Adaptive Optics simulation IDL code A User Manual
Contents
1. 0 257022 1 02809 7 00000e 07 1 00000 0 0100000 0 00000 CHAPTER 2 THE INPUT FILES Voltages on the DM to record interaction matrix Voltages on the TT to record interaction matrix Method of inversion modal or zonal Conversion matrix mode gt actuators Mode enable disable for modal inversion of eigenval to force to Zero in the inversion conditioning number of matrix to invert File to read interaction and command matrices fr bimorph or piezostack or zernike or tip t mirror par file for radial config only file that contains the influence functions Is there a tip tilt mirror 0 1 Maximum voltage on the electrodes deformable mirror hysteresis linear number of actuator piezostack only curvature hartmann WFS geometry radial or square Sensor par file radial config only WFS wavelength in meter or Linear number of subaperture SHWFS only Pixel size in arcsec SHWFS only subaperture size in arcsec SHWFS only e rms read noise SHWFS only Blur image to avoid discretization noise SHWFS Threshold on SHWFS in count if positive in percent if negative e g 0 5 50 and no threshold if 1 Extra focal distance in m CWFS only of point in a 1 4 wave of mm scanning function Diameter of the field stop CWFS only Gain of the running average on subaperture flux fraction of pupil diam obstructed in the CWFS includes the effec2. 0 01 and so on This is convenient for command line batches Here is an example of a parameter file File written on Wed Mar 21 11 18 45 HST 2001 Parameter file for simul pro Version 4 3 and later INTERACTION MATRIX ACQUISITION AND INVERSION zim v_dm v tt inversion mta mode enable forcenev cond number matrices MIRROR mir type parfile def_file tt v_limit hyst nxact 0 00000 SENSOR wfs type geom parfile lambda SHWFS only nxsub H pixsize subsize rnoise blur threshold CWFS only p 1 sinus fieldstop gainrav mask GUIDE STAR tes noise 1gs altitude thickness laser_2w0 lauch_diam lauch_pos lauch_ttm lauch_ttm_gain extended gsmap gsmapscale starmag skymag zeropoint TIP TILT GUIDE S ttgswfs ttgs starmag skymag psize fov lambda threshold gain noise ooo 5 0 95000 0 8000 00 0 300000 0 450000 0 00000 10 0 0 500000 0 500000 zonal PIE 3 22 1 0 0100000 mats fits 3 bimorph mirror par bim s64 p60 e19 fits 0 6000 00 50 curvature radial sensor par 7 00000e 07 parameters T 0 262033 2 09626 3 3 00000 arameters 0 150000 0 3 00000 1 00000 0 300000 1 0 800000 Os Jextended3 fits 0 0500000 8 00000 20 3000 2 00000e 10 TAR AND SENSING 10 0000 20 7000
3. gt gt batch 1 gt gt batch 1 gt gt batch 1 gt gt batch 1 res res res res APPENDIX A BATCH EXAMPLES 27 This is a quite elaborated example already The goal was here to investigate the performance of several subaperture configurations for a 36 or so curvature system pro optim36 define the common to be able to access the parameters common aostr ao Create a file named optim36 res to summarize the results com echo resultats optim gt optim36 res res execute spawn com Start simul this reads out the par file and put the variable into the common area simul par ao par optim 130 s0 define the configurations I want to evaluate configs 4 6 10 16 1 6 12 18 1 6 12 16 2 6 10 18 3 6 10 18 3 6 12 16 2 6 12 16 nconfigs size configs 2 frint ol Pe no Ho no N oh er frext loop on the configurations for i 0 nconfigs 1 do begin load the current configuration config configs i Some formating of the summary file openw 1 optim36 res append printf 1 444444444444444444 printf 1 config close 1 loop on ring radius within a given configuration for j 0 3 do begin r cwfs_design config 0 1365 r 0 r O frint j r n elements r 1 r n elements r 1 xfrext j 28 APPENDIX A BATCH EXAMPLES write the mirror and sensor par file for this configuration writ
4. 5 for an explanation of how they are normalized their format etc Optionnal files sensor par This is the sensor parameter file for radial configurations SHWFS defined on a square geometry do not use this file Used exclusively for curvature sensors in the current version 3 mirror par wfs_stat_ab par sort_modes par CHAPTER 1 INTRODUCTION This is the mirror parameter file for radial configurations PZT mirrors defined on a square geometry do not use this file Used exclusively for bimorph mirrors in the current version This file contains the zernike coefficient of the WFS path phase aberrations This file allow to sort the modes found by the modal decomposition in a given order The following is a list of output files written by simul simul res sensor fits pupil fits Influence functions mats fits mta fits simul avimage modevar dat cb_mes cb_com cb_modc cb_modc This is the journal file into which every operation executed by simul is dumped For each operation matrix acquisition inversion initialization loop run it contains the state of the system and the results of the operation Ascii and editable Fits file of the sensor configuration Just for information Image of the pupil For information only in the form pzt _s _p _ fits for the piezostack mirrors where the are the size of the array e g 128 if 128x128 are used the diameter of the pupi in pixel and t
5. appear as the 4 first modes of the basis In the current version the modal inversion works only with curvature wavefront sensors not with Shack Hartmann WFS ao zim mta Name of the matrix from which to read the mode to actuator transformation coeffi cients Usually called mta fits because is saved under this name by the routine that computes it included in simul This matrix is saved to save time during further executions of simul if using the same configuration of course ao zim mode enable Name of a fits file into which the previous user selection have been stored One is written each time the interactive mode selection routine is ran ao zim forcenev Number of eigenvalues to force to zero in the matrix inversion process Usually between 1 and 10 At least one has to be discarded the one corresponding to the piston mode and also the one corresponding to the deformable mirror tip and tilt when using an additionnal tip tilt mirror Sometimes one may wish to zero some other eigenvalues that are off the main group of eigenvalues because of instability or noise consideration ao zim matrices Name of a file recorded in a previous run and saved as mats fits that contains the interation matrix and the command matrix Be careful in using this option that the file you indicate is the right one You don t want to use a file taken with a different set of parameters and try to close the loop with it The purpose of this parameter is to sav
6. distance of star 1 in arcsec or 40 distance of star 2 in arcsec or N SOFTWARE PARAMETERS MONITORING AND DEBUGING 64 size of the standard arrays 60 0000 Pupil diameter in pixel should be lt size 4 5 Display the results on screen 0 1 0 compute modal decomposition 0 1 0 Compute MTF MTF SNR and Structure function o Output of meas mirror commands and modes Description of the parameters e zim substructure ao zim v_dm Voltages to apply to the deformable mirror to record the interaction matrix This should be large enough to avoid numerical errors and low enough to be in the dynamic range of the sensor A voltage that give a signal which is likely to be typical while on the atmosphere is recommended Typically of the order of 1 to 100 volts ao zim v tt Same for the tip tilt mirror Same order of magnitude e mir e wfs CHAPTER 2 THE INPUT FILES ao zim inversion modal or zonal zonal uses a Singular Value Decomposition algorithm and excludes ao zim forcenev eigenmodes modal uses a mode basis computed from the influence functions These modes are orthogonal in the phase space i e their cross product integral over the pupil is zero The modal inversion allow to cancel particular modes among all the modes interactively or not A pure piston tip tilt and defocus mode to be more precise modes which shape is the closest to these modes the mirror can achieve
7. in routines v_dm inversion mode enable matrices type def file v_limit type parfile nxsub subsize blur 1 fieldstop gs noise ttgsufs all gs gsmap gs starmag gs zeropoint do_inter init mkmatcom init init init init init do_inter loop do inter loop loop loop init loop loop do_inter loop do_inter Zim v tt Zim mta Zim forcenev geom lambda pixsize rnoise threshold sinus mask gs lgs all gs extended gs gsmapscale gs skymag do inter init mkmatcom init init init do_inter loop do inter loop init init ao tel diam init do inter Toop ao loop type ao loop n_iter ao loop aliasing ao loop start_skip dro screenl screen3 layer2 frac layerl speed layer3 speed layer2 alt ao isop enable ao isop diststar2 ao sim size ao sim monitor ao sim calcmtf loop loop loop loop loop loop loop loop loop ao loop delay ao loop im_lambda ao loop gain ao loop it_time directory screen2 layerl frac layer3 frac layer2 speed layerl alt layer3 alt loop loop loop loop loop loop loop loop loop loop loop ao isop diststar1 loop ao sim pupd ao sim calcmode ao sim cb output init loop loop Table 4 1 Parameter names and routines that uses it This is mostly to allow the user to determine if a new initialization or a new matrix or nothing is needed after updating parameters init stands for initialization d
8. laser 2w0 Beam diameter at the lauch telescope m A good value is 2 3 r0 laser wave length ao gs lauch diam Lauch telescope diameter m ao gs lauch_pos Lauch telescope position 0 means that the lauch telescope is behind the sec ondary 1 means it is just at the edge of the primary mirror ao gs lauch ttm 1 means that there is a Tip Tilt steering mirror on the upward beam Setting this parameter to 1 will also filter the TT from the AO WFS command to the AO system TT mirror the AO WFS TT on a LGS should be directed to the LGS steering mirror ao gs ttm gain Gain of the TT steering mirror if any ao gs extended Set to 1 if wavefront sensing has to be done on an extended object ao gs gsmap name of the fits file that contains the image of the extended object This image shall not contain diffraction of the telescope or the subaperture It is the image as seen from space using a telescope of infinite diameter a model of the object Use images of reasonable size e g 64x64 It is recommended that the object is well restrained within the image frame i e no discontinuities at the edges to avoid ringing when Fourier transforming it ao gs gsmapscale scale of the guide star map above in arcsec pixel ao gs starmag Magnitude of the reference star at the sensing wavelength ao wfs lambda ao gs skymag Magnitude of the sky per arcsec at the sensing wavelength In Shack Hartmanns the field i
9. month mostly when I was at CFHT It was initially developed to investigate the performance of the PUEO system 19 subapertures curvature system on the Canada France Hawaii Telescope located atop Mauna Kea in the big island of Hawaii I must warn the potential user that it is not bulletproof and certainly not bug free However IDL is kind with us and most of the time will crash gracefully and allow and easy tracing of the error I try to follow and solve bugs If you find one please let me know at frigaut gemini edu The files required to run the code are listed below More detailled descriptions can be found in the next section Some of the files are mandatory some other are optionnal the program check if they are in the current directory and read them out in this case Some files may be also required to define the system geometry in the case of a non square subapertures actuator pattern Mandatory files simul pro File that contains the main IDL procedure simul plus a bunch of other subroutines This file has to be somewhere into the IDL_PATH in order for IDL to find it It has to be compiled when entering IDL by typing run simul ao par This is the main parameter file It contains the parameter defaults To change it you can either edit it manually or use the editor provided in the simul GUI Phase screen s Phase screen s used to simulate atmospheric turbulence e g cscreen128 10 fits Those are fits files See below section 2
10. structure into the file The best is to run simul and to have a look at simul res file It should be self explanatory 23 24 CHAPTER 5 THE OUTPUT FILES Appendix A Batch Examples This is a simple example of batch It uses the batch command line language of IDL not a procedure To run it one should type IDL gt batch The goal was here to investigate the performance at several WFS integration time Start simul with parameter file ao par 14x14 Monitor is set to zero so no graphic window is created simul par ao par 14x14 mon 0 define the common to be able to access the parameters at this level common aostr ao Initialize init Create a result summary file spawn echo results from batch 1 gt batch 1 res Set the WFS integration time run a loop and store the results ao loop it time 0 002 loop spawn tail 21 simul res gt gt batch 1 res Same for a different WFS integration time etc ao loop it time 0 004 loop spawn tail 21 simul res gt gt batch 1 res ao loop it time 0 006 loop spawn tail 21 simul res gt gt batch 1 res ao loop it time 0 008 loop spawn tail 21 simul res gt gt batch 1 res ao loop it time 0 010 25 26 loop spawn tail 21 ao loop it time loop spawn tail 21 ao loop it_time loop spawn tail 21 ao loop it_time loop spawn tail 21 simul res 0 020 simul res 0 040 simul res 0 080 simul res
11. the phase is preserved Using this the effect of spatial aliasing is not taken into account so that one can estimate the effect of purely the noise or the servo lag This is only a first order approximation as the presence of the high order modes may affect the noise properties of the wavefront sensor for instance because the spots are extended ao loop gain Gain of the close loop This is a first order loop with a simple integrator NeWcommand Oldcommand gain x Acommand ao loop start_skip Number of iterations to skip before collecting performance statistics usually 10 to 50 ao loop it time Iteration time in seconds substructure ao atm dr0 D ro at WFS wavelength ao atm directory Path relative or absolute to the phase screen directory ao atm screenn with n 1 to 5 Phase screen file names See section 2 5 NONE OF THESE 5 PHASE SCREEN SHOULD BE THE SAME for statistical reasons the seeing statistics assumes the layers are uncorrelated if it is not it changes the statistics and the results quite drastically ao atm layern_frac with n 1 to 5 Fraction of the turbulence in each layers The three fractions must sum up to 1 If you want to use a single layer just set one fraction to 1 and the others to 0 ao atm layern_speed with n 1 to 5 Speed of the layers in meter per second ao atm layern_alt with n 1 to 5 Altitude of the layers above the telescope in meters e isop substructure 12 e sim CHAPTER
12. 1 0 0 for i 0 n elements config 1 do printf i rint i rext i config i 0 1 format f4 2 3x f4 2 3x 13 3x f3 1 3x i1 printf 1 1 8 close 1 end 29
13. 1 It has to be noticed that if monitor is set to 1 some interactive features are turned on which then requires an operator action beware of this in batches than may get stuck at this point If this is set to 0 in the parameter file and the GUI is not used then no window is created which therefore prevent any ulterior graphical display Usefull for batches where one does not want windows to keep the possibility to run the batch as a background process ao sim calcmode Toggle mode variance computation and output to a file ao sim calcmtf Toggle the computation of the Modulation Transfer Function the SNR on the MTF and the phase structure function ao sim cb_output If set the measurements the mirror commands and the mode coefficients are output to files resp cb mes cb com cb_modnc and cb_modc at each iteration Note 1 IDL defines the structure members type Integer float string the first time he reads it into an IDL session Make sure you assigned the right type to the parameters you modify in ao par For instance do not assign the value 1 to ao loop gain otherwise it will be taken as an integer by IDL and you will not be able to assign a non integer value to it later in the same session instead set it to 1 0 If this happens you have to exit IDL modify the parameter value in ao par and start IDL again Note 2 Do not change the syntax of a line You may change the parameter value but do not remove coma or bra
14. 2 THE INPUT FILES ao isop enable Set to 1 if isoplanatism computations are to be done It takes some time a fraction of what is needed for one loop iteration so do not use it if not necessary ao isop diststar1 Distance of star 1 to the guide star in arcsec ao isop diststar2 Distance of star 2 to the guide star in arcsec substructure ao sim size Size of the standard arrays on which the pupil and the phase is defined Should be a power of 2 for FFT purposes Usually 32 64 or 128 32 may be small for most of the purpose Array sizes over 128 lead to quite long execution times see section Running time considerations ao sim pupd Diameter of the pupil in pixel Should be smaller than the array size ao sim size Usually about 80 to 100 of the later parameter A good compromise is to have about 3 points per ro at the sensing wavelength so if for instance D ro equals 25 a pupil diameter of 75 is well suited More is even better up to 4 or 5 points per ro if one can afford the execution time Also sometimes it doesn t cost more to put larger pupil diameter for instance ao sim pupd 75 imposes ao sim size 128 so that one can set ao sim pupd to 120 without much extra cost Warning ao sim pupd is forced to the closest integer value that verify the condition that the number of pixels in the pupil plane should be constant for each subaperture ao sim monitor Graphic display of the results on screen flag equals to 0 or
15. Adaptive Optics simulation IDL code A User Manual Francois Rigaut March 21 2001 Contents 1 Introduction 2 The input files 21 a0 DAT 4 aa man BEDA 2 2 Dimensioning the system 2 3 sensor par 24 mirror par 2 5 The phase screens 3 The Main Routines Sado AMI Fina OO oe Moe ANAL 3 2 doimter 3 3 mkmatcom 3 4 loop 4 Running simul 4 1 X form interface to simul 4 2 Running Simulfromthecommandline 4 3 Running time considerations 5 The output files A Batch Examples 12 13 13 14 15 15 15 16 16 17 17 18 20 23 25 CONTENTS Chapter 1 Introduction This code is written exclusively in IDL Interactive Data Language language It should be executed in the IDL environment This code is intended to simulate adaptive optics systems featuring curvature wavefront sensors CWFS or Shack Hartmann wavefront sensors SHWFS in combination with piezostack PZT or bimorph mirrors BIM The code is organized into subroutines that allow to take interaction matrices invert them and run a dynamic close loop simulation that includes the temporal behavior of the loop The whole thing was developed by myself in an approximately 6 months period Later I added subroutines and a GUI Version 4 2 and 4 3 were developped in the mist of 1997 1999 I estimate the full developement time to something around 12 man
16. another form see next page You can modify any fields The modification is taken into account without doing a carriage return It will not accept parameters of an other type integer float or string than the type that has been defined for the parameter when IDL first read ao par see note 1 in the section ao par You may leave that parameter window open and push buttons on the main simul window After having edited modified the parameters if necessary you have to initialize before doing anything else Push the button initialize The code will perform the functions described in the section 3 1 and if ao sim monitor is set it will display a contour of the subapertures with the electrode actuator pattern superimposed in the configuration window if it has not exited in error because it didn t find some file before that Once initialized you have to have a command matrix to start a loop If the interaction and command matrices have been restored ao zim matrices set to a file name previously saved you can skip the interaction matrix acquisition and inversion steps You may also invert the matrix again with different inversion parameters This will of course override the command matrix If no matrices have been read then you need to acquire one with Acguire interaction matrix and invert it with Invert interaction matrix Again be careful you have set the deformable and tip tilt mirror voltages to a suitable value If ao sim monitor
17. aperture with respect to X axis in degrees Direction of rotation 1 clockwise 1 anticlockwise aaaaakaaaaaakaaaaakaaaaalakakakaabaalakaaabalabalakaaaalalalakaaaaalbakalaaalalalialalalalalalalalalalalalalaltallalalakalakakalalakalak start Radial 4 0 0 0 3135 4 0 1 0 3135 0 4985 8 0 1 0 4985 0 7410 16 O 1 0 7410 1 3500 24 0 1 The content of the file is explained in the header The code will build up the sensor subaperture s configuration from the parameters and print out the total number of subapertures Note The parameter which defines the direction of rotation is not used in the IDL version of the simulations there was a fortran version prior to this one which used it 2 4 mirror par This file sets the parameters required to define the geometry of the mirror when the radial configuration is chosen Remember this file is not read therefore not needed if a square configuration is selected Here is an example Mirror file For bimorph mirror Start after string start Description eo dede ae kkk gt gt od de ok gt gt ak aka kakak dead ak ze ak aka kakak ke kakak added line variable type option meaning 1 integer Number of ring central subaperture included 2 Real Real Mirror ellipticity Angle of incidence 3 R R I R I Inner and outer radius of ring 1 Number of subapertures in ring 1 Angular offset of first subaperture with respect to X axis in degrees Direction of
18. ckets otherwise IDL won t be able to read out the structure Note 3 For any modifications of the number of parameters You may add variables into the existing sub structures but do not add additionnal structures When adding parameters respect the syntax name value and then coma or closing bracket if end of substructure 2 2 Dimensioning the system It is recommended to e have several pixel per ro at the very least 2 e have more than 5x5 pixels per subapertures SHWFS or CWFS in the pupil plane That is if using a 10x10 SHWFS the pupil should be at least 50 pixels in diameter e don t go for values smaller than 0 04 for the CWFS extra focal distance 2 3 SENSOR PAR 13 2 3 sensor par This file sets the parameters required to define the geometry of the sensor when the radial configuration is chosen Remember this file is not read therefore not needed if a square configuration is selected Here is an example Sensor file Both for Curvature sensor and Shack Hartmann Start after string start Description sk sk sk sk ok ok ok ok sk sk kk kk ka kok ska EEE EEEE EEE EEEE EEE EE EEE EE EE EE EEE E E EE E E E E E E E line variable type option meaning 1 character 1 R or C Sensor geometry Radial or Cartesian 2 integer Number of ring central subaperture included 3 R R I R I Inner and outer radius of ring 1 Number of subapertures in ring 1 Angular offset of first sub
19. diameter m 0 421000 Central obs diam in frac of pupil diam P close Loop type close or open 0 Loop delay in sampling time 0 or 1 4000 Number of iteration in close loop 2 20000e 06 Imaging wavelength in meter 0 Aliasing 0 no action 1 remove low order 2 remove high order 0 600000 Overall gain of the integrator 50 Number of step to skip before averaging data 0 00100000 Iteration time s 19 0000 D rO at the WFS wavelength gt data Path to the dir that contains the screens cscreen128 16 fits Filename of the first phase screen cscreen128 18 fits Filename of the second phase screen cscreeni28_20 fits Filename of the third phase screen gt cscreeni28_10 fits Filename of the 4th phase screen cscreen128 12 fits Filename of the 5th phase screen 0 180000 Fraction of turbulence in layer 1 0 280000 idem for 2 Sum of fi should sum up to 1 0 180000 idem for 3 0 190000 idem for 4 0 170000 idem for 5 d 8 00000 speed of layer 1 in m s d 10 0000 speed of layer 2 in m s if negative goes backward d 15 0000 speed of layer 3 in m s d 25 0000 speed of layer 4 in m s d 15 0000 speed of layer 5 in m s 10 0000 altitude layer 1 in m 1600 00 altitude layer 2 in m 3600 00 altitude layer 2 in m 6300 00 altitude layer 2 in m 11600 0 altitude layer 3 in m SM 1 Compute isoplanatism effect 20
20. e execution time or allow the user to always use the same command matrix from one run to another substructure ao mir type Valid options are bimorph piezostack zernike or tip tilt ao mir parfile Parameter file for bimorph mirror on a radial geometry See below Usually mirror par ao mir def_file Fits file that contains the influence function for each electrodes actuators If not set that is set to an empty string the program will compute a new set of influence function from the mirror parameters This file is created by the code each time it computes new influence functions In that case the program will print on the screen the file name it saved the influence function in It is usually on the form bim_s128_p120_e56 fits or pzt_s128_p120_10x10 fits where the size of the arrays the pupil size in pixels and the number of electrodes actuators are quoted Warning There is no check that the actual file matches the other mirror parameters Please make sure that the entered file name matches with the array size the pupil diameter and the number of electrodes actuators respectively ao sim size ao sim pupd and ao mir nxact or ao mir parfile ao mir tt Whether there is a tip tilt mirror or not 0 1 ao mir v limit Limit voltage on the deformable and tip tilt mirror Used in the loop procedure to limit the voltages ao mir nxact For PZT mirror number of actuators in a pupil diameter For Zernike numbe
21. e of the denominator in the expression of the measurements Ni N m N No The running average has the form NG m NI Nad rav t where rav t g x Ni t No t 1 9 x rav t 1 CHAPTER 2 THE INPUT FILES This scheme was at one time proposed to increase the SNR of the measurement It turned out that the area over which the SNR is improved is quite narrow 3 5 photons subapertures according to M Northcott and marginal In every other cases this degrades the measurement SNR Usually I recommend to not use it i e set its value to 1 ao wfs mask CWFS Only Diameter in fraction of the pupil diameter of the mask if any placed directly in front of the microlens array Usually 0 Was implemented for PUEO where an inverted telescope scheme is used to expand the beam in front of the lenslet The mask is in that case the secondary mirror of this inverted telescope For PUEO this value was 0 3 compared to the central obstruction of the CFH tetescope 0 44 e gs substructure ao gs noise Flag to indicate that the program should consider noise read out noise and photon noise 0 or 1 ao gs lgs 0 or 1 depending whether the WFS is using a LGS 1 or not 0 Setting this parameter to 1 enable the LGS related parameters 7 following parameters ao gs altitude Altitude of the LGS m ao gs thickness Thickness of the layer m sodium layer or gated rayleigh a0 gs
22. e_parfiles config r Initialize in modal to get the mode to actuator matrix and mode expression This is to get the propagation of the noise ao Zim inversion modal ao sim calcmode 1 ao zim mta init do inter mkmatcom make the command matrix to be used in zonal ao Zim inversion zonal 0 ao sim calcmode mkmatcom run the loop loop grab the results from the output files and put them in the results summary file openw 1 optim36 res append amp printf 1 r 8 close 1 com cat modes_noise dat gt gt optim36 res res execute spawn com com tail 21 simul res gt gt optim36 res res execute spawn com endfor endfor com mv optim36 res optim36 res 130 s0 res execute spawn com end pro write_parfiles config r this is a procedure that writes par files sensor and mirror for a given config according to the format recognized by simul pro Just to avoid to pre type them all n n elements config rint shift r 1 rint 0 0 rext r rext n 1 1 5 openw 1 sensor par printf 1 start printf 1 R printf 1 n_elements config format i1 for i 0 n_elements config 1 do printf 1 rint i rext i config i 0 1 format f4 2 3x f4 2 3x i3 3x f3 1 3x i1 close 1 rint n 1 rext n 1 HK e DO openw 1 mirror par printf 1 start printf 1 n_elements config format i1 printf
23. he number of actuators respectively and in the form bim _s _p _e fits where the are the size of the array e g 128 if 128x128 are used the diameter of the pupi in pixel and the number of electrodes respectively Those influence fonctions are saved in order to be used for subsequent use of simul in the same configuration This file is written at the complexion of the initialization Fits file that contains the interaction matrix and the command matrix This file is written at the complexion of the command matrix computation routine Transformation matrix modes to actuators electrodes Written only if modal inversion is re quired Beware this is a fits file It contains the average compensated image over the whole loop run Written at the complexion of the loop procedure Ascii file Contains the mode variance in a loop run Written at the complexion of a loop run The parameter ao sim calcmode has to be set for this file to be written The cb xx are what is called circular buffer for historical reason Here they have nothing of circular nature This one contains the measurement One line written at each iteration of the loop run ao sim cb_output has to be set for this file to be written As cb _mes except it contains the mirror commands As cb _mes except it contains the mode coefficient of the modal decompositon of the input turbulent phase As cb _mes except it contains the mode coefficient of the modal decomposi
24. her with the interaction matrix and the command matrix These two matrices are then saved into a fits file called mats fits The user may rename this file to store it for later runs 34 loop This routine simulates a close loop including an evolving atmosphere and a one frame delay integrator The scheme is as follows 1 Read out the phase screens 2 Initialize statistic metrics 3 Start the main loop a Fill the phase with the translated and scaled phase screens values b Add the mirror shape to the phase deformation c Feed the phase to the WFS routine and get measurements d Compute the mirror commands by multiply the measurements by the command matrix e Compute new mirror shape to be applied on next iteration f Compute variances Strehls and average image g Display results on screen 4 Save performance metrics Chapter 4 Running simul The simul routine must be started at the IDL prompt After entering IDL the routine has to be compiled Just type run simul at the IDL prompt There is then two possibilities to run simul and perform the different operations From an X form or from the command line This is detailled in the next sections 41 X form interface to simul Just type IDL gt simul x An X form will pop up At this point you can modify Edit parameters or list the parameters show parameters or make other actions by clicking on a button Clicking on Edit parameters or Show parameters will pop up
25. in the right configuration Typically 128x 1024 or 1024x 128 or 200x1000 phase screens are suited You can emulate an atmosphere with less layers than 5 by setting the other layer fractions to 0 but all the screen name should be filled with a valid file name Chapter 3 The Main Routines The simul procedure is divided into 4 main subroutines When simul is started the file ao par is read and the ao structure is filled There is then basically four routines to achieve the simulation These are listed in the following subsections 3 1 init As indicated by its name this subroutine initializes the main arrays and variables Here is what it does Build up the pupil array filled with 0 and 1 Initialize WFS arrays i e build up an array that contains the subaperture number at each pixel on a scale that matches the pupil definition Initialize the mirror influence functions i e read the ao mir def_file if not set to a null string exit in error if file is not found otherwise compute the influence functions Computing the influence functions may take quite a while on large arrays 64 or 128 especially for bimorph mirrors which uses FFT on arrays 4 times larger e g 512x512 for array sizes of 128 Initialize size of various arrays Read out interaction and command matrices if ao zim matrices is not set to the null string Exit in error if file is not found Initialize the graphic display 4 windows even if monito
26. is set the eigenvalues the modes corresponding to the zeroed eigenvalues the interaction matrix and the command matrix will be displayed At this point you may start a loop simulation Once you click on Start loop some initializations are done and a message will appear on the loop window that tells you what operations can be entered while the loop is running Basically you can update D ro type d in the IDL root window the loop gain type g the extra focal distance CWFS only type 1 you can also reset the voltages on the mirrors type r toggle the 17 18 CHAPTER 4 RUNNING SIMUL graphic display type m or stop the loop type q This also works when loop is started from the command line Once done you can change parameters and start again It may not be clear for the user where to start from when some parameters are changed The following table gives a summary of the parameters and where to start from in the routine hierarchy init matrix acquisition matrix inversion loop in this order when the parameter is modified Note concerning X window management into IDL If the program exit in error while running simul with an X form you can always restore the control to the X form by typing IDL gt xmanager This will give back the control to the event manager and allow you to press new buttons 4 2 Running Simul from the command line simul can also be runned in a command line mode This is especially useful to ma
27. ke batches for instance to test a serie of different parameter sets without continuous user interventions To start simul type IDL gt simul The code will read the parameter file ao par and load the parameter default values into the ao structure This is the only purpose of the main routine simul For the ao structure to be accessible you have to declare it at the main IDL level by typing IDL gt common aostr ao You may now change parameter values e g IDL gt ao sim monitor 1 or start a subroutine e g IDL gt init As described above 4 subroutines are available init do_inter mkmatcom and loop In this mode 4 windows are created at the initalization level to which the graphic displays will be directed if ao sim monitor 1 You may also include part of the commands or all of the commands in a batch In IDL starting a batch is done by typing immediately followed by the name of the batch file For instance IDL gt ao batch An example of a batch file is given below and more examples are given in appendix A simul common aostr ao ao mir def file ao zim matrices ao wfs 1 0 30 ao wfs parfile ao mir parfile bim s64 p60 e80 fits mats s64 p60 e80 fits gt sensor par 80 mirror par 80 ao atm screen1 screen14 f 800 200 fits init ao gs starmag 14 ao loop gain 0 3 loop 4 2 RUNNING SIMUL FROM THE COMMAND LINE 19 Parameter name Used in routines Parameter name Used
28. lation code is not intented to provide the user with a versatile tool It is clear that the running time plus the necessity to have large statistical sample imply several minutes to several hours per configuration Hence a complete investigation of the parameter space which contains quite a lot of parameters is difficult if not impossible The goal is more to precise confirm results obtained with analytical code 21 4 3 RUNNING TIME CONSIDERATIONS ig dj PUDWWOJ Inang prosan snl py INT APM Irama RG LOT IRAQUE SI nbay Sagad POLE Sungompaed Jipa BEDE ANDAR SOG T PUP SSL We su Seely JOGUDE JALAN dar JIT agpang suc Figure 4 1 Main X form of the GUI 22 ao zim y dm 30 Zim y tt anzim inversion 30 21m mta CHAPTER 4 RUNNING SIMUL Parameter menu L ao zim mode_enable ao Zim forcenev ao zim matrices ao mir type ao mir parfile a0 mir def file ao mir tt ao mir v_limit an mir nxact ao ws type binorph nirror par 000 00 50 Eurvature 7 Parameter Help Done Figure 4 2 Parameter form Chapter 5 The output files These files are described in the introduction As mentionned in there simul res contains a complete set of information Each time one of the routine is ran it appends the results to simul res together with the complete set of parameters Parameters are loggued by writing the complete ao
29. ointer for acquire interaction matrix and mkmatcom for invert interaction matrix 20 CHAPTER 4 RUNNING SIMUL ao gs starmag 16 ao loop gain 0 15 loop ao gs starmag 18 ao loop gain 0 1 loop In this batch some environement parameters are modified their default values specified in ao par are overri den Then the routine init is called In this example the interaction and command matrices were computed in a former run and stored in a file This file ao zim matrices is here read by the init routine The next step is a serie of loop runs with different magnitude values At the complexion of each loop results are appended in a file called simul res as detailled in section 5 4 3 Running time considerations Note this section was written in 1994 Things have changed some since then but the order of the system we want to simulate have also gone up so it might still be lengthy to run simulations It is clear that you need a fast machine to run simul You need also quite a bit of RAM space 64 MB should probably besufficient for reasonable application but 128MB is better and is required for large system simulation 128x128 arrays for instance Simulating 36 subapertures CWFS on 64x64 arrays is fast lt 1s iteration on a pentium equiped machine or a Ultra Sparc workstation On the same machine a SHWFS of 14x14 subapertures on 128x128 arrays may take a couple of second per iteration This numerical simu
30. r is not set monitor may be interactively turned on off when the loop is running Print out the ao structure and the main parameters in a file called simul res 3 2 dointer This routine does the interaction matrix acquisition It first acquire a reference set of measurement by feeding a null phase to the WFS routine Each influence function scaled by the required voltage is then send to the wavefront sensing routine and the normalized measurements are stacked up in the interaction matrix Do not forget to set ao wfs blur to 1 when using a Shack Hartmann with a large pixel size now automatically turned on The noise is automatically turned off when doing the interaction matrix acquisition However the number of photon corresponding to the selected magnitude is used together with the selected threshold SHWFS 15 16 CHAPTER 3 THE MAIN ROUTINES 3 3 mkmatcom This routine inverts the interaction matrix using a Singular Value Decomposition method zonal or by an inverse modal filtering modal Following the standart SVD procedure for inversion the eigenvector matrices are swapped and transposed and the diagonal matrix containing the inverted eigenvalues is builded up The invert of the N lowest eigenvalues N equals to ao zim forcenev is forced to zero instead of a large number in this diagonal matrix If ao sim monitor is set the eigenvalues and the mode shapes corresponding to the zeroed eigenvalues are displayed toget
31. r of modes piston is inluded substructure ao wfs type curvature or hartmann are the only valid choices ao wfs geom WFS and mirror geometry radial or square are the only valid choices Note In the current version curvature sensors work only in radial and Shack Hartmann sensors only in sguare ao wfs parfile Parameter file for CWFS on radial configurations See below ao wfs lambda Wavelength at which the wavefront sensing is done In the results the variances turbulent phase variance corrected phase variance etc but not the Strehl ratios are expressed at this wavelength 2 1 AO PAR 9 ao wfs nxsub SHWFS only Number of subapertures in a pupil diameter ao wfs pixsize SHWSF only Size of the pixel in arcsec at the lenslet focal plane The program will print the subaperture size in arcsec during the execution You may then modify this parameter to match your requirements Usually between 0 05 and 0 5 arcsec pixel Will be rounded to the nearest multiple of the actual pixel size imposed by the sampling in the pupil plane ao wfs subsize Subaperture size in arcsec Taken into account only if smaller than the subaperture size imposed by the sampling in the pupil plane Updated with the actual value Will be rounded to nearest possible value multiple of updated pixel size One of the current restriction of the program is that subaperture size are usually limited to a rather lo
32. rotation 1 clockwise 1 anticlockwise 4 R Fixation radius aaa start 4 00 0 0 0 3135 4 0 1 0 3135 0 4985 8 0 1 0 4985 0 7410 16 0 1 1 0000 1 7000 24 0 1 1 78 14 CHAPTER 2 THE INPUT FILES The content of the file is explained in the header The code will compute the influence functions for each electrodes remember this is for bimorph mirrors only from these parameters and will write a fits file containing the influence functions It will print out the total number of electrodes and the fits file name Note The parameter which defines the direction of rotation the angle of incidence and the mirror ellipticity are not used in the IDL version of the simulations 2 5 The phase screens These should be fits files with a rO per pixel of 1 with the phase expressed in radian The conversion to a given D ro is done in the loop subroutine The phase screens are shifted across the pupil to simulate moving layers Screens 1 3 and 5 are translated along the X axis same direction and screens 2 and 4 are translated along the Y axis All the screens must have the same dimension The shorter axis must be at least equal to the pupil dimension in pixel The length of the long axis determines the maximum number of iteration achievable in a loop run The long and short axis position NAXISI greater or smaller than NAXIS2 in the fits file has no importance The phase screens are transposed by the routine if not
33. s limited by the size of the subaperture a field stop of the same dimension is assumed In curvature sensors the field of view is defined by the parameter ao wfs fieldstop ao gs zeropoint Zeropoint for photometric measurement i e number of photons detected by the WFS all subapertures added for a zero magnitude star at the sensing wavelength Must be around 10 or 101 for 4 m to 8 m telescope with good optical and quantum efficiencies e ttgswfs substructure ao ttgswfs ttgs If set to 1 it means that a Tip tilt guide star and a TT WFS are used Again to avoid conflict this also filters the TT from the AO WFS output ao ttgswfs starmag Magnitude of tip tilt guide star if any ao ttgswfs skymag Sky background magnitude arcsec ao ttgswfs psize TTWES pixel size ao ttgswfs fov SHWFS Field of view per subaperture in arcsec ao ttgswfs lambda TTWFS effective operating wavelength 2 1 AO PAR 11 e tel ao ttgswfs threshold TTWFS image threshold Same rules as for the AO WFS ao ttgswfs gain TT loop gain ao ttgswfs noise TTWES read out noise in photon substructure ao tel diam Telescope diameter in meters ao tel cobs Telescope central obscuration diameter in fraction of telescope diameter e loop substructure ao loop type close or open Self explanatory ao loop delay This is the pure delay of the close loop in addition to the one frame delay intrisic to
34. t of measurement noise 0 1 LGS 1 yes 0 no gt NGS Altitude of LGS Na layer thickness Beam waist 2 beam diameter Lauch telescope diameter Lauch Tel position in telescope radius units 0 center 1 edge TT mirror on upward beam if yes also filters out TT form HRWFS gain of the integrator of lauch telescope TT mirror Extended guide star 0 1 map of guide star 2D Must not include diffraction pixel size in guide star map arcsec pixel Magnitude of guide star Magnitude of background sky arcsec 2 normaly 20 7 Zero point phot number detected by WFS s for m Don t make it smaller than r0 Is a Tip tilt guide star and WFS used also filters HRWFS TT output Magnitude of tip tilt guide star Sky background magnitude arcsec 2 SHWFS pixel size SHWFS Field of view subaperture SHWFS wavelength SHWFS image threshold TT loop gain SHWFS read out noise 2 1 AO PAR 7 TELESCOPE tel diam cobs CLOSE LOO tloop type delay n_iter im_lambda aliasing gain start_skip it_time ATMOSPHERE atm dro directory screenl screen2 screen3 screen4 screen5 layerl frac layer2 frac layer3 frac layer4 frac layer5_frac layerl spee layer2 spee layer3 spee layer4 spee layer5_spee layerl alt layer2 alt layer3 alt layer4 alt layer5_alt ISOPLANATI fisop enable diststarl diststar2 SIMULATIO sim size pupd monitor calcmode calcmtf cb_output 3 60000 Pupil
35. the close loop principle If for instance you want to modelize a system using a CCD with a one frame delay induced by read out time set delay to 1 Value up to 3 are accepted ao loop n_iter Number of iterations for the close loop Actual number of iterations may be smaller as the loop subroutine exit the close loop when the phase screen has been scanned to the end or the user hits q at the keyboard ao loop im_lambda Imaging wavelength in meter The Strehl ratio and the FWHM are expressed at this wavelength ao loop aliasing If set to 1 then the correction modes are subtracted from the input turbulent phase Therefore the variance of these modes in the corrected phase is only due to aliasing and noise if any To compute the weight of aliasing in a system set this to 1 set the noise to 0 and run it on a long enough sample with a small enough sample time to reduce the servo lag error The corrected mode variance gives you the amount of aliasing An alternative for linear systems Shack hartmann is to determine the aliasing in open loop This allows to enlarge the sampling time therefore requires less time to scan a complete phase screen Beware that this method allow the determination of spatial aliasing only excluding cross coupling which may exist in real systems If set to 2 the high order modes are filtered out from the input turbulent phase that is only the component parallel to the mirror modes in the modal decomposition of
36. ton of the output compensated phase Chapter 2 The input files This section gives more details on the content of the parameter files 2 1 ao par This file contains the default parameters to be used in the code It is read out when simul is started simul can also read other par files by using the keyword parfile i e IDL gt simul x parfile ao par my config cwfs19 All the parameters defaults can be changed to accomodate user requirements This file is formatted so that IDL can read easily its content as an IDL structure It is in fact made of 9 IDL structures that are later gathered into one large IDL structure called ao The 8 IDL individual structures are substructures of ao namelly ao zim Interaction matrix parameters ao mir Mirror parameters ao wfs WES parameters ao gs Guide Star parameters ao ttgswfs Tip Tilt guide star and WFS parameters ao tel Telescope parameters ao loop Close loop parameters ao atm Atmosphere parameters ao isop Isoplanatism related parameters ao sim Simulation parameters It is very easy to list or modify each substructure elements when into IDL These elements can be addressed by the name of the structure dot the name of the substructure dot the name of the element For instance one can address the mirror type by the name ao mir type So one can modify parameters on the IDL command line by typing simply for instance ao mir type piezostack ao gs starmag 14 5 ao loop it_time
37. urce When closing the loop on a guide star blur should be set to 0 no way to blur the image in a real SHWFS except when using light intensifiers which have an extended PSF In the current version blur is put automaticaly to 1 when doing interaction matrices and to 0 for the loop run ao wfs threshold SHWFS only Threshold to apply in the centroid calculation in number of equivalent photons The threshold is subtracted to each subaperture s focal images and the negative pixels are forced to zero A threshold of 1 means no threshold A threshold between 0 and 1 is understood as relative to the image maximum e g 0 5 means threshold at 50 of the current image maximum ao wfs 1 CWFS only Extra focal distance in meter in a F 60 beam Usually between 0 05 and 0 5 depending on the D r0 at the sensing wavelength and the number of subapertures e g 0 07 for 19 subaperture and 0 6 arcsec seeing and a sensing in the visible 0 3 for 80 subapertures and 0 45 arcsec seeing ao wfs sinus CWFS only Number of point in a quarter period of the membrane mirror sinusoidal scanning function Usually it is better to chose a value greater than 7 to reduce discretization effect and lower then 20 to keep the computation time reasonable ao wfs fieldstop CWFS Only Size of the field stop in arcsec Used only to compute the sky background contribution to the subaperture s count ao wfs gainrav CWFS Only Gain of the running averag
38. w value This is a consequence of the sampling in the pupil plane and of the properties of the Fourier transforms Indeed the maximum size of the image plane is directly set by the sampling in the pupil plane following the relation 0 Awe PDP 5 4884e 6 D where s is the maximum field of view PDP the pupil diameter in pixels and D the pupil diameter in meters An 8 m telescope for which a pupil diameter of 110 pixel is used working at a wavelength of 0 7 microns leads for instance to a maximum size of 2 arcsec which is small and for some case smaller than what the user may want to use test The only way out in the current version would be to use larger pupil size 200 or more which therefore imply using larger array size 256 This is of course extremely CPU time consuming ao wfs rnoise SHWES only CCD rms read out noise The noise is assumed gaussian The CCD gain is 1 one ADU per photon Therefore this value is in equivalent number of photons ao wfs blur SHWFS only If set to one each subaperture focal images is convolved by a gaussian of FWHM roughtly equivalent to the size of the expected typical seeing This is to get rid of the undersampling problem when working with point like sources trying to compute a center of gravity is greatly affected if the source image is much smaller than the pixel This has to be used only when taking interaction matrices In the real word it is equivalent to taking an extended light so
Download Pdf Manuals
Related Search