HERA User Manual
Contents
1. Determination of the total integration time let cal time 1 15 15s to make a CAL COLD in min let int time nstep nummap 2 t let int time int timekcal timexint timelcal inter let int time int timex1 30 let cal num int cal_inter t 1 1 Determination of the offsets of the central channel offsets are relative to the source in OBS let count 0 let offx0 omx efsx 2 2 d dsta let offyO omy efsy 2 2 d dsta SET ANGLE SEC pause Ready to start pause Derotator CAL COLD SET OFFSETS BASIS for nsubmap 1 to nummap let q int nsubmaptnmx 1 nmx let p nsubmap q 1 nmx for n 1 to nstep let count count 1 let 1 int ntnidir 1 nidir let m n 1 1 nidir let offx offx0 ptm d p ds let offy offy0 qt1 d q ds OFFSETS offx offy BASIS if count eq cal_num 1 then CAL COLD BELL BELL let count 0 endif 21 START SAY SUBMAP OFF_LAM OFF_BET SAY nsubmap offx offy next BELL pause SUBMAP nsubmap next BELL say eziz ou MAP FINISHED return 22 24 L 1 syypo o A 24 0 24 right ascension offset Figure 8 Orientation of the HERA pixel pattern used in the oversampling OTF mapping mode The pixel pattern is rotated by x 9 5 in the equatorial system Two OTF scans are made separated by A 3 x As 12 in declination A second pair of such maps not shown offset in decl2. OBS gt set derot 2 S where S makes the derotator tracking the sky rotation and 6 is an angle constant in time by which the HERA pixel pattern is rotated with respect to the equatorial system counting east from north The derotator com mand is executed when a LOAD is made in the telescope control or in other words before a new scan is started As an example commanding 9 0 keeps the 9 pixel pattern of HERA stationary in the equatorial system so that the horizontal rows of 3 pixels e g pixels No 7 5 and 2 are always parallel to right ascension like shown in Fig 4 In this seemingly simple setup the HERA center pixel No 1 in Fig 4 is on the telescope s pointing axis and all peripheral pixels have fixed in time offsets in the equatorial system of Aa and Ad of 0 or 24 according to Fig 4 The angle by which the derotator can be physically rotated in the frame system is limited to 84 As the K mirror multiplys rotation angles by a factor of two this allows to rotate the pixel pattern on the sky within a range of 336 degrees However while tracking it might happen that the derotator approaches these physical limits To avoid loss off data the observer should keep an eye on the newly installed HERA monitor webpage ask the operator or the AOD to open this web page in the automatic update mode within the screen ensemble of the observer The current angle and tracking system is displayed in the left upper corner o
3. 2 Technical Description HERA has nine dual polarization pixels 18 Channels arranged in the form of a square center filled 3x3 array The distance between pixels on the sky is 24 i e close to twice the beam width FWHP at 230 GHz By means of a quasi optical K mirror derotator this pattern can be placed at any position K Schuster A Greve P Hily Blant P Planesas A Sievers C Thum and H Wiesemeyer contact A Sievers C Thum or K Schuster 50 om K Mirror Derotator with ellip Top Dewar Movable Mirror for Window internal Cold Load SH 2nd allip a Mirror Crossed Wire Grid 4 Daikin Crys trunc Lens Array Generator E corrugated Horns amp Coupler Mixer Units Figure 1 Optical layout of HERA including the derotator angle on the sky see also sect 4 The two polarization modules are named HERA1 and HERA2 Both modules can independently be tuned to different sky freqencies and be attached to different backends 2 1 Cryogenics HERA is cooled with a 3 stage closed cycle refrigerator system DAIKIN which greatly reduced technical overhead times and running costs The receiver can be warmed up in 12 hours and be cooled down within 36 hours Usually the receiver is kept cold for periods of at least 3 months followed by a maintenance period of typically 4 5 days 2 2 Optics HERAs optics fig 1 is designed to give maximum point source coupling of the individual beams and a
4. 90 main beam efficiency Beg 0 52 Beam size at 230 GHz 11 77 2 4 Derotator The Derotator is an optical ensemble in front of the HERA cryostat Fig 1 which allows to compensate for the rotation of the astronomical object in the focal plane of the telescope This rotation is due to the general char acteristics of an AZ Alt mounted Nasmyth telescope and a function of ac tual Azimuth and Elevation The derotator commpensates this rotation by computer controlled movements The end effect is that the pixel pattern and polarization can be kept fixed on the sky in the equatorial system or alternatively in the horizontal system Fields can therefore be easily sam pled homogeniously and a great number of particular observing strategies are possible The derotator is constructed as a K mirror a device which is frequently used in optical astronomy and the first implemented in the millimiter range for HERA To make efficient use of HERA it is essential that the user integrates a minimum of information about the derotator before observing The current control software allows to keep the HERA pixel pattern sta tionary i e derotate various hour angle dependent rotations in the following three coordinate systems F Frame Nasmyth coord sim Bolo stationary in the receiver cabin H Horizontal Az El Syst usefull to align array with wobbler dir S Sky RA Dec Syst most common mode The derotator is steered from OBS by entering e g
5. A 3K 3K aK R 3K 2K R SR Ok aK aK R SR Ok 3K R SR K k K k SR Ok k K R SR Ok 2K 2k ok k k k k INPUTS define real sx sy define real sr 19 define real cal inter t define real omx omy let sx 41 let sy 62 let sr amp 3 let cal_inter amp 4 let t 65 let 66 let 67 1 OUTPUTS define real offx offy define real int_time LOCALS define integer nmx nmy nummap define integer nstep nidir define real d ds define integer m 1 p q define real a define real beam define real cal_time define real fac define integer cal_num count define real offx0 offy0 define real efsx efsy L k k k k k K K k K K k k k k k kK KKK K k k k k k KK k k k k k k k k k k K k k k k k k k k k K K K bo Definition of some constants relative to HERA let beam 12 let a 24 let d beam sr some submap parameters ds size of the submap same in both dir np nb of points in one submap let ds 1 5x2xa d let nidir int ald let nstep nidir 2 nb of submaps along both directions let fac sx dstd if mod fac 1 1t 0 5 then let nmx int sx ds d 20 else let nmx int sx dst d 1 endif let fac sy dstd if mod fac 1 1t 0 5 then let nmy int fac else let nmy int fac 1 endif let nummap nmx nmy let efsx nmx dstd d let efsy nmy dstd d
6. and HEO1 is pixel No 1 of HERA see also Tab 1 All VESPA configurations are available with all observing modes Read out speed dump time in OTF phase time in switching modes is however limited to ca 10000 channels per sec by the VAX antenna control computer If the full number of VESPA channels is connected read out should not be made faster than once per 2 sec The new correlator is now the wide band usuable bandwidth 18 x 930 MHz WILMA backend offering 9 9 spectral bands The setup in OBS is simple the following OBS gt WILMA A Receiver HERA1 or HERA2 sets the first 9 spectral bands and 17 OBS gt WILMA B Receiver HERA1 or HERA2 sets the second 9 spectral bands The distribution of the 4MHz and WILMA backends and HERA1 and HERA2 is now done automatically by a new distribution box Only the 4MHz has to be switched manually between single pixel receivers and HERA Table 3 VESPA configuration table channel spacing bandwidth MHz number of channels usage kHz nominal actual raw data RED calibrated 1250 160 141 2 1152 9x 113 25 1250 320 281 2 2304 9 x 225 50 1250 480 421 2 3456 9 x 337 75 1250 640 511 2 4608 9 x 409 100 320 80 70 3 2304 9 x 225 25 320 160 140 3 4608 9 x 449 50 320 240 210 3 6912 9 x 673 75 320 320 280 3 9216 9 x 897 100 80 40 35 1 4608 9 x 449 25 80 80 70 1 9216 9 x 897 50 80 120 105 1 13824 9 1345 75 80 160 140 1 18432 9 1793 100 40 20 17 5 4608 9 x 448 25 40 40 35 0 9216 9 x 89
7. three scanning proce dures i e observing procedures exist data reduction is tested and they are handled by the web based time estimator version 2 4 and higher These three scanning procedures are associated with different orientations of the beam pattern rotation angle x with respect to the scanning line taken here as right ascension zero rotation labeled in the Time Estimator SL OTF not inclined coarse sampling SL OTF 8 arc sec oversampling SL OTF 4 arc sec In the zero rotation mode the beam pattern is scanned parallel to right ascension like in Fig 4 The points on each of the 3 scan lines are sampled three times but the area between the three horizontal beam rows is not sampled This mode may be useful for obtaining a first quick look on a weak extended source In the coarse sampling mode the beam pattern is rotated by x 18 5 This results in a slightly under sampled map where the scan lines are sepa rated by As 7 6 Fig 7 The map is homogeneously sampled apart from 60 wide vertical margins at the left and right ends of the scanning lines In the oversampled mode the beam pattern is rotated by x 9 5 This generates three bundles of three scan lines spaced by As 3 9 the separation between the bundles being A 3 x As Two OTF scans must therefore be made see Fig 8 which are offset by Ad in declination The scan direction of these two OTF scans may be the same as shown in the f
8. 6 50 40 60 52 5 13824 9 x 1344 75 40 20 70 0 18432 9 x 1792 100 20 20 17 5 9216 9 x 896 50 20 40 35 0 18432 9 x 1792 100 5 Data reduction Extensions for HER A Data reduction for HERA data is very similar to the one for single beam receivers The only difference is that the raw data contain now information about the derotator angle offset and the actual tracking system Up to now RED and OTFCAL generate only for obser vations with derotation in the equatorial SKY system the correct pixel offsets for all HERA channels For these standart observations the user will obtain calibrated single or otf 30m spectra with the correct offsets exactly equivalent to any other 30m spectral data files A current draw back is that currently OTFCAL and RED use the nominal derotation angle instead of the actual The usually negligtible errors in servoing of the derotator will not be taken into account Unfortunately this leads to wrong offsets in case of a stuck derotator Data are assigned to the various pixels together with the backens within the telescope parameter of the 30m scans 18 NEW Calibration via ODP All the spectral line data including some OTF maps and the 4MHz data are now calibrated on line on our linux file server mrt lxl For details of Online Data Processing ODP see the documentation http www iram es IRAMES documents projectDataProcOnline Current MAIN html We say some OTF maps since we calibrate only maps that h
9. H06 1 07 1H08 1H09 2H01 2H02 2H03 2 04 2 05 2 06 2HO7 2 08 2 09 VVILMA VVA 1 01 1H02 1H03 1 04 1 05 1H06 1 07 1H08 1H09 VVILMA VVB 2 01 2 02 2H03 2 04 2H05 2 06 2HO7 2 08 2 09 1 01 1H02 1 03 1H04 1 05 1 06 1 07 1 08 1H09 2H01 2H02 2H03 2H04 2H05 2 06 2 07 2H08 2 09 3 2 Focus HERA s derotator re images the telescope focal plane which is located about 28 cm in front of the derotator rotation axis onto the dewar window The separation between the HERA mixers and the subreflector is therefore con siderably longer than for the other SIS receivers Nevertheless HERA was installed in such a way that its telescope focus setting of the secondary mirror is within 0 5 mm of that of the other receivers Nasmyth vertical Nasmyth horizontal Figure 3 HERA beam pattern at the telescope focal plane in the Nasmyth cabin viewed from the receiver The derotator is at its default position set derot 0 F This is not the sky pattern The possibilities to focus HERA are quite similar to the pointing op tions There is no difference to single beam receivers for focusing onto the central element and one uses the OBS command set recoff to focus on off elements To focus on off center pixel No n e g because it is more stable in continuum than the center pixel one goes through the following sequence of OBS commands set derot 0 F set recoff xn Yn focus n corr k x F if fit is satisfactory set reco
10. HERA User Manual The commissioning team version 1 9 February 21 2005 1 Introduction The HEterodyne Receiver Array HERA is a receiver system with 18 SIS mixers tunable from 215 to 272 GHz The main purpose of this first multi beam spectral line receiver on the 30m telescope is to allow together with the related backends for rapid raster and on the fly imaging of spectral lines in the 1 3mm atmospheric window Other more particular observing modes continuum measurements and mapping polarization measurements and optimized wobbling on point sources are possible but not yet offered as standard observing modes HERA has been embedded into the existing setup of the 30m telescope as much as possible Observing commands OBS and data structure follow closely that of the single pixel spectroscopy receivers Only where necessary new or modified commands have been introduced and only these will be described in this manual However in order to make efficient use of HERA the user needs to be aware of some basic technical aspects of the instrument the particularities of the data stream and typical observing strategies The current version of the manual includes the description of observing with both polarizations and the new WILMA wide band correlator Complementary information can also be found in Schuster et al 2004 A amp A see IRAM 30m webpage Suggestions for improving this manual are welcome and should be addressed to A Sievers
11. LSB is the normal and preferred tuning mode as it offers superior HERA 55 Performance Mx Al Mx A2 Mx MxB1 Mx 82 Mx BS Mx 1 Mx 2 C3 RF GHz Figure 2 SSB receiver noise temperatures across the tuning range These temperatures were measured in the laboratory in front of the dewar window overall system noise However it must be recognized that LSB tuning results frequently in reduced total power stability which somtimes makes pointing on weak sources difficult The noise performance of HERA1 over the tuning range is given by the Fig 2 HERA2 has a very similar performance for frequencies up to 230 GHz but is higher in noise above 230 GHz In practice these noise temperatures have to be increased by a factor of 1 15 if VESPA is used as the backend IRAM s correlators are presently 2 bit 4 level Another small increase 3 5 K comes from losses on the derotator mirrors It is planned to upgrade the mixers of HERA in the future The local oscillator tuning range is 219 to 276 GHz Taking the IF frequency of 4 GHz into account this results in a RF tuning range of 215 to 272 GHz for LSB tuning standard tuning mode DSB tuning is also possible The IF bandwidth is 1 GHz Tuning is automatic and uses look up tables Because these look up tables have not yet been finalized the tuning has to be prepared in advance apart from a series of heavil
12. ath in the equatorial system syyo 24 0 24 right ascension offset Figure 4 HERA beam pattern on the sky when set to track the rotation of the equatorial system with zero offset angle OBS gt set derot 0 S The correspondence between pixel numbers and spectrometer sections is described in Tab 1 The upgraded web based time estimator version 2 5 and higher handles these modes in an approximate way Simple integrations on point sources are also possible 4 1 General aspects 4 2 Single Stamps and Observations of Point sources The very basic observations are single integrations with a fixed position on the sky This will result in 18 spectra on nine position of the sky see Fig 4 If requested the pattern can be turned around the central pixel with the help of the derotator command Before starting large maps it is allways recommended to do such a single shot in order to make shure that the system has been setup properly The pattern can also be offset from the pointing center by the usual commands of OBS If observed with the derotator tracking in the sky system the spectra will be written with the correct offsets into the spectra 30m file The use of an array receiver offers also a certain signal to noise advantage compared to single pixel receivers So far we tested several wobbler switching schemes where the source was ob served more efficiently than in the 50 available with the standard
13. ave a subscan on the reference position at the start and at the end of the on source subscan s Basically a pattern of RFF R that may also repeat The calibrated spectra are written in CLASS format to the file rart lx1 mrt data project directory spectraOdp 30m and and interme diate spectra used for on line display are written into the file mrt lx1 data mrt project directory spectraPlot 30m NEW On line display for ODP The on line display for ODP is not started automatically the observer needs to start a terminal session on mrt x1 iram es and then type rart tcs tools odpPlot At present this is only a quick look using CLASS to display the spectra as they are produced not yet a comfortable viewer 6 Appendix A raster map pro Vo a k k k k k K K k k k k k k k k k k k k K k k K k k k k K k k k k k FK k k k k k K k k FK k k FK k k ok k KK k K K 1 1 RASTER mapping vithout cross calibration consideration Inputs size sx X Sy sr sampling rate cal_inter time between 2 CAL COLD integration time per point omx offset of the center of the map along X omy offset of the center of the map along Y 1 1 Qutputs the offsets of the central channel of HERA the time needed to make the map I PHB 17 1un 2001 I PHB 25 jul 2001 1 ak akak ak R K SR S R SR SR R k R SR SR I SR SR SR
14. e 7 Orientation of the HERA pixel pattern used in the coarsely sam pling OTF mapping mode The pixel pattern is rotated by x 18 5 in the equatorial system This results in a uniform OTF map where the scanning lines parallel to right ascension are separated in declination by As 7 6 If uniform coverage is to be extended toward the north a second map is made with a declination offset of A 9 x As 68 spectroscopic baseline Flat baselines are obtained by setting the frequency throw equal to this value or multiples of it Satisfactory performance was obtained up to throws of 6 x 6 9 MHz 54 km sec Frequency switching rates up to 10 Hz were successfully tested with both HERA modules were connected to the VESPA backend sect 4 1 and Tab 3 set to 80 kHz resolution and 40 MHz bandwidth i e 50 usage of VESPA capacity However there is usually no need with the raster map procedure to switch faster than about 1 Hz 4 4 On the fly maps OTF mapping the most powerful spectroscopic mapping mode on the 30m becomes even more powerful with HERA So far we have expanded on the single receiver OTF in the following two ways scanning procedures some with special orientations of the HERA beam pattern reduce the time for mapping by factors somewhere be tween 3 and 9 14 e frequency switching is possible with the VESPA backend eliminating the need for reference observations Scanning procedures We support the following
15. e of pixels is aligned with the horizontal direction in which the subreflector was wobbling The wobbler throw is set equal to the pixel separation V2 x 24 when a diagonal pixel sequence is chosen and the derotator is commanded as OBS gt set derot 45 H The standard Wswitch command of OBS then executes two observations with telescope offsets of throw 2 The schema shows that pixel 2 sees the source in each observation for 50 of the time exactly as in standard wobbler switching with a single pixel receiver Additionally HERA pixels No 1 and 3 are also looking at the source but only during one of their phases Since the wobbler switched signal is the difference between the two observations pixels No 1 and 3 register the source signal at a level of 50 Averaging the signals from the three pixels with their right polarity therefore increases 2 the signal to noise ratio by J over the standard wobbler switching In 3Compared to single pixel wobbler switching the signal is doubled and the noise is increased by v3 This S N ratio increase of 15 is improved to 22 if optimum 1511 weights are used when averaging the spectra 2 2 7 from the three pixels 11 24 o o o 0 o o o 24 o o o 24 0 24 right ascension offset Figure 6 The simplest fully sampled raster map called a submap in the text has a size of 66 x 66 The points indicate the positions observed by ste
16. ee sect 2 4 when looking towards the subreflector The central pixels of HERA1 No 5 and of HERA2 No 14 are aligned with the tele scope s pointing axis and with the other SIS receivers to within better than lt 2 This degree of alignment was measured to be independent of the ac tual derotator position For pointing with central pixels it is therfore not necessary to set a particular derotator angle or tracking sys tem and the standard pointing model can be used unmodified for HERA This includes on line display of the central pixel on the RED screen and the Gaussian fits for deriving the pointing offsets The resulting correction parameters can be taken into account by the usual command OBS gt corr f f Pointing with off elements Although somewhat more complicated pointing with HERA is possible also with offset pixels The results are always referred to the position of the central beam This option may be used if an ofF pixel has a much better continuum sensitivity or stability than the central pixel Right now a few commands have to be entered in OBS to make pointings with off elements OBS gt set derot 0 F OBS gt set recoff xn yn OBS gt point n OBS gt set recoff 0 0 The first command sets the derotator to zero degrees its default position no rotation in the Nasmyth cabin Next OBS is informed about the offset in Nasmyth coordinates of the receiver pixel used for pointing Then OBS executes the pointing with pix
17. el No n including an on line display of that pixel Again the resulting pointing fits can directly be accepted as a pointing corrections for the array with corr F F no further offsets have to be given Fig 3 and Tab 1 show the correspondence between pixel No n and their offsets H and V in Nasmyth coordinates note that we have renumbered the continuum pixels for HERA1 The arguments xn yn of the OBS set recoff command have the opposite signs as these arguments are offsets for the telescope pointing which bring off pixel n back onto the pointing axis The nominal offsets of the pixels as given in Tab 1 have been determined from beam maps and are valid to a precision of better than 1 Do not confound these Nasmyth offsets to be used only with the OBS command set recoff with the offsets of the pixels from the pointing axis on the sky In the horizontal system the Nasmyth pixel pattern is rotated by the elevation angle In the equatorial system an additional rotation by the parallactic angle is made Table 1 HERA pixel pattern and corresponding spectrometer sections filterbank 4M HERA1 pixel No 1 2 3 4 5 6 7 8 9 HERA2 pixel No 10 11 12 13 14 15 16 17 18 Nasmyth H 24 24 24 0 0 0 24 24 24 offsets V 1 24 424 24 424 24 24 Xn 24 24 24 0 0 0 24 24 24 set recoff yn 24 0 24 24 0 24 24 24 spectrometers VESPA VO 1 4 1 01 1H02 1H03 1H04 1H05 1
18. f this webpage When the DEROTATOR box is turning yellow the derotator is still tracking normally but its time for action If it turnes red tracking is off due to the mechanical limits Mapping part of an extended source with the derotator stuck in a limit may result in un reproducible distortions If the derotator angle displayed in the derotator window is close to a limit yellow the derotator should be rotated by 90 Sky system in the right direction If the limit at 84 is approached set the angle to 9 90 where 6 is the angle demanded by the mapping mode see below A procedure is implemented in ASTRO which generates a plot of the derotator angle in the frame system as a function of time UT Open the input window in ASTRO Gildas version 2 and newer ASTRO gt input Then choose in the pull down window Pico Veleta the item Derotator angle This opens the input window where the source name mandatory the source position format hh mm ss s dd mm ss s and the desired dero tator angle often simply 0 After click on GO a plot of the derotator angle in the frame system is generated as a function of UT Note that the sign of this angle is still inverted with respect to the correct angle displayed in the derotator window 2 5 Mixers and RF performance HERA uses waveguide SIS mixers which can be tuned in LSB or DSB mode The coupling of local oscillator power is obtained over waveguide coupler modules
19. ff 0 0 reset pointing offset for central pixel Remark The focus procedure of the 30m telescope has shown some odds recently It is not unusual to have from time to time unsatisfactory focus fits which would result in very big offsets HERA has been proven to focus very closely to the position of the other SIS receivers Do not simply correct for proposed focus offsets of more than 2mm but double check these results with another focus observation 3 3 Calibration For HERA a specific calibration system is used consisting of dedicated hot and cold loads The cold load is integrated into the cryostat of HERA For the observer however this is transparent as the OBS command CAL COLD is redirected to control the HERA calibration system The only difference is that the cold load temperature may change slightly in time Because of this the temperature of the cold load is measured and displayed in the monitor window that also displays the derotator angles As this is not yet automatically taken into account in OBS the observer should check this temperature at the beginning of the HERA session on the HERA monitor ing screen Tcold eff and if neccessary change this value for HERA1 and HERA2 in OBS with the usual command OBS gt set chop Tcold eff receiver hera1 2 The image rejection is about 10dB in LSB tuning mode for all pixels Gn 0 1 for LSB and 1 for DSB tuning and set globally for both modules at the present time OBS gt set gai
20. ial Observing Modes 4 6 1 Rotated coordinate systems OTF mapping in coordinates other than the equatorial system are also pos sible If such a system possibly aligned with a molecular outflow is rotated by the angle relative to the equatorial system the derotator has to be given the same angular offset The corresponding OBS command then is OBS gt set derot S where x is as above the angle induced by the OTF mapping mode This situation is depicted in Fig 9 where the array scans along the A axis of a descriptive system centered at ao o and tilted by the angle with respect to the local meridian Change your observations into a rotated coordinate system needs some input into the hidden parameters of OBS OBSINP and a special setup for your source coordinates contact A Sievers in case 4 6 2 Backends HERA as a general purpose multi beam receiver has strong demands on spectral backends Tab 2 lists the available backends 16 Table 2 Backends for HERA type channel spacing bandwidth number status kHz MHz of units 4MHz FB 4000 1024 9 available VESPA 20 1250 20 640 9 36 available VVILMA 2000 1024 18 available The filterbank consists of 9 units with 256 channels each The channels have a half power width resolution of ca 5 MHz corresponding to a noise equivalent bandwidth of 6 4 MHz Channel spacing is 4 MHz The 9 units were successfully taken into operation in Febr
21. igure or opposite to each other At the end a homogeneously sampled map is obtained again apart from 56 wide lateral margins Note that for maps smaller than typically 2 on a side the size of these margins are a large fraction of the total area to be mapped Such small maps are therefore more efficiently done in raster mode Total power OTF is available with VESPA WILMA and the filter spec trometers The spectrometer readout rates and thus implicitly the scanning speed are limited to 0 5 Hz Observations with the 4 MHz filterbank clearly demonstrate the sensitivity of this observing mode to sky noise and gain vari ations in the receiver Together with the unavoidable nonlinearities of the backends sky noise and receiver gain variations lead to base line distortions and for the correlators to platforming Good and stable weather conditions are therfore mandatory for this observing mode The gain drifts are related to the temperature fluctuation of the cryostat which has an irregular cycle period of 3 to 5 min and sometimes intrinsic mixer instabilities Under good weather conditions Tsys 300 K the drifts which may be as rapid as a few seconds reach amplitudes of 0 2 K The gain drifts affect different sections of the bandpass differently and are statistically less affected near the times when a reference observation is made 15 Software is being developed to correct for the gain drift induced base line variations The mainl
22. ination by 6 x A 71 extends uniform coverage to the north 23 local A meridian lt b o 2 ol a Xo Figure 9 A rotated coordinate system is illustrated with an outflow source centered on the star x which is located at ao o in the equatorial system The new coordinate system has its equator A aligned with the outflow lobes and the orgin of the new system is at the star The latitude 8 makes an angle with the local meridian 24
23. n 0 1 1 receiver hera1 2 Right now the limited precision of the image rejection factors Gni the largest source of calibration errors The limited accuracy of the knowledge of sideband rejection leads to relative calibration errors in the range of 0 10 For high signal to noise mapping it is therefore recommended to undertake some additional steps for relative calibration or flatfielding see Sect XX A more precise rejection measurement using a line injection device is cur rently underway and will ultimately allow to retrieve individual rejections for each pixel from a look up table at each frequency 4 Spectroscopic observing modes As HERA is a heterodyne receiver we describe here only spectral line ob serving modes although continuum mapping is possible see section 5 Ina first step the basic observing modes are described and then we explain how the various backends can be attached Please consider that due to certain constraints mainly computer limitations observing modes and possible backend combinations are not independent e single stamps section 4 2 take nine point maps with a single inte gration e raster maps i e observations where the telescope steps through a series of points usually fixed in the equatorial system at each of which the telescope dwells for some integration time on the fly maps i e observations where data are taken while the tele scope moves at constant velocity along a linear p
24. pping the telescope pointing sequentially through the 16 offset positions Aa A 9 3 3 9 sect 4 3 relative to the telescope s pointing center The circles identify the instantaneous HERA pixel pattern at telescope offset Aa 3 and A 3 our particular observation Fig 5 weak signals are also detected in other pixels demonstrating that the source is actually extended The spectroscopic baseline is flat in all pixels no deterioration was detected compared to the standard wobbler switching with a single pixel receiver In a more severe test a long integration of a CO transition in a z 2 2 source was made The baselines were again flat and the line was detected Other wobbler switching schemes were tried where the full signal ampli tude was obtained not only for pixel 1 but also for pixels 2 and 9 The S N improvement in these more complicated schemes were however less than in the scheme described above they would become more interesting with larger arrays 4 3 Raster maps A slightly more complex but already very powerful observing mode is the raster where the telescope steps along a rectangular in the equatorial sys tem pattern of points At each point the telescope dwells for some integra tion time The special OBS procedure raster map pro not to be confused with the standard OBS raster command is used see Appendix A pg 19 OBS gt raster map size x size y ppb min i time Ad 12
25. s at 1 3mm wavelength wobbler switching This mode gives the best baselines on the 30m As the wobbler throw is at maximum 240 map and source size must be matched carefully if the reference beams are to be kept free of signal Note that contrary to position switching the position of reference beams refer to the current raster point not the map center and rotate with respect to the source frequency switching For many extended sources wobbler and even po sition switching may not be adequate and only frequency switching may be practical Owing to its position in the Nasmyth cabin HERA has the cleanest optical path of all 30m receivers Its frequency baselines are found to be the best recorded so far A ripple at 6 9 MHz probably originat ing from reflections between the subreflector and the mixers dominates the 4 This procedure raster map graphic together with the observing procedures RASTER MAP PRO OBSRASTER MAP2 PRO and OBSRASTER MAP4 PRO are copied to the project home directory in the subdirectory obs HERA on linux systems If you do not find them there you can copy these from mrt 1x1 vis default obs HERA 13 A R b za s l 72 b HE lt ede 24 lt p See eg 2 b lt eee 2 07 lt Yee 5 0 5 lt berk oem ey N lt J i lt sok 4 2 i N 24 0 24 right ascension offset Figur
26. t the same time to reduce deformations of the array geometry irrespectively of the position of the K mirror derota tor VVith this concept it is possible to obtain extremely simple observation procedures as it is not necessary to make field maps for calibration of the array parameters or other particular measurements before mapping The positions of the elements are simple functions of the position of the central elements and the chosen derotation angle The optics of HERA is largely reflective which reduces baseline ripples due to standing waves by a large amount 2 3 Optical data and efficiencies An internal cold hot load system is integrated into HERA and allows to calibrate all 18 channels in a single hot cold load procedure Due to its dual polarization and derotation concept HERA cannot be combined with other receivers Switching between HERA and other receivers is however possible within about 10 minutes HERA1 pixels have vertical in the Nasmyth cabin polarization and share the same local oscillator They have thus the same sky frequency HERA2 has horizontal polarization and a different LO its nine pixels can thus be tuned to a sky frequency different from that of the vertical pixels The relative pointing of the two polarization modules is better than 1 2 Both modules share the same focus setting The efficiency related parameters at 230 GHz are cold load temperature 97 K image band gain 0 1 forward efficiency Feg 0
27. uary 2003 The OBS command for setting up the filterbank is OBS gt 4MHz Receiver HERA1 or HERA2 The uncalibrated raw data are written to a LINUX computer An au tomatic calibration task generates a spectra 30m file which can be analyzed with CLASS The 4 MHz data are identified by their set telescope name The central pixel e g is designated as 30M 4M05 HERA The digital correlator VESPA offers spectral resolutions in the range 20 1250 kHz Tab 3 Up to 4 spectral bands per pixel are available for the current 9 pixel array The 4 bands can be placed anywhere in the lower half of the 1 GHz wide IF range A typical OBS command for setting up the correlator is OBS gt VESPA 1 320 160 0 0 Receiver HERA1 or HERA2 where one section of VESPA is connected to each HERA pixel Each section has a channel spacing of 320 kHz 160 MHz of nominal bandwidth and an offset of zero in MHz from the IF center This configuration uses 50 of VESPA hardware so that more sections could be connected possibly at off center frequencies Uncalibrated VESPA data from HERA are written by the VAX in the usual raw data format They are automatically calibrated displayed by RED and written to the usual spectra 30m file The calibrated HERA spectra are identified in CLASS by their set telescope name The central pixel e g is designated as 30M V01 HE01 where V01 stands for the first frequency band connected to the current vertical polarization array
28. where the seven arguments have the following meaning size_x size_y map size in x and y direction often simply in and in arcsec ppb number of points per beam 2 for Nyquist sampling min minutes before the next CAL COLD should be made i time integration time in minutes on each raster point Aa right ascension offset of the center of the raster map with respect to the telescope pointing position in arcsec A same in declination By setting ppb to 2 the instantaneous HERA pixel pattern is filled in like in Fig 6 and a fully sampled and homogeneous map is obtained consisting of 9 x 16 equidistant grid points Such a submap covers 66 square and it is command by OBS gt raster map 66 66 2 min itime Aa Ad Larger maps composed of several such submaps can be observed with the same OBS procedure Naturally coarser or denser sampled maps are also possible A raster simulation program is available The raster map procedure comes in three varieties position switching wobbler switching and frequency switching The observer uses the standard OBS commands for setting them before the raster map procedure is started position switching The raster_map procedure asks for the position offset of the reference position with respect to the map center It is possible to do several ON source observations per reference observation This saves observing time but was found to easily introduce poorer baselines Note that HERA operate
29. wobbler 10 M r wobbler telescope W status offset ON ATR ON 57 o darda OFF 0 Saxara x AN BD SP 270 1 ON 0 tor Av Au AA FD BE a ON throw 2 Xa Sey x AR 833 BD CG rf ON throw 2 dz A u MUZ gt azimuth 0 offset wobbler throw 200 100 Figure 5 Wobbler switching with HERA Right frame A linear sequence of pixels is considered the center one of which is located on the pointing axis of the telescope pointed at a source When the wobbler is switched on wobbler throw is equal to the pixel separation 4 beams are generated on the sky Plus and minus signs after the pixel number indicate the polarity of the wobbler phases negative ones are outlined in dashed lines Left frame Observation of an extended source IRC 10216 in 12CO 2 1 System tem peratures are below 300 K integration time is 2 min Apart from the usual signal in the center pixel two peripheral pixels detect a negative signal at the level of about 50 Horizontal scales LSR velocity km s and vertical scale antenna temperature K are identical for all pixels number in the upper right corner of each spectrum switching mode The aim was to combine the increased switching efficiency available with the array with the optimum baseline quality inherent in the standard vvobbler svvitched mode Our most successful schema is outlined in Fig 5 A linear sequenc
30. y concerned extragalactic observers are invited to contact Wiesemeyer wiesemey iramj fr about its current state Fur thermore it is planned to monitor the physical temperature of the mixers and use this information to re calibrate the receiver gain Frequency switched OTF is available with the same hardware limitations as with rasters sect 4 3 A high gt 1 Hz frequency switching rate is how ever more important here than with rasters Note that frequency switching OTF with VESPA is currently limited to a use of not more than 50 of VESPA This limitation is due to the limited readout speed of the current computer system 4 5 Ugly details Rasters are currently handled in a rather heterogeneous way The new procedure raster map pro introduced for HERA generates the desired posi tion stepping sequence of the telescope complete with flexibly interspersed calibrations But it cannot do more than one ON OFF Another proce dure obsraster map pro invoking the traditional OBS raster command does exactly that There are two examples given for obsraster map pro OBS RASTERMAP2 PRO and OBS RASTERMAP4 PRO for 2 resp 4 ON per OFF source integrations These procedures can be found as described in the footnote on pg 13 Frequency switching gives very good results with HERA The raster map procedures can be combined with frequency switch ing but the obsraster map should not be used as useless OFF source inte grations are done 4 6 Spec
31. y used frequencies The image rejection is about 10dB in LSB tuning mode for all pixels 0 1 and set globally for both modules at the present time A more precise rejection measurement using a line injection device is currently un derway and will ultimately allow to retrieve individual rejections for each pixel from a look up table at each frequency The limited accuracy of the knowledge of sideband rejection leads to relative calibration errors in the The present look up table of sky frequencies in GHz is 217 0 219 0 221 0 224 0 225 0 226 0 230 0 232 0 244 0 245 0 246 0 258 0 260 0 261 0 265 0 266 0 267 0 268 0 The fre quency you want to use should be within a range of 0 6 GHz range of 0 10 For high signal to noise mapping it is therefore recom mended to undertake some additional steps for relative calibration or flat fielding see Sect XX 3 Basic calibration observations This section describes the basic observations which serve to measure the telescope pointing and focus and to calibrate the antenna temperature scale 3 1 Pointing Pointing with the central pixel Default pointing scans are made with the central pixel of HERA1 and the usual command OBS gt point Pointing on the central HERA2 pixel can be made with the command OBS gt point 14 The HERA pixel pattern is shown shown in Fig 3 as it appears in the receiver cabin Nasmyth coordinations before rotation in the derotator as sembly s
Download Pdf Manuals
Related Search