS 64x64 DATA pro process 64 cube means variances gain readout_noise fourier psd median_psd mean_psd_per pix median RON median gain SYNTAX process 64 cube means variances gain readout noise fourier psd median psd mean psd per pix median RON median gain Processes the test data acquired with the LISA camera in order to extract the best possible pixels for the VINCI beams This procedure makes the computations for the 64x64 windows data cubes and produces the power spectral density for all pixels at different lighting levels This is the first step of the best pixels selection as described in the document LISA tests in Garching It also computes the gain factor and the read out noise for every pixel at the frequency used for the tests version 28 09 2000 Computes the mean value and variance of each pixel over the sequenc It excludes the last acquisition partly saturated print Means and variances means fltarr 6 64 64 variances fltarr 6 64 64 for n 0 5 do begin print for i 0 63 do begin for j 0 63 do begin means n i j mean cube n i j variances n i j variance cube n i j endfor endfor endfor
93. chnical CCD assembly The cubes have to be moved in translation to insert or remove them from the beams Three positions are available Beamsplitter cube directing the light to the Technical CCD assembly ALI1 ALI5 for alignment purposes Shutter blocking the light from the telescopes ALI1S ALI5S In the current definition the shutter positions will be used in engineering mode only Without any optical element in the beam OUT for stellar interferometer mode The ALI1 and ALI5 slides shall be positioned always in the same direction Req 4 in order to avoid any backlash in the mechanical motion Element Range Values ALI1 ALI1 ALI1S OUT ALI5 ALI5 ALIBS OUT 2 4 3 ALI Slide The two mirrors ALI3 and ALI4 are on the same moving slide They are used to redirect either the A or the B beam to the technical CCD assembly after the ALI1 and ALI5 cubes An intermediate position allows to see the output of MONA after reflection on ALI9 mirror movable by hand The ALI slide has to be moved in translation to insert ALI3 ALIA or no mirror in the beam The ALI slide shall be positioned always in the same direction Req 5 in order to avoid any backlash in the mechanical motion VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 12 Element Range Values ALI Slide ALI3 ALI4 FREE The user should be informed when the slide is moving and wh
94. ction effects and the effective area can be as small as 0 050m The average effectie area for the 10 MIDI objects is 0 084m The pupil shape of the siderostats is never circular but elliptic with an axis length ratio of 4 3 at best to 2 1 at worst The longer axis is about 80mm in all cases The resulting star image shape on the fiber heads will therefore not be circular and the injection efficiency will be reduced In the unfavorable case of the operation of VINCI without beam compressors in the laboratory the efective entrance pupil diameter is reduced to 10 cm 2cm x magnifying factor This results in a very small collecting surface and should be avoided as much as possible Table 2 Areas of the collectors primary optics Telescope type Effective area m Siderostat 0 084 average Siderostat without BC 0 039 Auxiliary Telescope 2 545 Unit Telescope 50 265 3 2 TRANSMISSION OF THE OPTICAL TRAIN 3 2 1 Photometry The overall light transmission efficiency of the VLTI optical train is given in Table 3 for the different light collectors including all the mirrors up to and including the folding mirrors on the VINCI table The transmissivity up to M16 is taken from RD1 A flat reflectivity for the 6 excess siderostats mirrors of 98 in the K band has been assumed Table 3 Transmission of the collectors in the K band Telescope type mirrors Transmission Siderostat 18 409
95. d obtained to reduce the stastical noise During the batches off source and with only one of the beams the piezo mirror could be stopped as OPD modulation is not relevant but the data produced is the same as on source four series of numbers 11 12 P1 P2 An observation is a collection of four batches On source the target is centered in the field and fringes are observed A few thousand scans are obtained during this phase Off source the telescopes are offset from the source in order to measure the sky background signal The number of scans Oft source is approximately the same as On Source but they are acquired faster as no synchronization is need between the piezo mirror and the LISA camera Beam A the beam B is obstructed by a shutter or by offsetting the telescope B and a series of about a hundred scans are obtained Beam B the beam A is obstructed by a shutter or by offsetting the telescope A and a series of about a hundred scans are obtained The order in which these batches will be acquired is not necessarily the one given here An observation represents 5 10 mm of observing time and its end product is the fundamental VINCI data unit to be saved in the VLTI archive No time critical operation is expected to occur between two successive observations The fast scan mirror INA3 is mounted on a piezo device which is controled through a command voltage Uc The following table lists the different scales that
96. d Sensitivity 10 12 July 2000 25 of 26 Calibrated Visibility Precision with VINCI VLT au Zerre kervella 57 Alf the Bed campufatrans are ve without fringe tracker Presse HI ONLY the dark Ais ceffs with the relevant vetoes the athers are Sofas Calibrator and target K magnitude Size of the calibrator mas Percentage of error on the calibrator angular size 95 Uncertainty on the ht of this sheet for a graphical display Precision siderostats without BC 9 Precision curves Baseline length m bservation Block integration time s wavelength fixed at 2 2 microns See the chart on the ri Precision siderostats with BC 9 Precision ATs 9 Precision UTs with AO 9 Exposure time calculator Magnitude of the target K band Estimated Uniform Disc angular diameter mas Correlated magnitude of the science target Requested calibrated visibility precision 9 EE Compute Exposure Time 235 nan oo ares ll Pa Maximum OB duration s Hide the chart on the right to compute faster press Esc to stop Siderostats without BC Figure 6 Zeta Geminorum observations 9 5 CALIBRATED VISIBILITY PRECISION CURVES FOR 40 M AND 195 M BASELINES A typical 2 3 mas 0 035 mas K magnitude 2 calibrator star is assumed with an observation time of 180 seconds The vertical scale is different between the two plots VLT TRE ESO 15810 2177
97. d of light m s inimum integration wavenumber m 1 fringe power aximum integration wavenumber m 1 fringe power dp oe 9 2 Acquisition frequency of the camera Hz frequency n scan speed lambda ar Spi VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 35 of 36 if frequency lt le 6 frequency le 6 end Scan duration dt l n frequency lambda Frequency bandwidth df lambda frequency n sigma max integ sigma min integ Signal S 2 A delta lambda 1le6 T I F0 10 m 2 5 2 dt 4 oe The 1e6 factor is to account for the W m2 MICRON reference flux 3 n oe hermal nois quivalent power squared nu c lambda delta nu c lambda delta lambda 2 c lambdat delta lambda 2 B 2 h nu 3 c 2 1 exp h nu k Temp 1 etendue E lambda 2 epsilon n thermal 2 h c lambda delta nu B etendue 2 Piston nois quivalent power squared piston l exp 1 5 1 fringes n lambda frequency 5 3 n piston S piston Detector nois quivalent power squared n detector 2 frequency RON h c QE lambda 2 Photon nois quivalent power squared n photon 2 h c lambda 2 A delta lambda 1e6 T F0 10 m 2 5 The 1e6 factor is to account for the W m2 MICRON reference flux Returns the SNR total noise n thermal 2 dt df n piston 2 n detector 2 dt df n photon 2 dt df snr
98. double correlated The readout in full quadrant mode worked properly during the tests The intensity values follow a negative scale The minimum exposure time is a bit more than 2 seconds due to the intrinsic readout speed limitations of the detector Three dark quadrants are displayed by the Infrared Real Time Display IRTD in addition to the useful quadrant This is a bit confusing and should be corrected 5 4 4 64x64 window double correlated Tis readout mode works properly once the window dimensions are selected The options to activate on the control panel are the following to read the lower left corner 64x64 window 1 HW window selected green 2 StartX 1 NX 128 3 StartY 1 NY 128 The readout frequency is much higher in this mode than in full quadrant mode see section 6 The display on the Infrared Real Time Display IRTD is correct but presents the same three dark quadrants as in the full quadrant readout mode 5 4 5 Windowed readout double correlated VLT TRE ESO 15810 2330 1 0 5 October 2000 23 LISA Test Report 0 0 fast register Figure 14 LISA windowed mode readout order The order of the data in the IRTD and in the data file is presented in the following sections 5 4 5 1 Beampix data During the tests of the windowed readout mode four beams each of 2x2 pixels the system behaved correctly showing the intensity of the light in the 16 pixels at the first and second reads of
99. e software Computation of the position error once the target is identified its coordinates are compared to the fiber head input coordinates as measured during the calibration phase The alignment of the optical train up to the laboratory is supposed to be good and only the injection optics will be moved to precisely position the star image on the fiber input Computation of the tip tilt mirrors required motions the difference is computed and converted to tip tilt positions of the injection mirrors The coefficients for this conversion two for each injection mirror have been estimated during the commissionning of LdV It will be also necessary to include an offset coefficient in the parameters due to the different positions of the star images in the visible and in the infrared This offset varies with respect VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 49 to the altitude of the star in the sky At zenith this is zero while near the horizon it is several arcsec projected on the sky but a direct mathematical relation allows its computation based on the altitude of the object the TCCD sensitivity curve mean wavelength and the mean wavelength of the LISA infrared band 6 Motion of the INA1 and INB1 mirrors commands are sent to the INA1 and INB1 motors to move in order to position the image of the star in the infrared at the same position as the fiber head as measured at th
100. e 3 Example of a focusing done by moving the fiber test 7 X scale in tens of microns Y in ADUS The direction of the focus motion necessary is towards the off axis parabola See Figure 4 VLT TRE ESO 15810 2330 1 0 5 October 2000 11 LISA Test Report Off axis parabola AFocus gt Focus on Focus the Hawaii infinity array collimated Figure 4 Direction of the applied focus motion 4 4 FOCUS CORRECTION TO APPLY TO THE DOUBLET As described in Section 5 1 5 the usable area of the detector is centered on the pixel X 33 Y 38 The focusing tests 6 and 7 see Section 4 3 have been done on pixel X 36 Y 35 which is very close to the center of the usable area Therefore the focus motion to give to the doublet should be computed assuming a measured focus offset of 0 214 millimeters on the fiber position as test 7 gives the best percentage of light in a single pixel and is very coherent with the measurement obtained on test 6 The direction of the motion to give to the doublet has to be determined knowing that the fiber focus correction has been done in the adequate direction Figure 4 It is mandatory to take into account the correct magnification factor between the off axis parabola and the doublet The focus correction has to move the doublet closer to the detector The ammount of motion is computed using the following formula AX doublet G AX per with G 9x 18 5 125 1 332 the nominal longitudina
101. e beginning of the night It is necessary to inject the light at a position offset from the visible outputs of the fibers as seen on the TCCD due to differential refraction of the atmosphere The computed offset will have to take into account the differential refraction which requires input from VLTI environmental parameters humidity temperature It is foreseen that the VLTI collecting telescopes send the infrared light directly on axis while guiding on the optical image off axis This means that the differential diffraction offset has to be taken into account directly at the telescope level This would make this procedure simpler as no variable offset would have to be applied to the INA1 and INB1 mirrors 7 Refined optimization At this stage the star should be positioned on the fiber head sending light into the MONA box up to the HAWAII detector The refined injection optimization can then be conducted see next paragraph 2 15 6 Off line use of TCCD images and LISA full frames At least one image of the object from each telescope obtained with the TCCD during the acquisition procedure should be saved in the data Req 40 This will enable the user during post processing to check visually the quality of the seeing the postion of the object relatively to the fiber heads during pre acquisition As a complement to the TCCD image s the computed data on the positions of the sources in the image as well as the refined injection op
102. e data acquired should be available on request from any system to the VLTI CS but no real time constraints are foreseen The interface requirements are currently under definition TBD Generally speaking the alignment toolkit should be integrated in the VLTI infrastructure as much as possible and follow the same control procedure Req 36 The allocation of ALIU alignment toolkit to the alignment software or the other instruments should be done by request By default it should be assigned to LdV software 2 13 INSTRUMENT USER MANUAL As LdV will be operated in several ways as a facility for the other VLTI instruments it is important to foresee a complete User s Manual for this instrument This will enable the science instruments teams to have a reference for the proper use of LEONARDO the alignment unit those two systems will be used directly by the other instruments and VINCI 2 14 SETTINGS DATABASE The operation of LdV on different baselines will require a database of settings internal OPD INB slide position TCCD focus positions to quickly find the fringes after a baseline change This database should be available on line on Paranal It is not part of the VINCI software and belongs to the VLTI system As a general rule for LdV there is no particular order in which the various functions should be set and the settings can be done in parallel VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software
103. e testing period with the siderostats The effect of the FSU is important on the precision of the visibility measurements reduced or canceled piston noise but not primarily on the limiting magnitude of the instrument With a FSU the maximum measurement accuracy can be reached on all reachable stars but the observable targets number will not be significantly larger than without FSU As VINCI and the FSU will supposedly have the same limiting magnitude fainter stars will not be accessible before the arrival of PRIMA and its dual beam capability With PRIMA though the limiting magnitude will become much higher Mag K 19 The Fringe Sensor Unit is absolutely essential for the most demanding observation programs such as the exoplanets and the efforts should be concentrated on having it available on Paranal as soon as possible The limiting precision reachable with the FSU is not yet certain but could be as low as 0 001 10 5 for all observable stars os EE VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 28 of 29 10 APPENDIX SIGNAL AND NOISE FORMULAE 10 1 USEFUL VALUES f Tr fs fringes n l l n At fringes Jos We 1 a 1 ii nax integration To A nin integration AF Fringes o Onin 10 2 SIGNAL AA Wavelength range micron A Effective single telescope collecting area m T Overall photometric efficiency of the optical train Overall interferometric eff
104. e tiff ONLY the dark Ap ceffs with the relevant valves he others are automatic 80 Calibrator and target K es magnitude 180 Size of the calibrator We mas Percentage of error on the calibrator angular size Precision curves 5 MM MM N Baseline length m Calibrated visibility precision for one Calibrator Science Target Observation Block 96 Observation Block integration time s T1087 Total error Unit Telescopes with AO Siderostats without BC Siderostats with BC Auxiliary Telescopes Uncertainty on the calibrator size mas Wavelength fixed at 2 2 microns See the chart on the right of this sheet for a graphical display Precision siderostats without BC Precision siderostats with BC 95 Precision ATs 95 Precision UTs with AO 9 Exposure time calculator Magnitude of the target K band Estimated Uniform Disc angular 18 Compute Exposure Time diameter mas iis Correlated magnitude of the science 324 Minimum OB duration s 1 Requested calibrated visibility 15 Maximum OB duration s 600 precision x its target Hide the chart on the right to compute faster press lt Esc gt to stop Siderostats without BC Siderostats with BC Magnitude K Auxiliary Telescopes 3 2 1 D 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Unit Telescopes with
105. each observation Altitude Azimuth beginning and end of each observation Mean tracking error RA and Dec over the observation 10 Observatory once per observation unless otherwise specified parameters read from the observatory environment monitoring system Environmental parameters seeing isoplanetic angle correlation time air pressure air temperature ground temperature humidity wind velocity and direction seismic activity seismic flag 2 11 2 3 Data Time Scales VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 38 In this section the data produced by LdV is sorted by time scale The data which is not available locally in the LdV LCUs and WS is indicated 2 11 2 3 1 Frame I1 12 P1 P2 data points data vectors in the future dispersed mode OPD introduced by the Piezo Mirror meters can be computed after the acquisition through the piezo motion parameters 2 11 2 3 2 Scan 9 Reference UT time at the beginning and end of each scan Reference sidereal time at the beginning and end of each scan LISA Offset of the fringe packet center quicklook meters this value is the one sent the the OPD controller in order to recenter the fringe packet Projected baseline length meters Baseline vector coordinates u v w OPD velocity meters second Hour angle seconds Theoretical Altitude Azimuth Number of data poi
106. ed aberration free fiber image in the autocollimator The aberration free field of the off axis parabola is only of the size of the fiber core so the precision of positioning must be very good The image should be a small yellowish disk of light with round shape and clean edge If the fiber is moved vertically or laterally use the OUT2 mirror to center the image back on the center of the autocollimator field of view Note the exact focus position of the micrometric knob used to move the fiber in focus and also in other directions This measurement can be done to a precision of 1 micron on the adjustment knob Be careful to have a precise reading Remove the autocollimator from the beam There should be a somewhat fuzzy image on the LISA detector VLT TRE ESO 15810 2330 1 0 5 October 2000 9 LISA Test Report 4 2 FOCUSING PROCEDURE The measurement procedure is based on the idea that we know the magnification factor between the fiber off axis parabola and the doublet inside the LISA dewar The knowledge of the focus offset between infinity and correct detector focus enables to compute the motion to give to the doublet or to the detector to focus the dewar optics correctly to infinity Therefore the measurement is done several times using the following procedure Focus the fiber at the focal point of the off axis parabola to produce a clean collimated beam procedure explained in paragraph 4 1 Take down precise
107. ed by their function for example Autotest Injection Adjust designates the hardware setup used to make the adjustment of the light injection in the fiber heads in the autotest mode The Idle setup which is found in all the instrument modes is used for engineering purposes or during observations for special needs It is the default starting setup if no setup is specified by the user when switching to a mode setup including the engineering and maintenance modes Req 54 In the Idle setup the instrument in online and ready to switch to another setup in the same mode or in another mode IN in the optical beam OUT off the optical beam N A not applicable ADJ adjustable 2 8 1 Autotest In this mode LdV can observe fringes without any other VLTI system involved The LEONARDO artificial star is used to produce the light which is sent directly to the VINCI table VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 18 Autotest Autotest Autotest Autotest Autotest cr pe Search om MOTOR steps COMU INB1 MOTORS OFF mm POM is COMU INA3 FAST SCAN pe 3mm ANRT ud S SAN COMU OUT1 MOTORS reer rr m ROTATION COMU POLA A POLA B mm uM M n soil ect A ai vc TT Posmon ALIU TCCD N A N A N A N A N A fmeesmw P C M P mM source LISA LISA Acquisition GNE ANE NN PIX LN PIX 4 PIX FRAME NOT SYNC SYNC SYNC 2 8 2 Autocollimati
108. en it has reached the working positions ALIB ALIA FREE 2 4 4 TCCD Assembly head and lens In order to focus the TCCD on a point in the laboratory situated about 5 meters away from the TCCD where the pupil of the VLTI is it is necessary to insert a supplementary lens in front of the telescope Two preset focus positions are selectable Req 6 one for the infinity focus without TLENS FOCUS1 and the other for the focus in the laboratory with TLENS at the foreseen distance of the pupil image projected by the VCM PRESET2 The focusing algorithm for the TCCD is a standard one based on the analysis of the FWHM of a point source see section 2 15 1 Element Range Values TCCD IN OUT Lens TCCD 0 25000 TBD microns Focus PRESET1 PRESET2 2 4 5 INB Slide The mirrors INB1 INB2 and INB3 are grouped on a moving slide to allow for the balancing of the optical paths in each arm of the interferometer The slide position can take continuous values to allow precise compensation of the residual OPD This motion will also be used to look for the fringes in the autotest mode It will allow to scan for the fringes spread over a length of a few tens of microns over a few centimeters length The INB slide shall be positioned always in the same direction Req 7 in order to avoid any backlash in the mechanical motion Element Range Values INB 0 25000 microns 2 4 6 INA1 INB1
109. ere 2x2 pixels in size giving 16 pixels For each pixel a series of 20 scans of each 1024 frames was analysed The frame frequency was set to the maximum possible value at the selected configuration see Section 6 for details e g 1656 Hz Therefore the frequencies investigated by computing the PSD of the sequences were in the range 0 828 Hz 7 4 2 2 Example of pixel 33 38 VLT TRE ESO 15810 2330 1 0 5 October 2000 35 LISA Test Report Power Spectral Density for pixel 33 38 amp 20 G15 Power Spectral Density a a 0 05 p 6 06 1 1 1 1 1 1 Il 1 1 1 i 1 1 1 i 1 1 L o 200 400 660 800 1006 Frequency Hz Figure 26 Overview of the PSD for pixel 33 38 read at 1656 Hz frame rate frequencies between 0 and 828 Hz The shutter was open background light present but the fiber spot was not illuminated The large peak on the left is precisely in the bin containing the 100 Hz frequency mean over 20 scans 7 4 3 Maximum frequency 0 1223 Hz 7 4 3 1 Data analysed This measurement is done using the windowed readout mode of LISA The windows used were 1x1 pixel in size giving 4 pixels For each pixel a series of 20 scans of each 1024 frames was analysed The frame frequency was set to the maximum possible value at the selected configuration see Section 6 for details e g 2545 Hz Therefore the frequencies investigated by computing the PSD of the sequences were in the range 0 1223 Hz 7 4 3 2 Ex
110. ersion in the VLTI VLT TRE ESO 15000 1989 v 1 0 25 01 2000 8 Factors affecting the performance of stellar interferometers B Koehler 9 Specifications for the feasibility study of PRIMA VLT SPE ESO 15800 1652 v 1 0 28 08 1998 10 PRIMA FSU Final report F Cassaing et al ONERA 16 08 1999 11 LdV Optical Definition VLT SPE MEU 15810 1000 v 2 0 21 06 1999 12 LISA Final Design Review Hardware R Hofmann 10 11 1999 ar Sens VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 5 of 6 2 PRECISION AND SENSITIVITY LIMITING FACTORS VINCI is designed to be a test instrument Though because of its particular optical fiber beam combiner it can achieve very high precisions in the measurement of visibilities FLUOR has demonstrated remarkable capabilities using this combination principle Table 1 give the list of all Known sensibility and precision limiting factors that will affect VINCI Precision and sensitivity are the two sides of the same problem and are considered simultaneously Though not all of the listed factors will cause both a precision and sensitivity degradation The factors limiting mostly the precision without affecting the sensitivity are in italic Table 1 Precision and sensitivity limiting factors in VINCI Optics Detection Atmosphere Astrophysics Area of the light Detector overall Piston noise direct sky Uncertainty on the collectors optics quantum
111. f the behavior of the fringes when affected by piston it is necessary to produce simulated interferograms affected by a typical atmospheric piston and then to compute the visibility using the standard pipeline The dispersion on the final visibilities as a function of the standard deviation of the OPD would give more realistic piston noise effects estimates 1 0000 0 9000 0 8000 0 7000 0 6000 0 5000 0 4000 Instrumental visdbility 0 3000 0 2000 0 1000 0 0000 0 100 200 300 400 500 600 700 800 900 1000 Standard deviation of the differential piston nm Figure 1 Differential piston effect on the instrumental visibility in the K band An aproximate formula for the high frequency power spectrum above 0 2 v B with v the wind speed and B the baseline of the OPD variation is RD9 FI sid 9 f 20 0039 m Hz 513 873 m of DE ST VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 12 of 13 v is the effective wind speed r the Fried parameter and f the frequency As r is proportional to ADD the power spectrum does not depend on the wavelength explicitely but only on the seeing conditions e g an ro of 15 cm at 0 5 um considered in the following as typical for Paranal corresponds to a value of r 5 7 m at 10 um r 1371 9 A m A value of 15 m s for the wind speed is assumed in the following formula oi Pdf 4 2910 Tv m Fy 1 T h which gives V T e T s
112. f the statistical uncertainty brought by the combination of the two visibility measurements coming from the two interferometric channels of VINCI is justified theoretically only if the two values are not correlated In reality this is not really the case but the approximation is usually good The uncorrelated approximation is assumed in this study but further investigations will be necessary once VINCI is working ar EE VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 16 of 17 5 9 STATISTICAL PHOTON SHOT NOISE Due to the statistics of the individual photons arriving on the detector a noise proportional to the square root of the number of photons or power is introduced in the measurements The contribution of this noise depends on the integration time RElated formulae are given in the Appendix ar EE VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 17 of 18 6 GLOBAL PARAMETERS 6 1 OVERALL PHOTOMETRIC EFFICIENCY Table 10 Photometric transmission coefficients Telescopes Siderostats Siderostats ATs BE UTs UTs with without BC with BC without AO AO Atmospheric transmission 0 90 0 90 0 90 0 90 0 90 Strehl ratio 1 00 1 00 0 31 0 016 0 48 Transversal dispersion 1 00 1 00 1 00 0 98 0 98 Optical train 0 42 0 40 0 35 0 35 0 35 Pupil lateral jitter 1 00 1 00 0 96 1 00 1 00 Fiber injection 0 78 0 78 0 78 0 78 0 78 Fres
113. fiber outputs are well aligned on the detector pixels The coordinates of the pixels used as well as the percentages of the light concentrated in each pixel are stored in the WS sw in order to be added to the data files headers Req 43 A LISA full frame image is stored after this process for reference during the data reduction Req 30 Automated output focus and rotation alignment is a second generation upgrade at this point but it will be mandatory in order to implement fully automated observation templates See section 2 20 2 for description of a possible algorithm 2 17 TEMPLATES The control software shall allow to build up templates which will carry out in a semi automated way an observation or a group of observations and use them to measure astronomical astmospheric or instrumental parameters Req 44 The observations with LdV based on templates are not yet completely defined The sequencing of LdV modes includes the intervention of the observer for some critical tasks such as injection optimization and output alignment The steps requiring the operator action are marked with operator 2 17 1 Autotest Mode Observations Standard Template mode Autocollimation Injection Adjust operator mode Autocollimation Output Adjust operator mode Autocollimation Fringe Search mode Autocollimation Data Acquisition archival of the data 2 17 2 Autocollimation Mode Observations Standard Template mode Autotest Injection Adjust
114. gth cold stars During the data reduction process it is important to know the effective mean wavelength of the observations in order to locate correctly the peak of the fringe power in the Fourier transform of the corrected interferomtetric signals ar Sens VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 22 of 23 If no shape factor correction is applied on the data the error on the visibility estimation can be as large as 1 for extreme cases This is especially important when the star and calibrator have significantly different spectral types It is possible to assemble a database of the shape factor corrections based on the spectral type and therefore to cancel the difference between the science targets and calibrators Though for peculiar targets for which the spectral type is not comparable to a standard type is will be necessary to acquire a low resolution K band spectrum to compute the proper correction Eventually the coupling of a low resolution spectrograph to the VLTI would be interesting to have a full calibration capability ar EE VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 23 of 24 9 CALIBRATED VISIBILITIES 9 1 LIGHT COLLECTORS The first fringes will be obtained with VINCI using the siderostats It is still unclear whether the beam compressors will be present or not There is a 1 4 limiting magnitude difference between the two configurations but the
115. he VLTI archive system In order to check for any previous observation of the instrument To reliably save the observational data the operating system should access the VLTI Archive at the end of the night to write the files of the data obtained during the night It would also be interesting to be able to read in the archive during the observations to check for the previous measurements on the targets VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 55 2 18 2 List of Parameters displayed by the Autotest Autocollimation Modes GUI Origin Description Typical Possibility Update of GUI Frequency Setting VLTI _ Seismic activity warning symbol Continuous No VLTI Time signals Local UT Sidereal Continuous No ICS Status of LdV hw warning symbol green ok red error Continuous No OS Data acquisition OPD range Observation Yes OS Data acquisition fringe velocity Observation Yes OS Data acquisition computed sample interval Observation No OS Data acquisition number of points per fringe Observation Yes OS Current LISA filter Observation Yes OS Technical CCD exposure time Observation Yes DCS Last Technical CCD images beam A beam B Observation No VLTI Positions of the Delay Lines one or two m 2 Hz 1 Hz Yes OS Internal Optical Path Difference m 2 Hz 1 Hz Yes OS Fringe search INB slide OPD increment millimeters 2
116. he perturbations introduced by the atmosphere LdV is made physically of two optical tables separated by a distance of 10 to 15 meters typically e LEONARDO a small optical table bearing the reference sources unit also called artificial star which can be operated without the main VINCI table This is the first table in the optical laboratory just after the telescopes light beams entrance In this document LEONARDO will be considered as a single source but it consists of several distinct light sources Visible Laser thermal source K band Laser on which a fiber is connected manually to send the light to the other parts of the instrument e VINCI ALIU the main instrument VINCI with the fiber injection optics and the alignment tools including the Technical CCD detector and optics for ALIU It is the last optical along the light beams path before the Fringe Sensor Unit FSU It is located just before the FSU on the west side of the laboratory MONA and LISA the fibered beam combiner MONA the output optics and the infrared camera LISA are located on the VINCI table Doc VLT SPE ESO 15810 1852 LEONARDO da VINCI Issue 1 1 Date 16 September 1999 Software User Requirements Page 6 Figure 1 Optical layout of VINCI VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 7 2 2 LDV SYSTEM OVERVIEW 2 2 1 Interferometry Room LdV is located in the VLT interferometry laboratory
117. i1 2 48486 mod 112 09 LE 0 05 then begin table freq i 1 0 endif else begin table freq i 0 0 endelse endfor for i 100 300 do begin if abs fft model i GT limit2 then begin E if i 2 48486 mod 112 09 LE 0 05 then begin table freq i 1 0 VLT TRE ESO 15810 2330 1 0 5 October 2000 49 LISA Test Report endif else begin table freq i 0 0 endelse endfor for i 300 513 do begin if abs fft_model i GT limit3 then begin if i1 2 48486 mod 112 09 LE 0 05 then begin table freq i 1 0 endif else begin table freq i 0 0 endelse endfor transfer function table complexarr 1024 100 for i 0 1023 do begin for j 0 99 do begin transfer function table i j fft chop i j fft model i R Normalization to the noise of the open shutter transfer function table i j fft chop i j fft model i fft const mean i endfor endfor for i 0 512 do begin transfer function i mean abs transfer function table i table freq i endfor plot transfer function xrange 0 600 yrange 0 5 1 5 end 11 7 MODEL GENERATION pro gen model period size min level max level rise model SYNTAX gen model period size min level max level rise model generates a model of the chopping with a defined rise time period and size version 21 09 2000 for n 0 fix size period 2 do begin i period n for j fix i fix itrise 2 do model j
118. iciency of VINCI VLTI m K magnitude of the star Fo Reference flux for mag 0 3 9 10 W m micron At Duration of one scan s S Signal in each of the four channels photometry and interferences m 2 cl AA rino This can also be written as Scan length um n Number of samples per fringe Nchanneis Number of channels onto which the signal is detected Effective central wavelength um AA Wavelength range um A Effective single telescope collecting area m T Overall photometric efficiency of the optical train m Fo S Overall interferometric efficiency of VINCI VLTI K magnitude of the star Reference flux for mag 0 3 9 10 W m micron Signal in each of the four channels photometry and interferences La Acquisition frequency Hz VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 29 of 30 E t S A AA riae WE 10 3 THERMAL NOISE Tx Temperature K Scan length m n Number of samples per fringe Effective central wavelength m AA Wavelength range m E Beam etendue Emissivity B Blackbody function W m sr m At Scan duration s Oerma Thermal noise equivalent power W Hz h B T A VA AKT e thermal 7 2hc gall 10 4 PISTON NOISE S Signal in each of the four channels X Fraction of the star signal eq
119. if telescope 4 ar T VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 33 of 34 A 50 265 Area of the telescope m2 T 0 00342 Photometric transmission I 0 722 Interferometric efficiency end ct These parameters are not lambda 2 2e 6 delta lambda 0 4e 6 Wavelength range m m magnitude K magnitude of the star FO 3 9e 10 K magnitude 0 flux W m2 MICRON 1 51 2 2 2e 6 Effective total scan length m l fringes 70e 6 Fringe packet length lescope dependant Effective central wavelength m oe oe oe oe oe n 5 Number of samples per fringe Temp 288 Temperature K E 1 Beam etendue epsilon 1 Emissivity QE 0 62 Quantum efficiency of the detector RON 10 Read out noise of the detector e h 6 6226e 34 k 1 38062e 23 c 3e8 sigma_min_integ 3e5 sigma_max_integ 6e5 oe Planck constant J s Boltzmann constant J s Speed of light m s inimum integration wavenumber m 1 fringe power aximum integration wavenumber m 1 fringe power o oe oe oe 9 2 Acquisition frequency of the camera Hz frequency n scan speed lambda if frequency lt le 6 frequency 1e 60 end Scan duration dt l n frequency lambda Frequency bandwidth df lambda frequency n sigma max integ sigma min integ Signal S 2 A delta lambda le6 T I F0 10 m 2 5
120. image produced by the array after cooling it down The detection of possible dead pixels can be done by subtracting the images obtained from a reference image and by adjusting the display levels to check for the pixels having no evolution between the two frames no noise associated with them During the tests of the camera a series of images of the full 512x512 quadrant have been saved for future reference No obvious variation of the dead pixels positions have been observed visually during the tests VLT TRE ESO 15810 2330 1 0 5 October 2000 38 LISA Test Report 9 MEMORY EFFECT 9 1 DESCRIPTION As stated in the LISA User s Manual the Hawaii detector may be subject to a memory effect which could affect the visibility measurements and even the observational procedure The symptom of the detector memory is that after a bright illumination of the photosite causing or not a saturation of the pixel a residual signal is kept for a while and affects the readouts for the following cycles Three kinds of behaviors are tested e Minute timescale Saturation of the detector followed by closure of the cold shutter This test ensure that the array does not need a rest time between a bright source and a faint source observation The effect of the saturation should disappear completely in a short time The tests were not conducted on the LISA camera but measurements have already been obtained by Gert Finger on the HAWAII detector His
121. in K min mean max V min mean max All available bands magnitudes U B V K L IRAS IUE min mean max All available color indexes B V J K with uncertainties All available angular diameter estimations theoretical models lunar occultations with publications references Computed angular diameters through different methods spectral type distance bolometric flux with uncertainties in milliarcsec and in radians Variability flag Variability type if applicable type code taken from the GCVS for example Binarity or multiplicity flag Spectral type s determination s with publications references Period seconds and days epoch julian date wavelength of measurement meters reference publication for periodic variables Comments entered by the observer LdV mode name or code LdV setup name or code Wavelength range mini mean maxi meters vector for dispersed fringes VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 40 LdV OPD offset meters Expected visibility on the current baseline mean over the observation with uncertainty dimensionless Technical CCD extracted parameters position brightness of the object Technical CCD refined optimization intensity map Number of scans saved Number of scans on source Number of scans off source Number of scans beam A with one telescope off source Number of scans beam B with one tele
122. is located at the lower left corner of the detector as seen on the RTD screen in order to have the best possible readout speed Though there are a number of dead or bad pixels in that region that should be avoided 5 1 3 Diffuse illumination The illumination of the detector with a diffused source gave a uniformly decreasing intensity from the center to the edge of the detector radial The decrease is nearly perfectly linear and the noise level seems to be approximately constant It is difficult to estimate the impact of this effect on the fringe measurements before the noise tests have been conducted on the system 5 1 4 Focused saturated spot As a second step the focused light source spot was moved over the array in order to estimate what fraction of the detector is usable to image the fiber head The problem here is that the beam is easily vignetted by the window of the dewar if its direction is too much tilted The practical experiment was done by measuring the intensities across one line and one column A problem was identified after the measurement the spot was heavily saturated due to a too long exposure time more than 2 seconds Therefore the usable area defined in this section will have to be checked with a lower intensity beam and the LdV four fibers bundle It is important to bring the spots as close as possible to the lower left corner of the array to optimize the readout frequency VLT TRE ESO 15810 2330
123. ital Unit ALIU The Alignment Unit DCS Detector Control Software DP Data Pipeline DL Delay Line DLCS Delay Line Control System FSU Fringe Sensor Unit GUI Graphical User Interface GEI Graphical Engineering Interface HW Hardware ICS Instrument Control Software IN Inserted IWS Instrument Workstation LCU Local Control Unit LdV LEONARDO da VINCI the whole instrument LEONARDO The artificial star subsystem LISA The HAWAII based infrared camera LISA WS The LISA LCU workstation MONA The fibered recombiner N A Not applicable OPD Optical Path Difference OS Observation Software OUT Removed SNR Signal to Noise Ratio SW Software TBC To Be Confirmed TBD To Be Defined TCCD ESO Technical CCD TCS Telescope Control Software VCM Variable Curvature Mirror VINCI The main optical table of LdV VLT Very Large Telescope WS Workstation 1 5 GLOSSARY Batch a hundred to a thousand scans In order to decrease the statistical noise many interferograms are acquired in a row The term batch designates this collection of scans Doc VLT SPE ESO 15810 1852 LEONARDO da VINCI Issue 1 1 Date 16 September 1999 Software User Requirements Page 3 Detector Control Software DCS the DCS is responsible to control one detector system It resides partly on the LCU for the direct interface to hw and real time issues partly on the Instrument Workstation for not real time issues For LdV we have 2 DCSs one for the IR science ca
124. ject once per observation Universal time of the beginning and end of the scans once per scan Number of data points saved in each scan once per observation Number of scans saved once per observation Number of scans on source once per observation Number of scans off source once per observation Number of scans beam A once per observation Number of scans beam B once per observation Name of calibrator s and reference code of calibrator files used for this target once per observation Name of the original file of the observations once per observation Refraction parameters air mass refraction correction applied by the VLTI beginning and end of each observation 7 Delay Line once per scan parameters read from the DL system Delay line position meters Delay line velocity meters per second Delay line piezo relative position meters Delay line piezo velocity meters per second Delay line VCM curvature curvature radius Delay line relative pupil position meters Delay line error signal meters 8 FSU parameters read from the FSU system Piston RMS value measured bu the FSU once per scan Residual FSU error signal once per scan 9 Telescopes once per observation unless otherwise specified parameters read from the telescopes system VLTI setup baseline used telescope types UT names Name of the delayed telescope RA Dec beginning and end of
125. justed to optimize the speed of the FFT computation depends on the algorithm used 2 9 6 3 Fringes Detection and OPD Offset The fringe detection follows a simple scheme The filtered signal is scanned for any point which is over 5 times the standard deviation If one such point is detected in the signal then the signal fringes detected can be issued A list of all these points is built containing their temporal coordinates for example one point over 5 sigmas detected at 0 019 s after the beginning of the scan If this list is empty no fringes were detected The median value of the list is computed and it corresponds to the position of the fringes centroid center Finally this temporal position is converted into the OPD offset taking into account the speed of the piezo scan and sent to the delay line 2 9 6 4 Alternative Algorithm for Fringe Detection and OPD Offset The phase function could be used to determine the precise location of the fringe packet but this should be considered currently as a second choice alternative to the previously described algorithm 2 9 7 Synchronized Data Acquisition Parameters SYNC VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 31 As the fiber outputs of MONA are imaged on a few pixels of LISA only those pixels are read during observations The coordinates of the pixels which are to be read are defined by the user Req 50 The shapes
126. k which is moved quickly back and forth 0 1 20 Hz in order to modulate the optical path difference between the two beams This is done to scan the fringes while they are recorder by the LISA infrared detector The synchronization of the motion of INA3 with the LISA detector is a critical real time process which is described in the section 2 9 2 5 SUMMARY OF LDV MOVABLE HARDWARE POSITIONS Unit Element Range Values ARTU BSA BSA1 BSA2 OUT ARTU BSB BSB1 BSB2 OUT ARTU ART3 ON OFF See section 2 8 6 VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 15 ALIU ALI ALI1 ALI1S OUT ALIU ALI5 ALI5 ALIBS OUT ALIU ALI Slide ALIS ALIA FREE ALIU TCCD IN OUT Lens ALIU TCCD 0 25000 TBC microns Focus PRESET1 PRESET2 COMU INB 0 25000 microns COMU INA1 0 25000 microns Focus COMU INA1 0 12500 microns Tip COMU INA1 0 12500 microns Tilt COMU INB1 0 25000 microns Focus COMU INB1 0 12500 microns Tip COMU INB1 0 12500 microns Tilt COMU OUT1 0 12500 microns Focus COMU OUT1 0 12500 microns Tip COMU OUT1 0 12500 microns Tilt COMU OUT1 0 20000 10 3 degrees Rotation COMU INA3 See section 2 9 COMU POLAA 0 360 TBD degrees COMU POLA B 0 360 TBD degrees COMU FILT K KPRIME H NARROW OPEN CLOSED
127. l magnification factor the fiber images are separated by 9 pixels of 18 5 microns while the fibers in the bundle are separated by 125 microns The real value may differ by a few percents from this value Therefore the resulting motion for the doublet is AX jouer 1 3322 x 214 380 um 4 5 RESULT OF THE FOCUSING 4 5 1 Procedure The focusing correction computed 380 microns was applied to the detector mount towards the doublet by machining the positioning reference device After this step the OUT light source was collimated precisely using an autocollimator and the resulting beam was sent to the detector on pixel 545 38 coordinates VLT TRE ESO 15810 2330 1 0 5 October 2000 12 LISA Test Report 33 38 from the lower left corner of the detector corresponding to the center of the usable field The beam position was adjusted at sub pixel precision to maximize the flux in the target pixel 4 5 2 Intensity profiles The resulting profiles in the X and Y directions are presented in the following figures measurement 10 In this measurement the 64x64 window readout mode was used at the maximum frequency The spot was not saturated on the detector Intensity ADU along column Xz33 1000 1000 3000 5000 7000 Intensity ADU along line Y 38 1000 1000 3000 5000 7000 4 5 3 Energy in the target pixel The energy contained in the target pixel is 56
128. lue is 19 18 ADU inverted scale in double correlated mode the higher values values are negative and the standard deviation is 2 06 ADU Using the gain factor computed in section 5 3 1 the resulting read out noise in double correlated mode is 6 electrons This value is a very rough first estimate of the effective readout noise of the array Please refer to Section 0 for precise measurements 5 4 READOUT MODES 5 4 1 List of modes The useful modes for the camera operation are listed in the following table They include the possibility to switch from one windowed readout mode to another to adapt the size of the read windows to the brightness of the star for example This list is likely to be updated VLT TRE ESO 15810 2330 1 0 5 October 2000 22 LISA Test Report Table 4 LISA readout modes Uncorrelated 512x512 pixels Double Correlated 512x512 pixels Uncorrelated 64x64 pixels Double correlated 64x64 pixels Double correlated 4 windows of 1x1 pixel Double correlated 4 windows of 2x2 pixels Double correlated 4 windows of 3x3 pixels 5 4 2 Off line windows generator In order to define the windows for the Beam readout mode four windows an off line generator was delivered by MPE This tool is especially useful during the commissioning and test phases The clock pattern files generated contain some syntax errors and require some manual correction to work properly 5 4 3 Full quadrant
129. ly the position of the micrometric knob P1 The reading can be done at a 1 micron precision but the final precision is probably something like 10 microns Position the beam on the detector at the position where the focus is necessary Obtain the best possible focus by adjusting the focus knob on the fiber head mount Take down the new position of the micrometric knob P2 The difference with the previous reading P2 P1 gives the focus difference between infinity and the current position of the doublet A multiplicative factor corresponding to the magnification factor has to be applied to have the physical motion to give to the doublet or the detector See Section 4 4 for the complete computation 4 3 EXPERIMENTAL RESULTS Table 1 Fiber head motion to obtain the focus on the Hawaii array X Fiber Energy Comments pix pix pix ollima Infinity i in one mm pixel White 4 911 4 615 0 296 Y 550 3 48 2 550 38 20 42 9 Black 4 951 4 614 0 337 New autocollimator background 550 38 Black 4 949 4 619 ERREUR New autocollimator background evaluation problem Black 4 952 4 659 0 293 Difficult reading of focus 5 529 17 43 46 2 Black 4 950 4 709 0 241 50 Very difficult measurement slow and new lateral positioning of the fiber Precise measurement 6 548 se 35 50 2 Black 4 942 4 730 35 730 0 211 54 35 Black 4 937 4 723 0 214 54 Precise mea
130. lysed ii E A 7 4 1 2 Examples of individual pixels PSD 7 4 1 3 Median low frequency PSD over the 64x64 pixels 7 4 2 High frequency 0 828 Hz as LR dese etie t eid er dues DER De ated ee 7 4 2 1 Data analysed 7 4 2 2 Example of pixel 33 38 7 4 3 Maximum frequency 0 1223 HELL 7 4 3 1 Data analysed E 7 43 2 Example of pixel 33 38 7 5 DISCUSSION 5 eae ne ee te ere as eae be eget ene ean enna manana ORIGINE 8 INDIUM LAYER STABILITY 5 1eeendie scuto eoo tpa ro rodea loeum retenu bana orbs eras esee pps as ene aerae era Dean Re eina sen EE 37 9 MEMORY EFFECT e 38 9 1 DESCRIPTION EE 38 9 2 MEASUREMENT PROCEDURE scsscsssssessseessseeeeseeseseesenesscsessceessceacseescsessesesseeecseeseaessceesseesseeaeaeeaseeeaseees 38 9 3 GEOMETRY OF THE CHOPPING SN STEM nennen enini tenne treni trennen enne tente 38 9 4 DATA REDUCTION EENS 39 9 5 DISCUSSION p 40 10 SUMMARY OF PROBLEMS wicssccsscsscscsscsocececescescsasscsasesosescnsesssseeseseesesassecteesseescassensesessesassscecsexesedecese 41 11 APPENDIX IDL SIGNAL PROCESSING PROCEDURES cese eren ette eene enata seta setae stnue 42 11 1 READ DATA de oise 42 11 2 PROCESS 64X64 E E NEE 43 11 3 Nee A4 11 4 PLOT GAXOA DATA etr tet ot t RES RR ER o dede t ee peg ee Ho 46 11 5 P
131. measurements show that the signal intensity is divided by 100 in 150 seconds e Millisecond timescale Array response to a chopped signal stimulus cut very quickly to zero in less than 1 integration time This measurement makes it possible to evaluate if the fringe visibility could be degraded by a memory effect of the detector The only delicate point in this test is that the shut off of the light intensity has to be very fast This is achieved easily by masking quickly the fiber head with an opaque material on the light injection side The very small diameter of the fiber core allow to achieve sub millisecond closing times A chopping device was used for that purpose 9 2 MEASUREMENT PROCEDURE Align the light source in order to have a good quality point like source on the detector Adjust the intensity of the light source to obtain an intensity on the detector just below saturation Close the cold mask of the detector for a while in order to avoid contamination by a possible long term memory effect Open the cold mask and select the K band filter Start the chopper wheel in front of the fiber injection Start the data acquisition at the maximum data rate on a four windows set containing the illuminated pixel Acquire two sets of data shutter open and shutter closed to have the top and bottom levels 9 3 GEOMETRY OF THE CHOPPING SYSTEM The geometry of the chopping system determines the shape of the input signal sent to
132. mera LISA and one for the TCCD Frame 4 pixel values A frame is a set of four numbers which are elementary values of the four signals coming out of LISA 2 interferometric flux values and 2 photometric flux values dt acquisition 1 millisecond They are the basic information elements provided by LISA Instrument Control Software ICS it is responsible to control the whole instrument hw except the detectors It resides partly on the LCU for the direct interface to hw and real time issues partly on the Instrument Workstation for not real time issues Instrument modes LdV is designed both as an engineering and observing instrument The instrument modes foreseen for LdV are e Autotest Autocollimation Stellar Interferometer Pupil Check Image Check Artificial Star LEONARDO alone Instrument setup The term setup designates the hardware setting of the different optical and mechanical elements on LdV All the LdV setups are associated with an instrument mode The setups are subsets of the instrument modes Several setups are associated with a single instrument mode Instrument status The current setup of LdV Observation four batches on source off source beam A beam B During an observation dt acquisition 1 to 10 minutes four batches are obtained off source about a hundred scans on source about a thousand scans beam A only about a hundred scans beam B only about a hundred scans A pointe
133. mpressors m Siderostats with beam compressors m ATs and UTs without Adaptive Optics same curve m UTs with Adaptive Optics DE EE VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 19 of 20 10 Precision on the visibility 96 d 12 13 K Magnitude 0 001 The maximum statistical precision theoretically attainable is about 0 002 for the brightest stars The limitation for the UTs comes from the photon noise and is due to the limited maximum speed of the piezo mirror V 4 OPD 1 cm s A practical limit for the observations can be set at 2 limiting precision in 100 scans After this limit the calibration of the observation can be problematic but the system can be usable up to a 5 precision This gives the limiting magnitudes listed in Table 12 It should be stressed that the figures given in this table are for the statistical uncertainties only They are useful to prepare an observation as they give an estimate of the limiting magnitude of a single star observation but the final calibrated visibilities will have a lower precision due to the uncertainties on the calibrator angular size or other factors see Section 9 for details Table 12 VINCI statistical limiting magnitudes Collector Limiting Limiting magnitude 2 magnitude 5 precision precision Siderostats without beam compressors 4 6 6 1 Siderostats with beam compressors
134. mum of A 8D Under normal conditions and assuming that the relevant corrections are applied this effect will not cause any flux loss and is neglected in the following computations 5 6 2 Inside the K band The K band spans from 2 to 2 4 microns Between the two extreme wavelengths of the band the atmospheric transversal dispersion will not be the same For a star far from zenith this results in a slightly elongated star image on the fibre head and therefore in a degraded injection The modeling done by Francoise Delplancke RD7 give a maximum flux loss of 4 696 for the UTs at 60 degrees of zenith angle and negligible values for smaller zenith angles and other telescopes 0 5 The following flux loss values are assumed in the following half of the maximum values m 0 for the siderostats m 0 2 for the ATs m 2 3 for the UTs 5 7 NON STATIONARITY OF THE ATMOSPHERIC CONTRAST LOSS Once the photometric corrections have been applied to the interferometric signals the visibility measured by the instrument should always be the same is no piston was present In reality the contrast loss due to the atmosphere itself not to fast turbulence and its large scale evolutions is still to be studied but it is certainly at a very low level below the current foreseen precisions without fringe tracker This effect was therefore not considered in the rest of this study 5 8 CORRELATION OF THE TWO INTERFEROMETRIC CHANNELS The reduction o
135. n Yes OS Current LISA filter Observation Yes DCS Last LISA full frame Observation No DCS Last Technical CCD images beam A beam B Observation No OS Technical CCD exposure time Observation Yes VLTI Positions of the Delay Lines one or two m 2 Hz 1 Hz Yes VLTI Velocity of the Delay Lines m s 2 Hz 1 Hz No VLTI Status of the Delay Lines track no track 2 Hz 1 Hz No OS Internal Optical Path Difference m 2 Hz 1 Hz Yes OS Fringe search OPD increment millimeters 2 Hz 1 Hz Yes OS Internal OPD current relative offset millimeters 2 Hz 1 Hz Yes DCS Visibilities plot time visibility 2 Hz 1 Hz No OS Visibilities plot settings axis Observation Yes OS Visibilities plot statistics last mean sigma 2 Hz 1 Hz No DCS Visibilities histogram visibility number of values 2 Hz 1 Hz No OS Visibilities histogram settings axis Observation Yes OS Fringes found warning signal 2 Hz 1 Hz No VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 58 2 18 6 Interface for the Engineering and Maintenance Modes The Engineering Mode is generally speaking less complex than the Stellar Interferometer Mode The user interface for this mode is basically the same as for the online modes with access to all the instrument parameters Req 48 either via a GUI or a command line During the maintenance and setup procedures the needs
136. n the intensity on the LISA corresponding photometric and interferometric pixels is continuously monitored as well as the position of the tip and tilt motors After the spiral motion has reached a radius of one speckle cloud 0 5 1 arcsec projected on the sky the data acquired is an array of tip and tilt positions in mm associated with intensity values in ADUs VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 62 3 Do the same thing with the other injection mirror INB1 which gives you another array of tip tilt intensity values 4 Fit the measured tip tilt intensity data with a 2 dimensional gaussian curve in order to find the precise position of the maximum intensities for the two beams This gives the offsets to apply to the injection optics to maximize the injected flux The main source of noise is the intensity variations due to turbulence 5 Correct the position with the tip tilt mirrors finally the computed offsets are sent to the INA1 and INB1 mirrors 6 Fringe search At this stage the injection should be optimized and the fringe search can begin 2 20 2 Automated Output Alignment This algorithm is TBD but will be for the focusing part very similar to the TCCD focus algorithm For this procedure LEONARDO artificial star is on thermal light source The focus setting of the fiber output is remotely adjustable The precise focusing of the four fiber outpu
137. n of 512 we can estimate the quantity of data contained in one scan if the data from LISA are coded on 16 bits one typical scan represents 4 kbytes of data Only 2 kbytes 256 frames centered on tzoppexp per scan are saved In a seond generation upgrade of the instrument it is foreseen to insert a dispersive element in front of LISA to disperse the four output signals This will result in an increase of the data quantity produced at each scan by a factor which could be up to 50 see section 2 20 3 The software system should be able to manage about 100 kbytes per scan Req 19 but the associated electronics hardware should not be considered for the first phase of LdV LEONARDO da VINCI Software User Requirements VLT SPE ESO 15810 1852 1 1 16 September 1999 27 A scan begins at a time ti We call Atpaa tpaa eng toaa sta the duration of useful data collection the ratio Atpag ti t defines the duty cycle i e the efficiency of LdV This duty cycle should be as high as possible 95 90 Req 20 2 9 3 Chronology of data acquisition The following chronology assumes that the quick look analysis is performed by LISA WS This is necessary because no real time processes can be done by the OS Time Definition LdV VLTI event VINCI ICS event LISA DCS event Wait for Reset LISA Delay line in READY status Piezo in READY status LISA in READY status Check that DAQ is requested
138. nce or twice per night the image of a photometric reference in order to estimate broadband magnitudes of the targets The procedure to obtain these measurements is not different from the pointing of the science targets The desired magnitude precision is 0 1 mag It is understood that this magnitude is not directly comparable to the standard systems magnitudes as it is a broadband integration based on the sensitivity curve of the TCCD This measurement could be useful to estimate the parameters for the infrared camera exposures Page laiss e vierge intentionnellement 3 LISA Test Report EUROPEAN SOUTHERN OBSERVATORY Organisation Europ enne pour des Recherches Astronomiques dans l H misph re Austral Europ ische Organisation f r astronomische Forschung in der s dlichen Hemisph re VERY LARGE TELESCOPE LEONARDO da VINCI LISA Test Report Doc No VLT TRE ESO 15810 2330 Issue 1 0 Date 5 October 2000 Prepared P Kervella eee Name Date Signature Approved A Glindemann Name Date Signature Released M Tarenghi eee Name Date Signature VLT PROGRAMME TELEPHONE 089 3 20 06 0 FAX 089 3 2023 62 VLT TRE ESO 15810 2330 1 0 5 October 2000 LISA Test Report CHANGE RECORD ae Date Section Page affected Reason Initiation Remarks 7 August 2000 First draft alignment procedure 6 September 2000 Includes first tests with the windowed readout mode 21 September 2000 Sections 5 6
139. ndle light in each pixel LISA exposure time seconds LISA full frame exposure time seconds LISA number of data points between the start of the acquisition and the start of the record left margin LISA number of data points between the end of the record and the end of the acquisition right margin LISA quicklook fringe detection level n sigmas Fast scan piezo mirror incidence angle radians Fast scan piezo total OPD range meters Fast scan piezo scan OPD range meters Fast scan piezo scan OPD total OPD Fast scan piezo voltage range Volts Fast scan piezo voltage offset Volts Fast scan piezo waveform signal type sine square sawtooth triangle Fast scan piezo waveform stiffness 96 Fast scan piezo waveform frequency Hz Fast scan piezo waveform delay seconds Fast scan piezo number of waveform samples Fast scan piezo waveform sampling frequency Hz Fast scan piezo output sample interval Hz Fast scan piezo sample time seconds OPD introduced by the Piezo Mirror meters can be computed after the acquisition through the piezo motion parameters once per frame 9999 9 99 9 VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 36 Piezo camera user settings OPD range fringe velocity sample interval FSU usage with without 5 Object reference data This section describes the data which should be sto
140. ned modes described in the following table Req 12 No low level command is foreseen to be accessible from outside of LdV software Unit Element Artificial Artificial Artificial Artificial Autocollim Star Star Star Star ation Off On On Off On Removed Inserted Removed Inserted Inserted ARTU BSA OUT BSA1 OUT BSA1 BSA2 ARTU BSB OUT BSB1 OUT BSB1 BSB2 ARTU LEONARDO OFF ON ON OFF ON Light source 2 9 DATA ACQUISITION ASPECTS OF LDV 2 9 1 Description The piezo mirror INA3 is used to quickly modulate the optical path difference between the two beams in order to scan over the fringes The frequency of its motion is adjustable from 0 1 Hz to 20 Hz Req 13 The most common frequency which will be used is 10 Hz The scan length is also adjustable from 1 micron to 360 microns Req 14 The wave used to control the piezo has a smooth triangle shape which is intermediate between a triangle and a sinusoid The corresponding factor for the softening ratio is the wave shape factor whose range will be between O triangle and 1 sinusoid Req 15 The generation of this curve is made at the LCU level based on the defined parameters VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 25 The most important difficulty for the piezo is that it has to be precisely synchronized with the LISA acquisition during the scanning of the fringes at a submilli
141. nel losses 0 85 0 85 0 85 0 85 0 85 Fiber injection tip tilt residuals 0 70 0 70 0 80 0 80 1 00 Triple coupler 0 75 0 75 0 75 0 75 0 75 Outputs imaging 0 30 0 30 0 30 0 30 0 30 Signal splitting 0 25 0 25 0 25 0 25 0 25 Quantum efficiency 0 62 0 62 0 62 0 62 0 62 TOTAL 0 00612 0 00582 0 00173 0 0000914 0 00342 6 2 OVERALL INTERFEROMETRIC EFFICIENCY Table 11 Interferometric transmission coefficients Telescopes Siderostats ATs UTs without UTs with AO AO Sampling losses 0 80 0 80 0 80 0 80 Telescopes and optical train 0 95 0 95 0 95 0 95 Triple coupler 0 95 0 95 0 95 0 95 TOTAL 0 722 0 722 0 722 0 722 ar EE VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 18 of 19 7 STATISTICAL VISIBILITY PRECISION CURVES 7 1 SIGNAL TO NOISE RATIO The signal to noise ratio is given by the following formula Q S 4 4 4 NH Dos O piston 4 photon 0 the fourth power on the NEP come from the fact that the signal is the squared visibility of the fringes 7 2 OPTIMAL OPD SCAN SPEED The effective precision attainable with VINCI will depend on the exposure time used on each star The optimized parameter is the OPD scan speed v in optical path difference which is twice the piezo mirror motion speed As a general rule VINCI will sample the fringes at a fixed n 5 points per fringe optimal value gives directly the readout frequency of the detector through
142. nflicts The default assignment of LEONARDO should be LdV All the LEONARDO light sources are not all immediately accessible on line Only one is available remotely at a time and the switching from one source to another has to be done manually in the laboratory The following table lists the light sources which are foreseen on VINCI reference document 7 Source type Bands Single multi mode Red laser diodes 2 Visible Single mode 2 3 microns laser K Single mode Thermal K Single mode VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 24 Thermal K Multi mode Thermal N Single mode Thermal N Multi mode Thermal Visible Multi mode This mode will be used mainly after the commissioning phase when the science instruments AMBER and MIDI are operational The electronic racks of LdV will be on as well as the WS sw thus allowing normal operation of the motors and light switches of LEONARDO The light source can be turned ON without being injected in the optical beams in order to pre heat it It can also be turned OFF with the cubes IN to check for the effect of the cubes on the optical transmission This mode and all the related data should be accessible from outside of LdV software Req 11 There is no remote intensity adjustment foreseen for any of the sources on LEONARDO The external instruments should access LEONARDO only through the predefi
143. ns of LdV Req 41 The main drawback is a slow down of the observations and the interruption of the automated template observing session Refined injection optimization is a second generation upgrade at this point but this is very important since as long as we do not have automated injection optimization it will not be possible to implement and test fully automated observation templates which is the target operating mode for the VLTI instrumentation See section 2 20 1 for description of possible algorithms 2 16 2 Output Alignment Procedure The four fiber outputs have to be precisely positioned on four pixels of the LISA camera to a fraction of a HAWAII pixel This means that the position of the fiber bundle output has to be precisely adjusted with respect to the HAWAII detector pixels This is done via the OUT fiber bundle output which is adjustable in translation 3 directions and rotation Only the focus and rotation aroud the optical axis are remotely controlled The alignment is checked on the HAWAII camera by acquiring full frames which cover a quarter of the chip surface i e 512x512 pixels while light is sent into the fiber bundle with the artificial star light source LEONARDO The output alignment is normally done before the observing night using the Autotest or Autocollimation setups However it is possible to do refocusing or rotational adjustment during the night if necessary as it is remote controlled A possible scheme
144. nt errors and the output wavefront is only affected by flux variations These variations inject some power in the frequency range of the fringes and therefore can affect the fringe visibility To reduce this effect the photometric signals are used to calibrate the interferometric outputs during the data reduction Further simulations are required to estimate the impact of this kind of noise on the measurements 5 4 INTERNAL TURBULENCE T VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 14 of 15 As stated in RD1 p 20 the equivalent r for the internal seeing is 125m at 2 2 microns Therefore the effect of internal seeing is neglected in the rest of this study 5 5 ATMOSPHERIC TRANSMISSION IN THE K BAND The following curves are taken from the www site http www eso org gen fac pubs astclim paranal h20 ISB n A 10mm H 0 WARUM CRM Mn uM CURL Nd d 2 TNI INMAL Wi nu Ti i if Aaa Ph FA ui 08 j 1 i AM 2 0 4 D 1 v Transmission o o A D _ em 1 1 1 1 a L 1 18 20 2 1 22 23 24 25 26 27 Wavelength im Figure 3 Atmospheric transmission above Paranal in the K band for two airmasses AM and 1 mm precipitable water vapor Typical values on Paranal are 1 5 winter to 2mm summer LLUTEZS4p Sel T T All Data megdian z min Winter Only median 1 S mm FREO Processing Lund Observatory Q 1 2 3 4 5 6 7 8
145. nts saved in each scan Observed internal OPD meters Available from the FSU TBD Piston RMS value measured bu the FSU once per scan Residual FSU error signal once per scan Available from the DL system Delay line position meters Delay line velocity meters per second Delay line piezo relative position meters Delay line piezo velocity meters per second Delay line VCM curvature curvature radius Delay line relative pupil position meters Delay line error signal meters 2 11 2 3 3 Observation 9999 Reference code of the observation can be sequential allows to quickly retrieve the data Names as taken from Simbad for example HD SAO HR HIP Name of calibrator s and reference code of calibrator files used for this target Type target calibrator type of artificial source Calibrator quality rating from 1 best to 5 worse Name of the original file of the observations Theoretical RA Dec LISA Calibrations TCCD last images telescope 1 telescope 2 possibly with the artificial light source on taken before fringe acquisition once per observation 9 9 VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 39 Opto mechanical computer controled elements positions readings meters angular degrees see list in the pre
146. o begin for j 0 63 do begin for k 0 128 do begin mean psd per pix i j k mean psd i j k endfor endfor endfor Computes the median PSD over the 64x64 pixels for dark exposures print Median PSD over the array at light level 0 median psd fltarr 129 for k 0 128 do begin median psd k median psd 0 k endfor Prints some data for checking print Gains print gain 0 10 1 print RON print readout noise 0 10 1 print print means 0 1 32 32 print variances 0 1 32 32 print psd 0 1 32 32 121 end 11 3 PROCESS SCANS pro process scans flux means variances fourier psd mean psd SYNTAX process scans flux means scans variances scans fourier scans psd scans mean psd scans Processes the test data acquired with the LISA camera in order to VLT TRE ESO 15810 2330 1 0 5 October 2000 45 LISA Test Report extract the best possible pixels for the VINCI beams This procedure makes the computations for the scans data cubes and produces the power spectral density for all the pixels used This is the second step of the best pixels selection as described in the document LISA tests in Garching Version 13 09 2000 Author P Kervella Computes the mean value and variance of each pixel over the sequenc It excludes the last acquisition partly saturated n pix n elements flux 0 0 n frames n elements flux 0 0
147. of the four windows is not necessarily rectangular as it might be more efficient for example to read an L shaped figure if the coma is too large The rate at which the four windows are read is set by the user Req 24 depending on the brightness of the observed target This setting can be computed by the WS SW from other parameters described hereafter entered by the user and that are related to the piezo mirror motion and fringes characteristics The total OPD range is the optical path length covered by the piezo motion the fringe velocity is the speed at which the fringes are moving during the scan There is a direct relationship between the fringe velocity and the piezo mirror scan speed The sample interval is the physical OPD length over which the pixel integration is made From the user point of view the parameters of the acquisition should be Parameter Range Values Data Acq Total 1 300 microns OPD Range Data Acq 1 2500 microns s Fringe Velocity Data Acq 0 1 3 microns Sample Interval 1 10 points per fringe After these parameters have been set by the user the LdV ICS computes the corresponding values of the frequency voltage range and optimal wave shape factor for the control of the Fast Scan Piezo the piezo frequency is directly given by Fringe Velocity in microns s OPD range microns the voltage range is a direct function of the s
148. of the user are mainly to have full control on every element possibly at the expense of the ease of use The user interface for this mode is basically the same as for the engineering mode with access to all the instrument parameters either via a GUI or a command line 2 18 7 Setting up the Instrument Parameters The instrument parameters for each standard setup listed in the section 2 6 should be loaded after the choice of the mode and optical setup of LdV Autotest Stellar Interferometer Pupil Alignment Image Alignment They will define the position of each hardware element motor light switches piezos for each mode The operator should have access via the GUI to the different parameters settings and be able to change them even after a setup has been loaded Req 49 Use of encoder values to set the functions shall be possible via a dedicated Hardware Setup interface VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 59 2 19 LIST OF NUMBERED REQUIREMENTS Req Subject Section 1 Status of hardware elements 2 4 2 HW warnings and alerts 2 4 3 Slide positioning direction BSA BSB 2 4 1 4 Slide positioning direction ALI1 ALI5 2 4 2 5 Slide positioning direction ALI 2 4 3 6 TCCD preset positions 2 4 4 7 Slide positioning direction INB
149. om then insert the bonnette if the pupil is less than 3m away preset the TCCD focus to the expected pupil position Once the focusing has converged on the pupil LED from the focus motor readout the software should derive a measured pupil longitudinal position Req 38 which can be compared to the expected pupil position provided by the VLTI OS The necessary setups have been described in the section 2 8 4 The focusing of the TCCD is done the same way as for the Image Check only the starting position rough focusing changes In the current design the role of LdV is not to provide the commands to set the mirrors of the optical train to align the pupil Only the display and shape measurements of the pupil are a requirement for the LdV software But this has TBC when a clear alignment procedure has been defined The procedure to obtain and measure the pupil with the TCCD is the following 1 Obtain an image with the TCCD in the Pupil Check setup with the Secondary Mirror red LEDs switched on 2 Provide the image to the user together with information on the quantity of light available and the position of the spot 2 15 4 Image Check This section refers to the optical configuration described in the reference document 2 section 3 9 3 As a part of the alignment procedure VINCI s TCCD will be used to measure the alignment of the VLTI regarding the position of the image It will be done by reimaging a light source multimode
150. ometric calibration signals For simplicity these windows are refered to as pixels in this document The camera measures the flux on each pixel while the optical path difference is modulated by the fast scan mirror INA3 2 2 4 Mechanical System The optomechanics of LdV need a main table VINCI ALIU surface of 2 4x1 5 meters plus another optical table 1 8x0 9 meters large for the LEONARDO artificial star VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 8 2 3 LDV UNITS Some setup change is required to switch from one instrument mode of LdV to another In the following section both manually movable and locally controlled motorized parts are described The intensity setting of the fibered K band laser is preset manually in the laboratory by offsetting the attached fiber head The LdV division in units is Req 55 2 3 1 Combiner Unit 2 3 1 1 Manually Movable Devices m ucc type a COMA3 COMB3 ee baa E ALIQ p E E INA2 Injection B Flat Mirror Manual INB2 Fc INB3 Fe a OUT2 2 3 1 2 Computer Controled Devices Control Comments type LISA Filter Wheel 6 positions FILT Output Fiber Head Motor 3 translations 1 rotation OUT1 INA3 Hz INB INB1 INA1 Doc VLT SPE ESO 15810 1852 LEONARDO da VINCI Issue 1 1 Date 16 September 1999 Software User Requirements Page 9 Polarization Motors Motor Two rotating motors one for each
151. on This mode give the capability to send light in the whole VLTI optical system up to the telescopes This light is then retroreflected to VINCI and fringes are measured This requires to inject light from LEONARDO using the BSA2 and BSB2 beamsplitter cubes positions This mode will be available on line without requiring any manual operation in the laboratory No external access from the other instruments is foreseen The switching to the autocollimation mode requires to send an OPD offset value to the delay line Req 8 compared to the Stellar Interferometer or Autotest modes The value of this offset is VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 19 constant and depends on the geometry of the artificial star injection It will be measured during the integration of LdV It should be possible to change this offset in the software if necessary Unit Setup name Autocoll Autocoll Autocoll Autocoll Autocoll Idle Output Injection Fringe Data Element Adjust Adjust Search Acquisition id id om COMU INA1 MOTORS OFF OFF ADJ OFF OFF SE end oes E e bebe TIP TILT FOCUS ROTATION COMU POLA A POLA B BEES id pe x or hr EC ac A aa a fe E P mm ms ee cU eo source FRAME NOT SYNC SYNC SYNC 2 8 3 Stellar Interferometer This is the standard mode for observations with the test siderostats the ATs or the UTs The light from the star is directed into the be
152. on conditions as the nitrogen load is purely axisymetric A second test with a higher mass 5 kg positioned precisely at the center of the upper plate did not provoke a visually detectable motion of the spot on the screen As a conclusion the stability of the dewar mount seems satisfactory except for non axisymetric loads which should be avoided as much as possible A concern might be for example the presence or not of the vacuum pump For normal observations absolutely no lateral load should be put on the dewar and no cable should be moved in order to avoid any tilt of the dewar VLT TRE ESO 15810 2330 1 0 5 October 2000 15 LISA Test Report 5 OPTICAL AND BASIC DETECTOR TESTS 5 1 FIELD OF VIEW 5 1 1 Description The Hawaii array is illuminated by the cold doublet lens in front of the camera It is necessary to evaluate the field of view seen by the detector in order to position the fiber images in an unvignetted part of the field 5 1 2 Procedure First roughly evaluate the LISA detector field of view by putting a diffuse extended light source just in front of the camera window while running the full frame acquisition Evaluate the uniformity of the light on the dectector Make a focused image of the fiber head on the LISA detector see Section 4 2 Measure the peak intensity of the spot as it is moved over the detector by turning the knobs on the flat mirror of the OUT source The center of the field
153. one can expect that the fine reduction can be done during the observations themselves Req 33 The scenario would be OS ICS DCS Request obj pointing Check TCCD LISA etc Request DAQ Observation n Get small batch off source Get small batch beam A Get small batch beam B Get batch on source loop on Wait for DL in track Get one scan Quick look Transmit internal OPD value Quick data reduction Display scan vis Get batch off source Receive data from DCS Transfer data to OS Request obj pointing Check TCCD LISA etc Observation n 1 Request DAQ Pipeline observation n Store data in archive Display results obs n Receive data from DCS Wait for DAQ request Get small batch off source Get small batch beam A Get small batch beam B Get batch on source Get batch off source Transfer data to OS VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 44 If observation n 1 is terminated before observation n is fully pipelined then observation n is stored in a buffer to be pipelined at the end of the night Req 34 13 During the routine operation of LdV many targets 50 are observed during the night 14 LdV End of the Night after the last target has been observed the termination procedure is started This includes e saving the opened files saving the observing log e performing a final detector calibration e crea
154. ope of the Signal vs Noise curve It is expressed in ADU electron The measurements were done in full quadrant double correlated readout mode The light source used was simply the background light it is nearly saturating the detector see section 5 2 for discussion The intensity was varied by very slightly opening the cold shutter in front of the doublet lens Temporal series of a few tens of values were obtained for each light level The gain is given by the inverse square root of the slope of the linear fit shown on Figure 13 ig 2 9 electrons ADU 2000 1800 1600 1400 1200 Varitatee ADU o o eo eo c o o o eo 400 200 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 Mean Signal ADU Figure 13 Variance as a function of signal for LISA The slope is 0 1182 giving a gain factor of 2 9 electrons ADU X axis is the mean signal in ADU Y axis is the observed variance of the temporal sequence in ADU The overall distribution of the points on the curve may not be linear It seems that a plateau occurs at least from 4000 to 12000 ADUs This point will be checked more thoroughly in the 64x64 and windowed readout modes fast readout Please see Section 7 for a much more precise study 5 3 2 Readout noise in full quadrant mode 512x512 pixels The shutter being closed a time sequence of pixel values is acquired The graph presented on Figure 10 shows that the signal mean va
155. or the WS sw can send the starting command for not synchronized acquisition for example START NOTSYNC The possibility to use the synchronized mode for the injection optimization has to be checked depending on the gain in terms of software development 2 9 9 LISA Full Frame Readout FULL FRAME The positions of the fiber outputs on the HAWAII chip have to be very precisely known and adjusted if necessary The mount on which the fibers are mounted can drift with time or temperature and so require a realignment of the outputs on the desired pixels This is done by taking a full quadrant image 512x512 pixels x 16 bits from the LISA chip and fitting the pixel positions with centroids The position of maxima together with the FWHM of the pixels are the informations needed to make the necessary adjustements In a first implementation the output optimization will be done manually during daytime relying on the stability of the camera and fiber mounts to ensure the correct alignment of the fibers on the VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 33 LISA pixels The operator actions will be based on the maximum FWHM informations displayed in near real time by the system with a computed value of the percentage of light in a single pixel Req 26 The full frame image should also be available on the display Req 27 Checking of the output pixels should be available online during the obse
156. overed by the Cohen is typically 2 0 to 2 5 mas 0 03 to 0 04 mas The extreme values of the catalogue are 1 6 and 10 mas The following figure shows the precision on the interferometric transfer function estimate assuming a typical 2 3 0 035 mas calibrator bold curve and a very favorable 1 0 0 015 mas calibrator thin curve for different baseline lengths in the K band 2 2 microns Transfer function estimation maximum precision for two calibrator size Transfer function EB theoretical 2 3 0 035 mas precision 1 4 1 0 0 015 mas 1 2 1 0 8 0 6 0 4 0 2 0 0 20 40 60 80 100 120 140 160 180 200 Baseline length m ar Sens VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 21 of 22 On the brightest stars VINCI is able to reach a statistical precision of 0 002 96 This means that the final visibility estimation will be limited mainly by the calibrator size knowledge for baselines larger than a few tens of meters meters It is important to realize that it will not be possible to increase this precision by observing the calibrator for a longer time as it is a fundamental uncertainty not of statistical origin One should also mention that the error on the transfer function measurement impacts the final calibrated precision in a way that is not multiplicative but by division This means that usually if the observations on the target are very good and the calibrator
157. ptember 1999 Software User Requirements 30 2 9 6 Quick Look Fringe Detection Algorithm This section describes the algorithm which will be used to detect the fringe packet and then to compute the OPD offset to be sent to the delay line for the next scan 2 9 6 1 Construction of the Combined Interferometric Signal The two signals from the interferometric outputs are combined in order to increase the signal to noise ratio As the two signals are antiphased max light in one is min light in the other which means that they are substracted to obtain the combined signal The simple substraction of the two signals is not optimal because the intensities in the two channels are not exactly the same It is necessary to compute a coefficient o which minimizes the quantity Ip ab eventually to have the mean of equal zero The computation of o is done by minimizing the quadratic error on the value 2 9 6 2 Frequency Filtering In order to reduce the noise on the combined interferometric signal the second step of the algorithm is to filter the signal in the frequency space We know in advance precisely the range of frequencies covered by the fringe signal from the bandwidth of the K band or other filter used in LISA So it is necessary to make a fast Fourier transform of to cut the frequencies below Fmin and over Fmax and then take the inverse Fourier transform of the resulting signal The number of frames in the scan can be ad
158. r of each calibrator by comparing it to the other observed known stars This comparison can only be done in a limited segment of time as the transfer function of the whole VLTI VINCI System must not evolve between the measurements As the catalogue of the observed sources will grow so will the cross calibrated calibrators catalogue A first step could be to check the internal consistency of the whole southern sky Cohen catalogue 300 stars This would require to observe each star of the catalogue together with two other Cohen or other external reference stars the number three is chosen to stay in a short time segment Each set of three stars will provide the calibrated diameters for two of them while the third star can be used as a link to the rest of the catalogue The total number of single star observations would therefore be 3 stars x 2 observation per star x Nta 2 900 observations Assuming 80 observations per night one every 6 minutes this could be completed in 12 nights It will also be possible to extend the Cohen catalogue by observing stable stars down to smaller sizes and calibrating them with Cohen stars 8 2 KNOWLEDGE OF THE TARGET SPECTRUM SHAPE The spectral distribution of the energy received from the star can be far from beeing perfectly flat over the K band Depending on the spectral type of the star the mean slope of the star spectrum can be either nearly flat hot stars or rising steeply with wavelen
159. r to the relevant calibrators observation files is included in the header of the file This is the largest self consistent data set and thus it has to be stored in a single separated file Observation block a few star observation calibrators observation An observation block consists of interferograms obtained on a science star the astronomical object of interest on one hand and on a calibrator star on the other hand which is used as a reference to calibrate the science data The calibrator data is mandatory to produce scientifically significant visibility values from the science target raw data The calibrator star gives a reference for the evaluation of the transfer function of the instrument During an observation block a few observation pairs star calibrator are acquired to sample the transfer function variation dt acquisition 15 minutes to 1 hour The final estimation of the Doc VLT SPE ESO 15810 1852 LEONARDO da VINCI Issue 1 1 Date 16 September 1999 Software User Requirements Page 4 transfer function variations takes into account all the calibrators used during the night by linearly interpolating between the transfer function values Still each observed object is associated specifically with one or several calibrators to which references should be included in the saved file Observation Software OS it coordinates the activities of DCSs ICS and VLTI and interfaces with the VLT Data Flow System It runs only on
160. ram of pixel readout noises at 6 66 Hz frame frequency The median value is 20 18 e and the maximum number of pixels is found at 18 e There are still a number of pixels at less than 15 e The pixel 33 38 has a readout noise of 17 76 electrons There is an indication that the readout noise varied with the frequency being lower for higher frame rates see next section 7 3 4 Readout noise at the highest frame frequency A series of scans was acquired at the highest possible frequency for four 1x1 pixels windows with the cold shutter closed The readout noise from this data was found to be 12 24 electrons assuming a constant gain of 6 9652 e ADU Compared to the 17 76 electrons value obtained for the same pixel at 6 66 Hz frame rate it seems that the readout noise is significantly lower 3096 when acquiring at a higher frequency The reason for this behavior is 7 4 PIXEL NOISE POWER SPECTRAL DENSITY The second step in the photosites selection is to evaluate the color of the noise i e the absence of power peaks at certain frequencies that would betray the incorrect behavior of the detector noise VLT TRE ESO 15810 2330 1 0 5 October 2000 32 LISA Test Report 7 4 1 Low frequency 0 3 33 Hz 7 4 1 1 Data analysed For the low frequency evaluation the power spectral density PSD was computed for all the pixels up to the Nyquist frequency of 6 66 2 3 33 Hz This frequency is half of the in
161. rame frequency through Frequency Number of Frames in Scan Min DIT 6 2 EFFECTIVE MAXIMAL READOUT FREQUENCIES The maximal frequency reachable by the detector depends on the position of the four windows on the detector and on their size The smaller and the closer to the lower left corner the faster the readout The selected pixel windows for the windowed readout tests were centered for the 1x1 or 3x3 pixels or had their lower left corner for the 2x2 pixels windows on the pixels 29 26 25 34 33 38 37 30 The distance between these pixels was chosen to be 9 pixels to mimic as much as possible the behavior of the real spots readouts Table 5 Maximum readout frequencies for LISA Mode Maximal frequency Hz 4 windows 1x1 pixel each 2545 4 windows 2x2 pixels each 1656 4 windows 3x3 pixels each 1062 64x64 double correlated 28 6 64x64 uncorrelated 54 1 512x512 double correlated 0 48 512x512 uncorrelated 0 95 The maximum frequency reachable with the 1x1 pixel mode is better than 2 5 kHz therefore fulfilling the requirements stated in the LISA Technical Specifications document For information the pixel readout rate in 64x64 mode was 117 kHz This means that 4 windows of 1x1 pixel located in the lower left corner of the array 0 0 0 1 1 1 1 0 could theoretically be read at a maximum frame rate of 29 KHz VLT TRE ESO 15810 2330 1 0 5 October 2000 26 LISA Test Report
162. red as a reference for the data reduction and analysis The different fields to be stored will depend on the definition of the VLTI reference catalog of sources and calibrators which is TBD As a general rule it is important to store any information which influenced the choices the observer has made for his observations type of variability brightness spectral energy distribution This will be necessary in order to build a posteriori an image of what was the knowledge of the observer about the target at the time the object was observed 9999 9 Names as taken from Simbad for example HD SAO HR HIP Type target calibrator type of artificial source Calibrator quality rating from 1 best to 5 worse RA Dec taken from the Hipparcos catalog or the FK5 epoch J2000 0 Proper motion with uncertainties Parallax measurements Hipparcos Separated magnitude estimations in K min mean max V min mean max All available bands magnitudes U B V K L IRAS IUE min mean max All available color indexes B V J K with uncertainties All available angular diameter estimations theoretical models lunar occultations with publications references Computed angular diameters through different methods spectral type distance bolometric flux with uncertainties in milliarcsec and in radians Variability flag Binarity or multiplicity flag Variability type if applic
163. relate the command voltage to the OPD generated The gain factor g takes into account geometric considerations sign conventions and the stiffness of the piezo support blades Its absolute value is typically 1 5 but it has to be calibrated during integration VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 26 Piezo command voltage Uc 5 V or 0 V 5 V or 10 V Piezo input voltage Uin 0 1000 V Piezo mechanical extension 0 micron 180 microns OPD generated 0 micron 180 x g microns The voltage ramp that generates the scan is of smooth triangle type The OPD modulation rate v is called the OPD velocity or fringe velocity The data are collected only when v is stabilized i e during the linear part of the ramp Fringe velocity range from 0 to 2500 microns s For each observation the user requests via the OS a given velocity and DAQ OPD typical values are 660 microns s and 200 microns respectively Depending on those parameters the total duration of data collection in a scan Atpag OPDpagy can range between 0 05 s and several minutes We shall adopt 0 1 s as a nominal and most common value The user also choses a sample interval i e the OPD interval between two successive frames This is translated by the OS into a frame rate The combination of DAQ OPD length fringe velocity and frame rate determines the number of frames recorded per scan After acquisition
164. related data should be accessible from outside of LdV software Req 9 LdV instrument is exclusively assigned to the instrument requesting the image check Only the ALIU and ARTU LEONARDO units are concerned but no other operation can be conducted simultaneously The images and other data position of the center of the pupil obtained in this mode by an external user instrument should be made available to this instrument No real time constraint is foreseen in this mode Image Check image Check Image Check Idle Telescope A e escope B VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 23 COMU INB1 MOTORS N A N A N A TIP TILT FOCUS COMU INA1 MOTORS N A N A N A TIP TILT FOCUS COMU _ INA3 FAST SCAN N A N A N A PIEZO COMU OUT1 MOTORS N A A A TIP TILT FOCUS ROTATION COMU POLA A POLA B N A MOTORS COMU LISA Filter Wheel N A ALIU ALI SLIDE POSITION NOMIRROR ALI3 ALI4 ALIU TCCD OUT OUT OUT LENS POSITION ALIU TCCD FOCUS N A ADJ ADJ PRESET1 PRESET1 ARTU BSA POSITION OUT P TIERE ARTU BSB POSITION OUT OUT OUT ARTU LEONARDO light N A N A N A source 2 8 6 Artificial Star In this setup LEONARDO can be operated without the main VINCI optical table Req 10 As a general rule LEONARDO should be considered as an integral part of the VLTI infrastructure The user instrument requests the exclusive assignment of LEONARDO in order to avoid co
165. ric transmission I 0 722 Interferometric efficiency end Siderostat with BC if telescope 1 A 0 084 Area of the telescope m2 T 0 00582 Photometric transmission I 0 722 Interferometric efficiency end AT if telescope 2 A 2 545 Area of the telescope m2 T 0 00173 Photometric transmission I 0 722 Interferometric efficiency end UT without AO if telescope 3 A 50 265 Area of the telescope m2 T 0 0000914 Photometric transmission I 0 722 Interferometric efficiency end UT with AO if telescope 4 A 50 265 Area of the telescope m2 T 0 00342 Photometric transmission I 0 722 Interferometric efficiency end ct These parameters are not lescope dependant lambda 2 2e 6 Effective central wavelength m delta_lambda 0 4e 6 Wavelength range m m magnitude K magnitude of the star FO 3 9e 10 K magnitude 0 flux W m2 MICRON 1 51 2 2 2e 6 Effective total scan length m l fringes 70e 6 Fringe packet length oe oe oe oe oe n 57 Number of samples per fringe Temp 288 Temperature K E 1 Beam etendue epsilon 1 Emissivity QE 0 62 Quantum efficiency of the detector RON 10 Read out noise of the detector e h 6 6226e 34 Planck constant J s k 1 38062e 23 Boltzmann constant J s oe c 3e8 sigma min integ 3e5 sigma max integ 6e5 Spee
166. roved 4 Glindemann Name Date Signature Released M Tarenghi Oe a ea Name Date Signature VLT PROGRAMME TELEPHONE 089 3 20 06 0 FAX 089 3 20 23 62 VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements i CHANGE RECORD Section Page affected Reason Initiation Remarks EM 6 July 1999 uses release v6 0 draft E August 1999 List of numbered reqs 2 19 leslie MN corrections after the GUI display parameters 2 18 LdV FDR Templates overview 2 17 16 September Maintenance and engineering Includes comments from J P modes precised 2 7 2 18 Dupin and A Longinotti Some numbered reqs added VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements ii TABLE OF CONTENTS 1 INTRODUCTION 1 1 1 2 1 3 1 4 1 5 Scope Applicable Documents Reference Documents Abbreviations Acronyms and Typographic Conventions Glossary 2 LEONARDO DA VINCI SOFTWARE USER REQUIREMENTS 2 1 Instrument Concept 2 2 LdV System Overview 2 2 1 Interferometry Room 2 2 2 Optical System 2 2 3 LISA Camera 2 2 4 Mechanical System 2 3 LdV Units 2 3 1 Combiner Unit 2 3 1 1 Manually Movable Devices 2 3 1 2 Computer Controled Devices 2 32 Alignment Toolkit Unit 2 3 2 1 Manually Movable Devices 2 3 22 Computer Controled Devices 2 33 Artificial Star Unit 2 3 3 1 Manually Movable Devices 2 3 3 2 Computer Controled Devices 2 3 4 Infrared C
167. rvations Req 28 Eventually the output optimization could be done before every observation to maximize the effectiveness of the instrument but this requires that the output optimization process be automatic The output optimization automatization is potentially difficult for the control software Though as it is foreseen to eventually achieve the complete automatization of the LdV operations this possibility should be left open in the software design The readout of the full LISA image should be possible at maximum rate Req 29 as a bright artificial source will be used during this process Though the command required to start the full frame acquisition should take into account an exposure time setting in order to accommodate for different artificial light source intensities Parameter Range Values Full Frame 1 1000 milliseconds Exposure Time The command to start the exposures automatically one after the other should be available to the user and the WS sw The computed parameters maxima positions FWHM energy in one pixel should be displayed to the operator Before each observation the last full frame image should be saved for reference Req 30 This will be necessary at least during the commissioning phase but this could also be useful afterwards and in any case after every automatic optimization when it is implemented 2 9 10 Engineering Mode Data The data in the Engineering mode is simpler than
168. s Precisions all telescopes plot table siderostats without bc 1 100 1 table siderostats without bc 1 100 6 table siderostats 1 100 1 table siderostats 1 100 6 table ats 1 table ats 6 table uts witho ut ao 1 table uts without ao 6 table uts with ao 1 table uts with ao 6 Noise sources sids dp plot table siderostats 1 table siderostats 7 table siderostats 1 table siderostats 8 ta ble siderostats 1 table siderostats 9 11 3 SNR COMPUTATION SNR VINCI function snr snr vinci scan speed magnitude telescope This function computes the signal to noise ratio as a function of the VINCI observations for a given detector frequency version 17 05 00 dp oe oe These parameters change with the telescope Siderostat without Beam Compressors if telescope 0 dp oe A 0 084 Area of the telescope m2 T 0 00612 Photometric transmission I 0 722 Interferometric efficiency end Siderostat with BC if telescope 1 A 0 084 Area of the telescope m2 T 0 00582 Photometric transmission I 0 722 Interferometric efficiency end AT if telescope 2 A 2 545 Area of the telescope m2 T 0 00173 Photometric transmission I 0 722 Interferometric efficiency end UT without AO if telescope 3 A 50 265 Area of the telescope m2 T 0 0000914 Photometric transmission I 0 722 Interferometric efficiency end UT with AO
169. s is the time interval after the end of the scan during which the piezo mirror is stopped to wait for the OPD offset computation and correction to finish After that the piezo goes backwards to start the next scan in the other direction This means that the software system should compute the fringe position quickly enough quick look algorithm and correct the delay line position so that it is possible not to waste any time by stopping the piezo at its extremal position to wait for the computation to end This puts a constraint on the computation time topp otiset Which iS topp ottset lt scan end tpag eng IN any case the dead time ti tscan eng should be minimized in order to obtain the maximum efficiency of LdV but the delay line has to be in the track status before starting the next data acquisition VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 29 D Piezo rine position OPD offset computation and correction Data Acquisition E Time lscan toaa toaa lscan start start end end The time critical operations are Transfer of scan data from LISA DCS to LdV OS Quick look analysis at the LISA WS level Transfer of OPD ofrser value from LISA WS to the delay line through the DLCS Interface module Transfer of timing information for the next scan from LdV OS to VINCI ICS and LISA DCS Meanwhile the timing information tscan star
170. scope off source Refraction parameters air mass refraction correction applied by the VLTI beginning and end of each observation 9 9 Available from the telescopes or observatory systems VLTI setup baseline used telescope types UT names Name of the delayed telescope RA Dec beginning and end of each observation Altitude Azimuth beginning and end of each observation Mean tracking error RA and Dec over the observation Environmental parameters seeing isoplanetic angle correlation time air pressure air temperature ground temperature humidity wind velocity and direction seismic activity seismic flag 2 11 2 4 Data Format After a meeting with representatives of the AMBER and MIDI teams the two science instruments of the VLTI and ESO Data Management Division it has been agreed that the FITS binary table format was the most suitable for the storage of the data The exact internal formatting and keywords of these files are currently under definition The header of the FITS file contains all the parameters which are defined less often than every scan i e night observation block observation The interferometric data itself is stored in the body of the FITS file The TCCD images are stored in the body of the data file before or after the interferometric data The following figure shows the typical data acquisition during an observing night and what is stored in the data files
171. se set plot X end 4 LdV Precision and Sensitivity 10 EUROPEAN SOUTHERN OBSERVATORY Organisation Europ enne pour des Recherches Astronomiques dans l H misph re Austral Europ ische Organisation f r astronomische Forschung in der s dlichen Hemisph re VERY LARGE TELESCOPE LEONARDO da VINCI LdV Precision and Sensitivity Doc No VLT TRE ESO 15810 2177 Issue 1 0 Date 12 July 2000 Prepared P Kervella eee Name Date Signature Approved A Glindemann Name Date Signature Released M Tarenghi eee Name Date Signature VLT PROGRAMME TELEPHONE 089 3 20 06 0 FAX 089 3 2023 62 os EE VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 1 of 2 CHANGE RECORD Issue Date Section Page affected Reason Initiation Remarks Error Reference All first issue Source not found VLT TRE ESO 15810 2177 LdV Precision and Sensitivity o 10 12 July 2000 2 of 3 TABLE OF CONTENTS Table of Contents 1 INTRODUCTION EN 4 1 1 SCOPE t4 ee LM IM ER Melen 4 1 2 REFERENCE DOCUMENTS ener ens enres terne n tese ttes et tes eene tense sense ae estes e rr enean 4 2 PRECISION AND SENSITIVITY LIMITING FACTORS 0 ccsscssscssccssccssccssccssccssccesscesscesssesscessees 5 Ba OPTICS ES 6 3 1 AREA OF THE COLLECTORS OPTICS trennen nennen nnne eene neret inns ee nns eene trn tenes eere eene ne 6 3 2 TRANSMISSION OF THE OPTICAL TRAIN csssccsssssscsessecc
172. se is introduced on the detector by the infrared light emmited by the instrument The temperature taken into account is the of the warm optics We adopt the mean value of the temperature in the laboratory of 15 5 C 4 3 2 Emissivity The worst possible case is an emissivity of 1 This means that the instrument is assumed to be a perfect blackbody The thermal noise is negligible in the K band at the temperature considered here 15 5 Celsius degrees and therefore the influence of this parameter is also negligible 4 3 3 Beam etendue The beam etendue is the product of the angle seen by the detector by its illuminated surface It is 1 when the detector is seeing all the thermal light emitted by the instrument this is not the case in reality thanks to the cold stop in the detector dewar For the same reason as for the emissivity low thermal noise this factor has a negligible influence on the results As it is in addition relatively difficult to model we take into account the worst possible case i e a beam etendue of 1 ar Sens VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 10 of 11 4 4 SAMPLING LOSSES The HAWAII detector is used as a non blocking sampler The maximum efficiency in the transmission of the fringe contrast is reached for a sampling of 5 points per fringe corresponding to an efficiency of 8096 4 5 DETECTOR DEFECTS The HAWAII array of VINCI is an engineering grade detector This means that
173. second precision goal 0 5 ms minimum 1 ms Req 16 The synchronization does not have to be done at the frame level but only at the beginning of the scan This scheme assumes that the internal clockings of the two LCUs piezo LCU and LISA WS are accurate enough to guarantee no significant deterioration of the synchronization during the scan duration 2 9 2 Terminology and typical values There are four fiber outputs in LdV named 11 12 P1 P2 The four fibers are arranged through a fiber bundle in a square which forms the optical output imaged on the LISA focal plane array Ideally each fiber core is imaged onto a single pixel In case the light cannot be put on a single pixel it will have to be collected in a window of 1 to 25 pixels Req 50 that might not be adjacent The LISA DCS reads out only those pixels located in the four windows It then returns four numbers which correspond to the total energy output of 11 12 P1 and P2 respectively This operation readout and sommation is called a frame Frame rates range between a few Hz and a few kHz a typical value is 1500 Hz A scan is a collection of frames While observing the astronomical source they are obtained while the optical path difference OPD is modulated by the fast scan mirror INA3 There are real time issues related to the acquisition of successive scans which are detailed below A batch is the collection of a number of scans from about a hundred to a thousan
174. ser instruments should access LEONARDO only through these predefined modes Req 12 No low level command is foreseen to be accessible from outside of LdV software VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 45 The thermal light source can be turned off during normal operations but if it is required to switch it on again the preheating will take a few minutes in order to achieve a good stability Req 53 This does not apply to laser sources The commands issued by the instruments should follow the same format as the commands for the other systems of the VLTI telescopes delay lines Generally speaking LEONARDO should be integrated in the VLTI infrastructure as much as possible and follow the same control procedure Req 35 This means that it will act independently from VINCI and the alignment toolkit ALIU The user instrument should request the allocation of LEONARDO before sending commands to it in order to avoid conflicts with the other instruments By default LEONARDO should be assigned to LdV LEONARDO should be automatically allocated to the master observing instrument so as to avoid the accidental insertion of the LEONARDO cubes in the beams 2 12 3 Alignment Toolkit Interface with the VLTI instruments The VLTI instruments AMBER and MIDI as well as the alignment software of the VLTI should have access to the Pupil Check and Image Check modes and setups Req 9 All th
175. sesseeesessecesessecesesaesceessesecesaeeeseseeceesaeseeeaeeeseseeens 6 3 2 1 PROTON CEs EE 6 3 2 2 Vibrations and polarization contrast losses Rs 6 3 2 3 Static wavefront distortion due to the optical train 7 3 2 4 EE 7 3 3 FIBER INJECTION STATIC LOSSES lt se cscscccsscssescssestesvescsssasavssseusavestcsveseussasgucsbensaveshsostsesbiasesesdessaveshevstiosveds 7 3 4 FRESNEL ge EE H 3 5 FIBER INJECTION DYNAMIC LOSSES sise EE EEA EAEE y AE EEE AEE EEEE AREAN 8 3 6 EFFICIENCY OF THE TRIPLE COUPLER EE 8 3 7 IMAGING OF THE FIBER HEADS ON THE DETECTOR een emen eterne eren nenne trennen trennen 8 4 DETECTION Sec 9 4 1 DETECTOR QUANTUM EFFICIENCY ecce etta Ere des E Ea ve Wein dass d eve RE NE Wain tees Seen sen ttes 9 4 2 READOUT NOISE creme retail tues tues rr de ve e o e eb e e ib t e ean re teet 9 4 3 THERMAL BACKGROUND NOIS Ees inent eneeier tian tenn teas e Ee ben eire e E D E eR 9 4 3 1 y TATE 9 4 3 2 E eee Ness uae tU e m e LM i loo cd a ee De cut Oe 9 4 3 3 RE 9 4 4 SAMPLING LOSSES 1 17 25 20205 EEN 10 4 5 DETECTOR EE KE 10 5 1 ATMOSPHERE E 11 5 1 PISTON NOIS E EE 11 5 1 1 Theoretical aspects ettet re ti trs ER lee e eere denen 11 5 1 2 FLUOR IOTA experimental results 12 5 1 3 Fringe Sensor Unt iei Re oe SR EEN 13 5 2 ND OI WEE 13 5 3 PHOTOMETRIC NOISE EE 13 5 4 INTERNAL TURBULENCE oer enee EE e n e o RR E d 13 5 5 ATMOSPHERIC TRANSMISSION IN THE K BAND 14 5 6 TRA
176. sing light level by opening slightly the cold shutter Cubes of 500 frames of 64x64 pixels were recorded 150 ms exposure time of which 256 were processed adequate number for fast Fourier transform computation This test was conducted at seven different light levels of which six were kept for processing the seventh contained some saturated pixels 20 total T 2 2 O photons T O Ron Oo Sanu ES Se Sapu Ke RON G The gain was estimated as the inverse of the slope of the linear regression of the Variance f Signal curve All the processing was done using IDL The listing of the procedures is included in the Appendix 7 2 2 Examples of gain curves The following curves show the fits achieved for three typical pixels VLT TRE ESO 15810 2330 1 0 5 October 2000 27 LISA Test Report Gain plot for pixel 31 38 2000 1500 Variance 1000 50G E ET EEE ET ET TEE eE T3 CARTE DS pap 043 Ep bp e Eo b d 3a GC o 2 0x10 40x10 6 0x107 8 0x lt 10 1 0x10 1 2x104 14104 Mean ADU Figure 15 Gain plot for pixel 31 38 G 5 1096 Gain plot for pixel 32 38 2000 LE S E A O O E E E E E EE 1500 1000 Figure 16 Gain plot for pixel 32 38 G 6 8939 VLT TRE ESO 15810 2330 1 0 5 October 2000 28 LISA Test Report 2000 1500 1000 Figure 17 Gain plot for pixel 33 38 G 6 9652 7 2 3 Gain map The gains are available numerically
177. surement The mean AFocus is 0 275 mm standard deviation of 0 053 mm but the focus point is different for different parts of the detector The reason for this could be a curved focal plane for the doublet or a bad parallelism between the detector and the doublet or even a non planeity of the detector surface due possibly to decay of the indium layer observed near the edges A plot of the evolution of the focus position as a function of the distance to the center of the detector lower left corner of the used quadrant is presented on the following figure X axis is the radius in pixels 18 microns from the center of the detector lower left corner of the quadrant used Y axis is the fiber displacement in millimeters VLT TRE ESO 15810 2330 1 0 5 October 2000 10 LISA Test Report Delta focus 0 400 0 350 0 300 0 250 0 200 0 150 0 100 0 050 0 000 42 0 43 0 44 0 45 0 46 0 47 0 48 0 49 0 50 0 51 0 Delta millimeters 0 01408 R pixels of 18 microns 0 924 Using this formula it is possible to compute the focus correction to give to the doublet or to the detector depending on the part of the detector that will be used for the actual observations Though it is necessary to evaluate the usable area of the detector to determine the best focus to apply Intensity 5000 1 o 6000 7000 8000 Intensity 9000 10000 11000 12000 Figur
178. t The alignment of the VLTI is supposed to be good enough so as to bring the stellar light up to the VINCI table and on the TCCD chip The following procedure is foreseen Req 39 1 The star image is acquired The exposure time is computed based on the brightness of the source in the optical wavelengths It is taken from the observing template or the online catalogs database Measurement of the source s the resulting image shows the source and possibly other objects with a calibrated intensity in ADUs The sources are then measured by fitting two dimensional gaussian curves to the intensity profiles present in the image This fit could be done via a fast Fourier transform of the image or other numerical method The resulting data is a set of coordinates intensities and profile widths This information and the TCCD image will be saved in the observations data Identification of the target in the observing template a flag should indicate if a close bright source is expected to be in the same field In case this flag is up the identification could be done based on the expected relative brightness of the target compared to the other object s or visually by the operator of the instrument In case this flag is down the identification is normally unambiguous though a very slight risk of confusion exists with an unexpected source other very close star asteroid This risk should not be taken into account for the design of th
179. t the status of LdV and the VLTI including error and warning messages For example the position of the delay line the pointing direction of the telescopes the status of the movable mirrors the readings of the encoders and the status of the light switches e Provide the observer with software control over every computerized element relevant for the Stellar Interferometer Mode such as the movable mirrors and white light sources VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 54 Read and modify the observing list to allow for real time tuning of the observing schedule The observing conditions can influence the selection of the sources for example the faint sources will not be observable during bad turbulence periods Perform basic observational data pre processing such as averaging multiple scans values The raw data shall be kept safe during these steps and saved together with pre processed data in the archive Real time display the observational data This is the key point to allow for a real time appreciation of the quality of the data and to check for possible malfunctions The visibilities could be displayed in chronological order as well as in the form of an histogram Allow the user to perform near line visibility data processing To check for unexpected variability in the targets visibilities and possibly trigger extended observation on the most interesting objects Access t
180. the Instrument Workstation This implies that it cannot deal with any real time issue for which DCS and or ICS must be responsible Optical Path Difference OPD this is the difference in the physical length traveled by the stellar light between one arm of the interferometer i e from one telescope and the other Scan a few hundred frames The interferogram itself covering a few hundred microns OPD with thousand frames sampling 5 pixels fringe It shows fringes on the interferometric channels and the photometric variations dt acquisition 0 1 to 1 second Doc VLT SPE ESO 15810 1852 LEONARDO da VINCI Issue 1 1 Date 16 September 1999 Software User Requirements Page 5 2 LEONARDO DA VINCI SoFTWARE USER REQUIREMENTS 2 1 INSTRUMENT CONCEPT LdV is at the same time an interferometric beam combiner designed to coherently add the light coming from two telescopes either test siderostats auxiliary telescopes or unit telescopes a reference source system for the VLTI and an alignment toolkit The key element of the instrument is the fibered triple coupler MONA which uses single mode in the K band fluoride glass fibers to guide and mix the stellar light coming from the two telescopes It will provide four signals two interferometric outputs and two photometric calibration signals The interferometric outputs carry the scientific information the fringes visibility while the calibration signals are used to compensate for t
181. the LISA camera This will set the reference for the alignment of the optical axis of the autocollimator Be careful to switch off the laser after use Install the autocollimator and adjust its height precisely For that it is necessary to have a correct horizontality of the autocollimator optical axis There is an integrated bubble level for that purpose The height of the center of the input lens should be set at 320 mm above the table surface Position the axis of the autocollimator between the OUT2 mirror and the LISA camera window mm precision This part is not very easy as there is no reference to set the direction of the refractor One possible method is to use the laser of the OUT source send the laser light into the OUT source check that the laser dot is at the center of the entrance lens of the autocollimator but do not look into the eyepiece be careful to switch the laser off and plug in the thermal light source do a first rough autocolimation of the refractor by using a flat mirror in front of the entrance lens The goal here is to focus the image of the cross while looking into the eyepiece move the back of the autocollimator until you can see the fiber head near the center of the autocollimator Do a second autocollimation to focus precisely the autocollimator to infinity Move the OUT2 mirror to center the fiber image on the cross of the autocollimator Adjust finely the fiber position to have a focus
182. the formula f n w A This means that the readout frequency of the detector and therefore the exposure time is completely defined by the fringe speed As the star is brighter the exposure time can be shorter and therefore the piston noise impact on the fringes can be reduced It is important though to have a sufficient number of photons in each OPD bin so as not to be limited by the detector readout noise but still not too large to avoid the photon shot noise The determination of the optimal exposure time is therefore a trade off between the piston noise the photon noise and the detector noise in order to minimize the total noise on the flux measurement For the faint stars the problem is the same but the photon shot noise will be negligible compared to the readout noise of the detector and the piston noise The formulae for the different noise sources are given in the Appendix Section 10 The minimization is done on the total signal to noise ratio Q as given in Section 7 1 assuming that only the OPD scan speed is variable The piston has an effect on ly on the fringes part of the acquired interferograms This means that the time used to compute the fringes blurring is the the time necessary to scan the 70 microns of the fringe packet see Section 5 1 for details 7 3 CURVES WITH THE DIFFERENT LIGHT COLLECTORS WITHOUT FSU For 100 scans and combining the two interferometric outputs From left to right m Siderostats without beam co
183. the possibility exists that the detector presents some discrepancies from the mentioned characteristics either in readout noise frequency response or other parameters No defect has been taken into account here DE ST VLT TRE ESO 15810 2177 LdV Precision and Sensitivity 10 12 July 2000 11 of 12 b ATMOSPHERE 5 1 PISTON NOISE 5 1 1 Theoretical aspects Fringe motion caused by the atmosphere is the most important precision limiting factor for VINCI The differential piston between the two beams blurs the fringes and therefore affects their visibility Figure 1 shows the instrumental visibility as a function of the standard deviation of the piston over one scan duration see RD8 for details For small phase difference fluctuation o T 2 Vs Ss e with T the duration of the scanning of the fringes and o T the standard deviation of the piston in radians Vinst decreases from 100 when no piston is present to 10 for a standard deviation of 4 3 at 2 2 microns In the following the uncertainty on the visibility estimate is assumed to be the same as the instrumental visibility degradation A qualitative reason for this is that the energy in the interferometric fringes can be affected randomly both positively and negatively by the atmospheric OPD variations The visibility degradation computed here corresponds to the typical negative energy case and therefore can be taken as a standard deviationvalue For further analysis o
184. timization data should be stored in the observation of the star The same comments apply to the LISA full frame images which will help during the data reduction process to track misalignments of the fiber outputs They should be saved before each observation and in any case after every output alignment procedure 2 16 INJECTION AND OUTPUT OPTIMIZATION PROCEDURES 2 16 1 Refined Injection Optimization The principle for this optimization is that we are looking for the very precise position of the stellar image peak intensity The stellar Airy disk size is 0 065 arcsec in FWHM for UT and 0 290 arcsec for AT and the fiber head projected size on the sky is the same but the speckle cloud which replaces the Airy disk without adaptive optics is about 0 5 to 1 arcsec in size for both The star image and the fiber head transmissivity have gaussian profiles VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 50 The precision required for the superposition of the two gaussian curves is of the order of 7 milliarcsec for the UTs It will not be possible to center the images using the TCCD because the star detection precision is not sufficient This is a critical point because all the light which is not injected in the fiber is lost As the automation of this procedure is very demanding for the software system the operator should manually search for the optimum injection in the first phase of the operatio
185. ting a summary of the observations of the night a summary of the measurements obtained during the observing session is displayed to the observer and possibly printed or sent by e mail to the interested users The final environmental parameters could also be added to the saved data A summary of all the technical problems and important error messages from the operating system could also be displayed and saved to disk e switching off the instrument the status of LdVis changed to off 15 Reduction Pipeline the automated data reduction process is started after the end of the night It reduces all the data obtained during the night which is not already reduced and saves the resulting files to the LdV night data set 16 Off line Data Analysis using standard MIDAS IDL or specific data analysis tools the astronomer processes and combines the data 2 12 2 LEONARDO Interface with the VLTI Instruments The VLTI scientific instruments AMBER and MIDI should have access to LEONARDO artificial star mode Req 11 All the setups of the Artificial Star mode see desctiptions section 2 8 6 should be available to the VLTI instruments through the VLTI CS Artificial Star Off Removed Artificial Star On Inserted Artificial Star On Removed Artificial Star Off Inserted A Autocollimation On Inserted The switching between these setups will require a few seconds in order for the motors to achieve the necessary motions The u
186. ts 21 No real time constraint is foreseen in this mode The pupil check is done by imaging the pupil of the two telescopes on the TCCD During normal operation the pupil is situated on the VINCI table on the parabolae INA1 and INB1 After this image has been saved the image of the artificial star provided by LEONARDO star image at infinity is obtained By comparing the positions of the pupil and artifial star the user can evaluate the quality of the pupil alignment The LdV TCCD will be used in this mode during the alignment of the pupil of the VLTI but the control of the mirrors of the optical train is not part of LdV SW but of VLTI software It will only provide the coordinates of the pupil and artificial source images to the VLTI alignment software which will adjust the relevant mirrors The corresponding interface is TBD The TCCD can image the pupil from 1 meter to infinity It is necessary to insert an additional lens TLENS achromat lens focal length 2000 mm in front of the TCCD in order to focus to pupil distances to the TCCD refractor lens center of less than 3900 mm The focal length of the TCCD refractor is 300 mm The pupil longitudinal positioning question is adressed in the document 2 section 4 2 4 The detailed description of the optical design for pupil check can be found in the document 3 p 14 The size of the pupil image on the TCCD will be 6 mm for INA magnification 0 3 and 8 mm for INB magnification 0 4
187. ts on the LISA detector requires to monitor precisely their FWHM on LISA full frame images and then to apply the necessary focus motion on the fiber bundle head OUT1 To set the rotation of the fiber bundle it is possible to monitor the light concentrated in the four pixels corresponding to the light maxima and to compare it with the flux in the surrounding pixels The target here is to have as much of the incoming light concentrated on a single pixel of the detector This could result is positioning the four output pixels on four different lines and columns i e on a square with non horizontal and vertical sides 2 20 3 Spectral Dispersion The optical design of LdV makes it possible to disperse the light from the four fiber outputs before it enters the LISA camera This feature is very interesting because it would greatly extend the capabilities of LdV in the field of stellar and extrasolar planets observations It is thus important to keep in mind that the LdV software has to be able to accommodate for four lines readout of LISA instead of four pixels The dispersion would cover about 50 pixels This would mean for the LdV software a proportional increase in the quantity of raw data produced The LdV software should be able to read 200 50x4 individual windows on the HAWAII detector The fiber output alignment procedure would also be affected in this configuration of the instrument 2 20 4 Sensors VLT SPE ESO 15810 1852 LEONARDO da VIN
188. ts tpao sta should be determined within 0 5 ms 1 ms across LdV OS VINCI ICS and LISA DCS Req 16 2 9 5 Delay Line Control In order to follow the motion of the fringe packet induced by the atmospheric piston effect it is necessary to send an offset to the position of the delay line This is done after each scan in which fringes have been detected Req 22 The OPD offset is transferred to the delay line control system through the OPD controller The OPD offsets shall not be sent to the delay line when the Fringe Sensor Unit FSU is used Req 62 A switch should allow the user to disable the OPD corrections by LdV The total time between the start of the quick look analysis and the end of the delay line corrective motion LdV informed that the DL is again in track mode should be less than the slow down time of the piezo t ena toag ena Req 23 This time lapse depends on the shape factor applied to the piezo command signal It is the shortest when the piezo wave shape is set to its most triangular value This setting should set the constraint on the OPD offset correction time Though not all the scan frequencies can be used with all the wave shapes generally speaking the faster the scan the more sinusoidal the wave shape to reduce hysteresis The precise minimum time interval available is thus variable depending on the frequency and wave shape 5 10 ms range VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 Se
189. uivalent to piston noise Oniston Piston noise equivalent power W Hz ous 5 37 Another expression is the following see Section 5 1 for details Scan length um n Number of samples per fringe Effective central wavelength m LL Acquisition frequency Hz Viringes Fringe speed m s piston Piston noise equivalent power W Hz fringes n 2 ES Hn P fag 1 e ge piston VLT TRE ESO 15810 2177 1 0 12 July 2000 30 of 31 LdV Precision and Sensitivity 10 5 DETECTOR NOISE La Acquisition frequency Hz fringes Fringe frequency Hz n Number of samples per fringe Read out noise e n Quantum efficiency Deetector Detector noise equivalent power W Hz 2 2 4 9 hc Do 7 E ES N pixels ha At AF 10 6 PHOTON NOISE A Effective central wavelength m AA Wavelength range um A Effective single telescope collecting area m T Photometric transmission of the optical train m K magnitude of the star E Reference flux for mag 0 3 9 10 W m um Dphoton Photon noise equivalent power W Hz 2 2hc X son EE AAT F 1075 J ALAF 10 7 UNCERTAINTY ON THE VISIBILITY By combining the visibilities of the two interferometric channels and assuming that they are not correlated the resulting uncertainty 9 on the visibility for one scan is for N interferograms with
190. und increase the internal OPD by a fixed ammount X entered by the observer can be positive or negative once the fringes are found by the quicklook algorithm the delay line motion speed is set to the sidereal rate in order to have the fringes at a fixed position The fringes are then repeatedly measured and the scans are stored The duration of the observation is selected by the observer as a number of good scans for example At the end of each fringe scan while on source a quick look analysis is performed which results in an updated value for the internal OPD to be communicated to the delay line 12 0n line Data Quality Control the LdV operating system provides the user with an on line display of the data acquired Req 18 This enables the observer to check the quality of the visibility measurements and gives an estimation of the atmospheric perturbations evolution during the night The display of these values can be done with a delay of 0 5 s 1 s and or by packets several scans displayed simultaneously to reduce the real time constraints on the system The basic data reduction should be done by the DCS quick reduction in real time on the LISA WS but displayed every 0 5 s 1 s to the user and the fine data reduction by the OS in a near line regime after each observation This implies that the fine data reduction algorithm should be resident on the LdV WS Since the LdV WS will be essentially idle during an observation
191. ure 10 11 VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 42 of calibrators should be apparent in the list The database is a key element of the VLTI SW The definition of this database is a point to discuss with the other instruments and the VLTI The observation scheduling SW tools are not part of the VINCI SW but rather part of the general VLTI SW since they will be used by all the instruments LdV Initialization the instrument status is changed to online Short term Observations Scheduling in order to fine tune the observing program of the instrument the observer has access to the observing list and can modify it by hand The possibility remains of having an automated system to optimize the observing schedule for the night Telescopes Pointing position the telescopes at the target coordinates Telescopes Focusing focus the telescopes in order to feed a collimated beam in the Coude train Delay Line positioning the delay line is sent to the position required for optical path length equalization between the two beams and it is moved at the computed rate for the observed target Technical CCD Images Acquisitions images of the target are obtained by the technical CCD on both beams to check for the positioning of the pupil and the image quality This requires the motion of the LdV mirrors ALI2 ALI3 and ALI4 ALI6 Positioning of the fiber input this is done b
192. verse of a single 64x64 frame exposure time The power spectral density was computed simply as the squared modulus of the fast Fourier transform along the time sequence axis for each pixel in the temporal data cube 64x64x256 7 4 1 2 Examples of individual pixels PSD Power Spectral Density for pixel 31 38 T T T T T 5 0104 T T T T 4 0x10 3 0x10 Power Spectral Density 2 0x10 LLJ L 1 L L 0 50 100 1 Relative Frequency Maximum pix 128 3 33 Hz 0 026 Hz bin m a Figure 22 Power spectral density for pixel 31 38 from 0 to 3 33 Hz X 128 Y scale arbitrary VLT TRE ESO 15810 2330 1 0 5 October 2000 33 LISA Test Report Power Spectral Density for pixel 32 38 T T T T T T 5 0 10 T T T T T soxot 3 0x10 Power Spectral Density S o E T 1 0x10 LLLLLLLLLLLELLLELEE DELE E LE EE LE EE LEE EE EE LEE EEL LLLI il AN 1 0 50 Relative Frequency Maximum pix 128 3 33 Hz 0 026 Hz bin m a Figure 23 PSD for pixel 32 38 from 0 to 3 33 Hz X 128 Y scale arbitrary Power Spectral Density for pixel 33 38 5oxto T T T T T T T T T T T soxot 3 0x10 Power Spectral Density 2 0x10 1 0x10 LLLLLLLLLLLLLLLEEEE ELE E LEEE EE EE LEE EE D 50 Relative Frequency Maximum pix 128 3 33 Hz 0 026 Hz bin m a Figure 24 PSD for pixel 33 38 from 0 to 3 33 Hz X 128 Y scale arbitrary 7
193. version 12 09 2000 Power Spectral Density set plot X set plot PS device inches xsize 7 0 ysize 7 0 p position 0 15 0 1 0 9 0 9 n pix n elements mean psd 0 n frames 2 n elements mean psd 0 1 window 0 xsize 800 ysize 600 plot frequency mean psd 12 yrange 0 1 xtitle Frequency Hz ytitle Power Spectral Density title Power Spectral Density for pixel 33 38 write bmp psd scan X33 Y38 overview bmp tvrd plot frequency mean psd 12 yrange 0 0 2 xtitle Frequency Hz ytitle Power Spectral Density title Power Spectral Density for pixel 33 38 write bmp psd scan X33 Y38 bmp tvrd plot frequency mean psd 12 yrange 0 0 05 xtitle Frequency Hz ytitle Power Spectral Density title Power Spectral Density for pixel 33 38 write bmp psd scan X33 Y38 zoom bmp tvrd wdelete 0 end 11 6 TRANSFER FUNCTION COMPUTATION pro transfer fft model fft chop transfer function transfer function table SYNTAX transfer fft model fft chop transfer function transfer function table tft const mean Computes the modulation transfer function of the camera fft chop contains the individual ffts of each scan version 21 09 2000 Identification of the frequencies sampled table freq fltarr 1024 limit 180 limit2 130 limit3 80 for i 0 100 do begin if abs fft model i GT limit then begin H if
194. vious sections LISA number of frames acquired per scan LISA number of frames saved per scan LISA pixel coordinates associated with 11 12 P1 P2 set of coordinates for windows LISA percentage of the fiber bundle light in each pixel LISA exposure time seconds LISA number of data points between the start of the acquisition and the start of the record left margin LISA number of data points between the end of the record and the end of the acquisition right margin LISA quicklook fringe detection level n sigmas LISA full frame LISA full frame exposure time Seconds Fast scan piezo mirror incidence angle radians Fast scan piezo total OPD range meters Fast scan piezo scan OPD range meters Fast scan piezo scan OPD total OPD Fast scan piezo voltage range Volts Fast scan piezo voltage offset Volts Fast scan piezo waveform signal type sine square sawtooth triangle Fast scan piezo waveform stiffness Fast scan piezo waveform frequency Hz Fast scan piezo waveform delay seconds Fast scan piezo number of waveform samples Fast scan piezo sampling waveform sampling frequency Hz Fast scan piezo sample interval meters Fast scan piezo sample time Seconds Piezo camera user settings OPD range fringe velocity sample interval RA Dec taken from the Hipparcos catalog or the FK5 epoch J2000 0 Proper motion with uncertainties Parallax measurements Hipparcos Separated magnitude estimations
195. work only if the motors of the INA1 and INB1 mirrors are replaced by piezos which is not the case in the current implementation of LdV It is taken from a proposal by Steve Ridgway NOAO The idea is to scan the image rapidly across the fiber in order to freeze the seeing But then it takes multiple scans to build up signal to noise This is not a problem because you really want the measurement to average over the seeing fluctuations Since you have a piezo control it is possible to put a ramp signal on the piezo and scan repeatedly the fiber head over a range of 1 arcsec Synchronized with the scans the signal is coadded into a vector When finished you should have a peak in the vector which corresponds to the nominal position You can then move the piezo to the correct position Then repeat in the other direction This should be faster more accurate and more sensitive than alignment by hand 2 20 1 2 Slow image scan algorithm Another possible scheme for the automated optimization of the injection in the fibers is 1 Obtain quick exposures from the LISA camera in not synchronized mode in order to obtain a nearly continuously frequency of a few Hz monitored value for the flux arriving on the HAWAII detector The maximization of this value will be the goal of the optimization 2 Start a slow spiral motion of one of the injection mirrors say INA1 to slowly move the stellar image disk in front of the fiber head During this motio
196. ws that the baffling defect takes place optically between the dewar window and the filter wheel These tests show that the background is as computed in the study phase of the camera at about 40000 e s pixel The solution to reduce this background would be to add a cold stop in front of the doublet lens Currently only the filter wheel is playing the role of an imperfect cold stop Though it is expected that the short exposure times will not suffer from this background A picture of the LISA cold mechanics is presented on Figure 11 and Figure 12 They show the cold mechanics down to the detector support A possible location for an additional cold stop would be close to the black part on the left of Figure 11 As a conclusion the addition of a cold stop is not absolutely required for now but would be an important plus to the current design It will be mandatory for an upgrade of the camera to a lower readout noise detector LISA Test Report b ELE EE 5 October 2000 20 Figure 11 LISA cold mechanics and baffling seen from the front photo MPE Figure 12 LISA detector support and baffling seen from the back photo MPE 5 3 DETECTOR READOUT NOISE IN FULL FRAME MODE All the measurements described below are done in double correlated readout mode which is the one used for the scientific observations LISA Test Report b ELE E 5 October 2000 21 5 3 1 Aproximate gain factor The gain factor is computed from the sl
197. xels 21 5 4 READOUT MODES si epe tedio pedet bugie eee p pg SEENEN dee 21 5 4 1 Listof mod6es siia eaa healed itio a alatis eode n eH eitis 21 5 4 2 Off line windows generator ss 22 5 4 3 Full quadrant double correlated eene 22 5 4 4 64x64 window double coreloted esee eene eene eene eene entes 5 4 5 Windowed readout double correlated 5 4 5 1 Beampix data is Aere eegene 5 4 5 2 Flux data 5 4 5 3 Quicklook data 6 READOUT FREQUENCIES 6 1 INTRODUGCTION dei ATE C ERR SER IRE CEINTURE RE ERE EE Ye ETE dg 25 VLT TRE ESO 15810 2330 1 0 5 October 2000 LISA Test Report 6 2 EFFECTIVE MAXIMAL READOUT FREQUENCIES iii 25 T BEST PHOTOSITES SELECTION iscsccscsesscacssesssesssesssssceccsscssenssacsassssssscsseescecssocssecessaseosesssssnetecsseestensoss 26 7 1 PRINCIPLE EE 7 2 PIXEL EE 7 2 1 Measurement procedure eet stats deett ete ded it e niet 7 2 2 Examples of gain CUPVES 4i Ae esee te dre Pre aee de eaten cente ka 7 2 3 Gain MAP E 7 2 4 Histogram of pixel gains eet e eet ee tarte e tetigit 7 3 PIXEL READOUT Ce 7 3 1 Measurement proc dure eege ien ee etel eie Ee eege 7 3 2 TEE 7 3 3 Histogram of pixel readout noises sen 7 3 4 Readout noise at the highest frame frequency ss 7 4 PIXEL NOISE POWER SPECTRAL DENSITY ie 7 4 1 Low frequency 03 33 Hz as A CARB HAGO edente ROB On acte eR 7 4 1 1 Data ana
198. y maximizing the flux on the camera pixels The setting of the injection parabolae is motorized and the flux maximization is done on the stars during the observations This first setting during the initialization phase will help finding the maximum light position on the scientific targets by giving a reference point As an upgrade of the system it should be possible to do this process automatically Acquisition of a LISA full frame this will enable the observer to check for the quality of the images of the fibers on the HAWAII detector and track any systematic problem If the centering of the fiber heads is bad then an output optimization procedure can be started Off source and mixed batches In order to compute the transfer matrix of the system it is necessary to obtain three small batches about a hundred scans Off source the telescopes are pointed to the background sky beamAonly beam B only Fringes Finding long scan and Scanning fast scan Req 51 during the fringes finding phase the optical path difference is slowly modulated by moving the delay line while scanning with the fast scan mirror INA3 By doing this motion a segment of optical path difference is scanned for the presence of fringes The delay line motion is superimposed to the sidereal speed The algorithm is simple VLT SPE ESO 15810 1852 LEONARDO da VINCI 1 1 16 September 1999 Software User Requirements 43 while fringes are not fo
199. y of the off axis parabola using the two reference pods on the back of its support This can be done with a micrometric comparator or as an alternative with a precise slide caliper The reference for the measurement should be the optical table itself but as a second choice the base plate of the OUT source can also be used VLT TRE ESO 15810 2330 1 0 5 October 2000 8 LISA Test Report 4 DOUBLET FOCUSING 4 1 ALIGNMENT PROCEDURE 9 Collimate roughly the artificial light source following the Meudon procedure on a distant wall for example The adjustement is done by moving the fiber head itself The image produced by the laser should be a tiny spot of light less than 1 mm across with round shape and sharp edge Two small fainter dots on each side of the main spot should be visible they are due to the microgrooves of the off axis parabola Turn back the source in the correct direction Measure the relative positions of the two benches on the VINCI table The references are the center of the LISA dewar window and the center of the OUT2 flat mirror The precision required depends on the distance between the source and the dewar the further the lower the required precision A precision better than 1 mm relatively to the table edge should be sufficient assuming a distance of 1 or 2 m between LISA and OUT2 Send the laser in the OUT system and adjust the OUT2 mirror only to have the red dot centered on the entrance window of

Download Pdf Manuals

Related Search

Related Contents

Art.-Nr. SM 210.1  Inscription Formation 14-15 juin 2007  Ryobi P2002 User's Manual  WP-MS11  MODE D`EMPLOI  DEVELOPMENTAL Vol.3, No.7  User Manual - Title page  Einleitungshandbuch - Minerva Metrology and calibration  Produktanleitung  

Copyright © All rights reserved. Failed to retrieve file