Home

View/Open - University of Lethbridge

Contents

1. IRMA PWV E Select Date Read IRMA Start gt Range and gt datafor gt Bin Data J Unit numbers date range Separate Calibration and Observation Data PWV Conversion Unit ae E IRMA Apply coefficients to uni iier observation data to profile determine infrared atmospheric flux Convert temperature sensor values to flux using the Planck eqn and Atmospheric Load weather model data Use weather data and spectral power data to interpolate PWV values from site model Use pre determined lid coefficients to calculate effective infrared flux of the blackbody Use multi variable regression on calibration data to calculate coefficients for internal temperature sensors Plot save desired output Figure 4 11 A flowchart describing the steps involved in calculating the PWV from the raw data these programs were written by a previous student I was able to learn the code structure maintain and make bug fixes in the code I also improved the performance of the programs and added new functionality as needed With the development of a new calibration method new calibration software was written The new calibration program IRMA PWV was written in IDL because of its SECTION 4 4 SUMMARY 71 efficient handling of large arrays of data and built in graphical display functions Though IRMA P
2. Y Axis Min Y Axis Max As at zenith ie Data sec El W Add info balloons _ Use full resoution data W Linear Calibration method Bin graph over 0 seconds Show difference between datasets Overlay SMA phase data Print Graph Save Window Dump data to file Calibrate detector Calibrate temp diodes Display position map 3 Display axes relative to moon centre Subtract baseline fi ition E Sa lo Start UT date time 2008 01 10T15 38 12 End UT date time 2008 01 10716 3812 Last 10 mins Last Hour Display cooler data Overlay Humidity Overlay command history Overlay Pressure Overlay Temperature Figure 4 10 A screen shot of the IRMA Archive Interface GUI 4 2 1 Requirements and Execution As with the other Perl based programs which have been discussed the IRMA Archive Interface requires that additional Perl modules be installed as well as custom IRMA modules to be present The IRMA Archive Interface depends on many of the same Perl modules as the IRMA Control Interface a complete list of required modules and installation instructions are given in Appendix C 2 The Archive Interface is located in the same directory as the IRMA Control In terface IRMA and is executed with the viewirma pl command SECTION 4 3 IRMA PWV 69 4 3 IRMA PWV As will be discussed in 85 1 the IRMA units were once calibrated using a sim ple two point calibration method However it was disco
3. The Daily Tasks application within Auto Tasks may be managed through the Daily Tasks sub menu located under the Engineering menu There are two items under the Daily Tasks sub menu The first Edit Task allows the operator to create or edit task files without requiring any knowledge of the file format described in 3 5 1 3 The task files may then be loaded to the current unit through the second option Organize Tasks SECTION 4 1 IRMA CONTROL INTERFACE 65 Automatic Tasks Options micron15 Humidity Cooler Daily Tasks Sky Maps _ Enable Units 3 Enable _j Enable Units _j Enable Units Delay 30 scnd 3 Remove Detayfiz0 scnd 3 Remove I Delay a0 send i Remove Delay 30 scnd i Remove ee Level W Temperature E Run if Humidity Nknoven Kr ii KI LS ma J Osc Freq Comp Amp BS Ea 4 Run if Humidity Nknown i aad Apply Figure 4 8 A screenshot of the dialog box that the operator uses to configure Auto Tasks Units to Control 1 Remove y Start Up Functions performed automatically on start up _j Query Unit Status _ Load Data Plots Scanning sx Automatic Manual W Open Shutter Queue Server Information Server M Units Last 10 mins Automatic Functions performed p Change Unit automatically at IP Ico cl Set Point 170 K certain intervals E Query Unit Status Port MM SSH ture E Port fez Functions perfor
4. The MC AAC communication protocol uses simple handshaking The MC sends the serial communications packet to the AAC which then converts it from ASCII to binary and performs the function Once the task was complete the AAC responds with the same packet except with data fields populated if applicable 3 2 5 Alt Az Controller The AAC software exists as a bootable software image that resides in the flash memory of the Rabbit microcontroller Similar to the MC MicroC OS II real time kernel SECTION 3 2 RABBIT BASED IRMA 43 gt Response uc Figure 3 7 IRMA serial communications protocol between the Master Controller and the slave Alt Az Controller MC AAC Serial Metronome Comm Task Task Flag do job Flag servo tick Single Axis Job Task gt Move Task Long duration Dual Axis task dispatcher Move Task Flag move done Figure 3 8 A schematic of the IRMA Alt Az controller task structure was used to provide a real time multitasking environment which is required for the servo control loop The task structure of the AAC shown in Figure 3 8 is very similar to that of the MC The serial communication task analogous to the MC s dispatcher task waits for commands from the MC It is a real time task devoted to monitoring serial data traffic When it reads a STX character it proceeds to read up to 80 characters or when an ETX character is encountered If the packet is received without error t
5. calibration schema which was verified through test campaigns both in our laboratory in Lethbridge and on a remote mountain top in Chile In the second part of Chapter 5 a case study is presented demonstrating how the software was able to be reconfigured to acquire scientific data in the event of an unexpected hardware failure Chapter 6 concludes the thesis with the future directions of the IRMA project as well as some final thoughts 12 Chapter 2 IRMA Hardware 2 1 Overview The IRMA concept has evolved from a barebones prototype into a compact au tonomous remotely operated instrument During this evolution most of the original pro totype was redesigned The IRMA hardware can be divided into three main categories Optics Mechanical Design and Electronics A brief history along with the current design will be presented in this chapter 2 2 Optical System The IRMA optical system is the most important part of the unit since it enables the unit to observe the sky as well as calibration sources The prototype IRMA I shown in Figure 2 1 was developed by a previous graduate student Graeme Smith 1 The unit mounted on a three legged optical table and consisted of a plane scanning mirror which directed light to an off axis parabolic mirror which subsequently focused the light at the SECTION 2 2 OPTICAL SYSTEM 13 Figure 2 1 A photograph of the original IRMA instrument detector The scanning mirror provided a ra
6. conditions exist the unit will point up to zenith and resume normal operations Additional autonomous control is handled through Auto Tasks 83 5 3 3 3 AAC Updates Though the MC processor was replaced with a PC 104 the AAC remained Rabbit based Therefore much of the original AAC software remained intact However I identified several upgrades which made the AAC run more efficiently 3 3 3 1 Faster Initialisation During initialisation the previous version of the AAC would seek both the counter clockwise and the clockwise limits This made initialisation a time consuming task taking up to 10 minutes to complete Since the physical distance between the limits remains con SECTION 3 3 PC 104 BASED SYSTEM 50 stant once the unit is built it is only necessary to seek one limit and load the predetermined distance to the other limit Also the original software was only capable of initialising one axis at a time An additional function was written to allow for a dual axis initialisation which further improved the startup time 3 3 3 2 Software Initialisation In some cases such as after a power cycle the position of the Alt Az is known to the operator but the unit has not been initialised In these cases it would be desirable to be able to load the position directly without having to perform an initialization This additional functionality was added to the latest version of the AAC software 3 3 3 3 Uninitialised movement
7. or variables must be separated by commas and there must not be any whitespace between the quotes The reason for the prohibition on whitespace is that the IRM Ascript interpreter divides statements into their constituent parts tokens along whitespace divisions Two special literals can be used within printf strings the s symbol defines a single whitespace while the n symbol defines a linefeed and is often called a newline character Output can be directed to an open log file by including the log modifier imme diately after the print command The methods for defining the string format is identical to the standard print command The localhost command has methods to open and close logfiles Furthermore the assign command can be used to define strings that can be assembled using the print command A 3 7 System Commands The following group of commands are responsible for controlling and or reading data from IRMA s hardware components which includes the AAC NOTCH_FILTER The notch_filter state filter commands enable or disable the 60 Hz notch filter The filter is enabled with the 60hz_in parameter and disabled with the 60hz_out parameter Reading 60 Hz notch filter state can be done with the notch_filter read 60hz command A return value of 0 zero indicates that the filter is not enabled while a return value of 1 indicates that the filter is enabled BANDPASS FILTER SECTION A 3 IRMASCRIPT LANGUAGE DEFINITION 121 The ban
8. signal will result in the units being shown in the appropriate calibrated units C mbar humidity Double clicking on a small graph will cause this graph to exchange places with the graph currently in the large graph area SECTION 4 1 IRMA CONTROL INTERFACE 64 The data from the cryocooler can also be monitored using the IRMA Control In terface This is accomplished by selecting Display Cooler Stats under the Engineering menu which opens a new window and displays the cryocooler temperature and amplitude Time ranges are selected in the same manner as the main graphing windows 4 1 1 5 Message Window The message window is located at the bottom right of the GUI seen in Figure 4 1 It displays information such as connection attempts raw output from IRMAscripts and other debug messages 4 1 1 6 Managing AutoTasks As discussed in 3 5 Autotasks may be configured manually on the IRMA unit Autotasks can also be managed from within the IRMA Control Interface The options for the autoTasks conf file described in 83 5 1 5 can be set with the IRMA Control Inter face by selecting AutoTask Options under the Options menu Once selected the Control Interface connects to the unit and retrieves the current autoTask conf file and displays the information in the dialog box shown in Figure 4 8 At this point the options may be modified The new options are then written to the remote autoTask conf file when the operator selects Apply
9. value The longer the ADC integrates the analog signal the greater the accu racy or resolution of the digitized sample Table A 3 lists the different sample resolutions in terms of noise free resolution bits integration time in millisec onds and word rate Input span of digitization can be either unipolar where A D values contain values ranging from 0 to 24 1 or bipolar which allow signed values ranging from 2 to 2 1 Table A 4 lists constants and their respective numeric values that can be applied to the polarity field Table A 3 CS5534 ADC sample resolution settings in IRMAscript S5531 RES2L SLOW 21 O55534 INTEGG9 69 60 S5531 RESIT SLOW 17 0S55311NTEGIO 99 480 S5531 RESIT FAST 17 CS553LINTEG 6 57 0 adc read csr channel The contents of each of the CSR channels can be read using this command A colon delimited string having the following format is returned channel gain word rate polarity SECTION A 3 IRMASCRIPT LANGUAGE DEFINITION 128 Table A 4 CS5534 ADC polarity settings in IRMAscript Polarity Value CS5534_UNIPOLAR C95534 BIPOLAR 2 adc init rw_test Primarily used for troubleshooting and verification the read write test command tests the ADC to ensure that the IRMA software can communicate with it An arbitrary value is written to one of the CS5534 s offset registers then that value is read back from the offset register If the two values are identic
10. weather monitoring cryocooler logging daily tasks and skymaps Each of these tasks are described below 3 5 1 1 Weather Monitoring The IRMA unit must be able to operate over a wide variety of environmental con ditions The weather must be continually monitored If the weather becomes unfavourable as determined by the relative humidity the unit will close the weather shutter and move to a parked position The relative humidity threshold is set in the autoTasks conf file with the humidity humid parameter If an external weather station is available at the site a special function can be added to the IRMA Weather module to parse the output This information is then read at a specified interval If external weather sources are unavailable the unit can be configured to use its own on board humidity sensor 3 5 1 2 Cryocooler Logging The cryocooler must be observed constantly to ensure that it is maintaining a proper temperature It is also important to record the amplitude of the cooler drive as SECTION 3 5 AUTOTASKS 53 Connect to Queue Server daily_tasks delay Weather Cryo Cooler Monitoring Monitoring Daily Tasks Task Task Figure 3 10 A flowchart of the AutoTasks execution path this may lead to early detection of a vacuum leak It is possible to monitor the cooler temperature drive amplitude and oscillator frequency with the cooler monitoring function by setting their respec
11. 2 IRMA Archive Interface o 4 2 1 Requirements and Execution aa a AS IRMA PWV hs 6404 ib hd ba ee wk a dh d Rh de ee es Be E A n Ges i se Ge Ae ki ee i SE a or ae Instrument Calibration and Field Performance 5 1 Calibration s sr e porel 5 1 1 Calibration Basics ee 5 1 2 Calibration Issues 2 ee 5 1 3 Improved Calibration Method i i ela Rolis lt A aoc A E erage eae ey Sec eh ct oe 5 2 Field Performance 2945 46 4254 aa w fee Boe Beate we es sa e e D3 SUMMAT 2 5 4 4 ec 3 ek oh ea ee eS e aa eee ee ee Ee Future Directions 6 1 Overview aocor ordoni ena makk e RAA ue E ROE CA ENA AGOR ea E a ie aa A 6 2 Future Mechanical and Optical Development 6 2 2 New Cryocool t 1 2 a ea woe ERS ee eA wee re a 62 2 Larger Primary Mirror 3 2 44 6 0 5 Roe bon a OS Sak eae es 6 2 3 New Weather Shutter Design o e e 6 3 Future Software Development e 6 3 1 System Software Enhancements o 6 3 2 Front End Software Development 6 4 Final Thoughts sans parte eh bee eee ee a Ee a Se ew g vil 30 30 30 31 39 40 42 42 45 46 49 49 91 52 52 54 99 56 56 57 66 66 68 69 69 72 72 72 74 79 85 96 102 TABLE OF CONTENTS A IRMAscript A 1 Overview A 2 IRMAscript Language Summary o A 3 IRMAscript Language Definition A331 List Man
12. 2007 MCT infrared detector KMPC19 1 SP Kolmar Technologies Inc URL http www kolmartech com C Lee P A R Ade and C V Haynes Self Supporting Filters for Compact Focal Plane Designs ESA SP 388 p 81 1996 Robin R Phillips Vic Haynes David A Naylor and Peter Ade Simple method for antireflection coating ZnSe in the 20 um wavelength range Appl Opt 47 7 870 873 2008 URL http ao osa org abstract cfm URI a0 47 7 870 Rabbit Semiconductor Inc Davis CA RCM2100 RabbitCore Data Sheet 2005 URL www rabbitsemiconductor com products rcm2100 rcm2100 pdf WinSystems Inc 715 Stadium Drive Arlington Texas 76011 USA URL http www winsystems com Diamond Systems Corporation Emerald MM DIO Quad RS 232 48 Digital I O PC 104 Users Manual Newark CA 2002 URL www diamondsystems com Honeywell Hymatic Redditch Worcestershire UK URL http www hymatic co uk Rabbit Semiconductor Inc Davis CA RCM2000 RabbitCore Data Sheet 2005 URL www rabbitsemiconductor com products rcm2000 rcm2000 pdf US Digital Corporation Vancouver WA E6 Optical Kit Encoder 2005 URL www usdigital com data sheets E6 Data Sheet pdf US Digital Corporation Vancouver WA LS7266R1 Encoder to Microprocessor Inter face Chip 2004 URL www usdigital com products 1s7266 lan Sean Schofield The IRMA III Control and Communication System MSc thesis University of Lethbridge Lethbridge AB 2005 S
13. 43 3 8 A schematic of the IRMA Alt Az controller task structure 43 3 9 A Flowchart of PC 104 based system software 48 3 10 A flowchart of the AutoTasks execution path 53 LIST OF FIGURES X 4 1 4 2 4 3 4 4 4 5 4 6 4 7 4 8 4 9 4 10 4 11 5 1 5 2 5 3 5 4 5 9 5 6 5 7 5 8 5 9 5 10 5 11 5 12 5 13 5 14 5 15 5 16 5 17 5 18 5 19 5 20 5 21 A screen shot of the IRMA Control Interface GUI 57 A screenshot of the Queue Status area of the IRMA Control Interface showing the blackbody ON command running with the Refresh command waiting in the QU sisama eee a we eG Be eb os Ae aia e 58 A screenshot of the Current unit status area of the IRMA Control Interface showing the current status unit showing the chopper on the shutter moving and the blackbody of 0 02 02000002 eee eee 59 A screenshot of the dialog box that the operator uses to point IRMA 60 A flowchart showing the execution path of the IRMA Control Interface 61 A screenshot showing a zoomed region of the Unit Stat area of the GUI showing that unit 12 is connected while unit 3 is disconnected 62 A screenshot showing a zoomed region of the graphical display area of the IRMA Control Interface The right hand graph shows a calibration cycle followed by sky observation aooo a a 63 A screenshot of the dialog box that the operator uses to configure AutoTasks 6
14. 4815 36 45 2002 Blue Sky Transmittance and Radiance Atmospheric Model Product in formation from Blue Sky Spectroscopy Lethbridge AB Canada URL http www blueskyinc ca H J P Smith D J Dube M E Gardner S A Clough F X Kneizys and L S Rothman FASCODE Fast Atmospheric Signature Code Spectral Transmittance and Radiance Technical Report AFGL TR 78 0081 Air Force Geophysics Laboratory Hanscom AFB Massachusetts U S A 1978 REFERENCES 140 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 L S Rothman D Jacquemart A Barbe D Chris Benner M Birk L R Brown M R Carleer Jr C Chackerian K Chance L H Coudert V Dana V M Devi J M Flaud R R Gamache A Goldman J M Hartmann K W Jucks A G Maki J Y Mandin S T Massie J Orphal A Perrin C P Rinsland M A H Smith J Tennyson R N Tolchenov R A Toth J Vander Auwera P Varanasi and G Wagner The HI TRAN 2004 molecular spectroscopic database and Radiative Transfer 96 2 139 204 2005 Journal of Quantitative Spectroscopy D A Naylor R T Boreiko T A Clark R J Emery B Fitton and M F Kessler Atmospheric emission in the 20 micron window from Mauna Kea Publications of the Astronomical Society of the Pacific 96 167 173 1984 Richard Querel IRMA As A Site Testing Instrument MSc thesis University of Lethbridge Lethbridge AB
15. 5 1 3 3 Calibrated sky measurement When the IRMA units are operating at remote sites periodic calibrations are performed using the internal blackbody which acts as a secondary standard By applying SECTION 5 1 CALIBRATION 84 Figure 5 8 Three IRMA units on the roof above our laboratory at the University of Lethbridge The lower unit is fitted with a heating cable and insulating jacket cardboard to test the ability of the fitting routine to correctly account for and remove the systematic offset due to heating the coefficients determined in Equation 5 9 the effective flux seen at the detector can be calculated A new set of internal temperature coefficients are then calculated by applying a linear least squares fit similar to Equation 5 7 except replacing Pjpp with Perum Once these new coefficients are calculated the infrared flux from the atmosphere can be determined by equation V Vo D Ci Dd Bey W 5 10 CeffLID The flux associated with this temperature is found using equation 5 2 This flux value along with the local temperature and pressure can be converted to PWV using a pre calculated SECTION 5 1 CALIBRATION 85 2006 11 19UT 1 18 Calibration sequences T Se e gt O O jav 9 aA Atmospheric AA lt measurements b o O KACI oO AA ey D AA Ler Prfy fi a p mn atari O E Pe A A A e E A a i EEE DER GSE 2AF
16. Alt Az controller AAC is responsible for pointing the IRMA within its Alt Az mount The AAC a custom built electronics board designed around the Rabbit Semi conductor RCM2010 21 Figure 2 6 connects to the PC 104 over a serial connection through the umbilical cable It acts as a slave processor to the main computer and its primary function is motor control which consumes the majority of DIO lines on the AAC Rabbit SECTION 2 4 ELECTRONICS 24 5 p ANALOG 6 gt MUX u 7 gt PC 104 J m A gt MASTER E 10 lt MUX 1 DIO w COOLER Y LE gt wx2 BOARD E lt gt ANALOG OUT t E IR CONTROLLER g 4 MUX 3 RS232 a e DRIVER 2 ct Elala lol GPs 13 gt analog Y 14 p MUX Lx 2 a AS ke ms GPS TIMEMARK Log gz 6 gt DIFFERENTIAL TTL K 17 gt DRIVER aos 18 gt Tx S 19 _______ 2 ay tu z ANALOG OUT TO SLAVE RCM 2010 3 3 AAC RESET PAZ NA ly cia sol fe ken E ScLK pak as psh 500 POWER 3 LINES j SUN SHUTTER 1 ER SHUTTER READ ar ETHERNET 4 LINES 553442 ADC BANDPASS FILT ASSHZ ON OFF NOTCH FILT 60HZ CHOP ENAB BB SHUT OC LATCH RESET BB HEATER Be SHUTTER D Z INTOA tional p SHUTTER LIMI E Sl iri p SHUTTER Lime 58 CHOP
17. An Infrared Radiometer For Millimeter Astronomy MSc thesis University of Lethbridge Lethbridge AB 2000 Ian M Chapman and David A Naylor Fourier Transform Spectroscopy Hyperspectral Imaging and Sounding of the Environment In edited by editor Technical Digest CD paper HTuD2 OSA 2005 Infrared Windows Infrared Processing and Analysis Center URL http www ipac caltech edu Outreach Edu Windows irwindows html O P Lay MMA Memo 209 183 GHz Radiometric Phase Cor rection for the Millimeter Array ALMA Memos 1998 URL www alma nrao edu memos html memos abstracts abs209 html M C Wiedner and R E Hills Imaging at Radio through Submillimeter Wavelengths In Mangum J G and Radford S J E edited by editor 217 327 2000 S J E Radford and R A Chamberlin MMA Memo 334 Atmospheric Transparency at 225 GHz over Chajnantor Mauna Kea and the South Pole ALMA Memos 2000 URL www alma nrao edu memos html memos abstracts abs334 html J Duan M Bevis P Fang Y Bock S Chiswell S Businger C Rocken F Solheim T van Hove R Ware S McClusky T A Herring and R W King GPS Meteo rology Direct Estimation of the Absolute Value of Precipitable Water Journal of Applied Meteorology 35 830 838 1996 D A Naylor I M Chapman and B G Gom Proc SPIE Vol 4815 p 36 45 Atmospheric Radiation Measurements and Applications in Climate In Shaw J A edited by editor
18. IRMAscript name gt IRMAscript xls where lt box number is the assigned unit number for the IRMA and lt IRMAscript name gt is the name of the script to be executed This script must be located in the IRMA SCRIPTS directory The IRMAscript xls is shown in square brackets to show that it is optional It directs the interpreter to the Excel file which defines the formal language of IRMA script Unless the file is renamed it is not necessary to include it as a parameter since the interpreter will use the default filename As an example if an operator wanted to open the weather shutter on IRMA unit 1 and had written a script named shutterOpen irma he would copy the script into SECTION 3 2 RABBIT BASED IRMA 39 the IRMA SCRIPTS directory of the CP for IRMA Unit 1 He would then execute the following command from the base IRMA directory IRMA HelperProgs irmaExec pl 1 shutterOpen irma 3 2 2 CP MC Communication The binary network packets generated by irmaExec pl are sent over the network to the MC Due to the fact that the CP MC communication is over Ethernet transmission could not be guaranteed For this reason a communication structure based on the Euro pean Space Agency ESA Packet Telecommand Standard 27 was used This protocol is shown in figure 3 4 After the command is received the MC responds by sending an ac knowledgment packet a packet indicating the requested activity has begun a data packet if applic
19. PWV image showing the location of storm front The asterisks represent the raw IRMA skydip data while the smooth curve represents the modeled scaled curve of growth based upon the independent CSO tau data Figure 5 25 shows the corresponding data three hours later when the storm from was approaching the Mountain At this time the Teso 0 062 PWV 1 37 mm and the IRMA skydip deviates from the modeled curve of growth Generally there is seen to be good agreement between the IRMA measured and the CSO tau derived curves of growth when the storm front was far away but as the storm approached Mauna Kea there was an increasing discrepancy between the two data sets These differences can be attributed to the fact that CSO and IRMA radiometers view SECTION 5 2 FIELD PERFORMANCE 101 1 08 1 07 1 06 1 05 Signal V 1 04 A 1 03 XK 1 02 Airmass ans Figure 5 24 A GOES 10 satellite image of the Hawaiian Island chain taken at 05 00 UT April 24 2007 left The IRMA skydip asterisks and the modeled curve of growth derived from the independent CSO tau data taken at the same time right There seem to be good agreement between the IRMA and the CSO tau data 1 10 x x x 1 08 X ra a s x Gj L KT 1 06 m Sees n L ae JA S s 1 04 SN I EX 1 021 pepe qoj juj rey qa e o AAA Lay 1 2 3 4 GOES 10 WV 10 Airmass 1 cos igure 10 satellite image of the Hawaiian Island c
20. The following table lists all IRMA system commands addressable within the IR MAscript language Non system commands such as flow control commands are not listed PSTARTPROG SOCKET OPEN SSCS PENDPROG SOCKET OLOSE SSCS PORYO SAME PON owo SE On oyo SET MANUAL MODE owo o fer AUTO MODE JI oyo o eer STOPPED MODE O Oo ovo ea OOMPAMP SSCS oyo o ea seron S ovo ea o O SSCS yo kea ooe Ooo ooo y vo ea osem SSCS vo ewm Jon Ooo ooo S ovo mw ooe OOO SSCS s ea baere SSCS GPS hea oe OJ s eab iman S SCS s SERIAL OPEN GPS SERIAL OLOSE SSCS ADO ur REYNO SSCS ADO IN RES SSCS ADO IN AE dhan gam wordRate polarity SECTION A 2 IRMASCRIPT LANGUAGE SUMMARY 114 PAD kE CONFIGREGISTER SSCS SHUTTER STATE OPEN JI SHUTTER STATE OLOSE JI READ UMP S SHUTTER READ OVERCURRENT SSS SHUTTER SET OGRESET SSCS CHOP MOTOR_ STATE_ ON SSS SS CHOP NOTOR STATE or lt CSCS CHOP MOTOR STATE MEASURERPMON SSS CHOP MOTOR STATE MBASURERPN OFFJ SSS CHOPNOTOR READ STATES SSCS CHOP MOTOR_ RFAD__ RPM_ SSCS BB STATE JON TJ mo EV or A Bm READ SA HALT TI PADTAZ STATE JREBOOT Jo PALTAZ IN PNG PADTAZ IN A o oS PADTAZ r SR WO OJ PADTAZ IN MOTOR SSCS PALTAZ READ POSITION OJ TI PADTAZ READ TASESTATOS SS PALTAZ READ JAGrOFFSET SCS PADTAZ READ AZOFFSET SSCS PALTAZ READ POSLOG RANGE OOO O oo PALTAZ READ POSLOGDATA
21. The master controller software is not affected altaz move_to axis alt_d alt m alt_s az_d az_m az_s speed Servo controlled movements which track a theoretical velocity versus position profile are performed using the move_to command Three parameters must be provided the axis to be moved the destination angle and the axis rota tion speed specified in degrees per second Options available for axis include altitude azimuth and dualaxis The destination angle is defined in degree minute second format where altitude degrees minutes and seconds occupy fields 4 5 and 6 respectively assuming field 1 refers to the altaz symbol For single axis movement altitude or azimuth destinations should be written to fields 4 5 and 6 while fields 7 8 and 9 should be zero filled For dual axis movements altitude should occupy fields 4 5 and 6 and azimuth should occupy fields 7 8 and 9 Field 10 is populated with the desired axis speed In the case of dual axis movement the speed refers to the diagonal speed between the two moving axis or rather the speed required for both axes to meet at the final altitude azimuth coordinate Since this is a servo controlled move command movements are continuous and are consequently limited to the speed options provided by the given Alt Az mount s gearing The slowest speed possible with this command occurs when the axis motor is driven at 0 volts which corresponds to 500 motor RPM The axis will rota
22. as a running task altaz read alt_offset altaz read az_offset These two commands respectively return the currently defined altitude and az imuth offset values in optical encoder units There are 8192 units per revolution SECTION A 3 IRMASCRIPT LANGUAGE DEFINITION 133 altaz read poslog_state The poslog commands are used primarily for Alt Az servo tuning They allow the user to collect axis motion data necessary for tuning the AAC s PID servo control loop The read poslog_state command returns the current operation mode of the position log the table in the AAC that is used to store servo and position data Three states can be reported code 1 indicates the position log is enabled Code 0 indicates the position log is disabled Code 2 is returned if the position log was not initialized during AAC start up This can happen if an extended memory allocation failure occurred on board the AAC Rabbit processor altaz read poslog_range Calling this command returns the dimensions of the position log a memory array aboard the AAC containing position and servo data A four field colon delimited string is returned min array index max array index curr array index NULL The range of data readable from the AAC s position log is found between the minimum array index and the current array index inclusive Reading values beyond the maximum array index will result in a memory read error on the AAC altaz read poslog_data index Given some
23. catastrophic if for example the command in the queue is a weather protection command Future software development will likely see IRMAscript reduced from a full scripting lan guage with variables and flow control to a series of instrument control commands The flow control would then be handled by the program interfacing with the IRMA unit Reducing SECTION 6 4 FINAL THOUGHTS 110 IRMAscript in this a manner would not only reduce the complexity of the IRMA code but make it easier for an operator to write custom interfaces with the unit 6 3 2 Front End Software Development As discussed in 84 3 improvements in the calibration method required writing a new data reduction program IRMA PWV which was written in IDL and performed some of the functions of the IRMA Archive Interface 4 2 Future development of IRMA PWV will include adding a GUI interface as well as a complete suite of data reduction and viewing functions rendering the IRMA Archive Interface obsolete It is also foreseeable that IRMA control commands will be added as well which would incorporate the functionality of the IRMA Command Interface As a result all of the data reduction and control functions would be contained in one convenient package which could be run on any platform using the IDL Virtual Machine 6 4 Final Thoughts Since the development of the IRMA prototype in 1999 the IRMA project has been continually progressing To date the IRMA project has be
24. derived coefficients 92 Time series PWV data from three co located IRMA units in Chile based on sky calibration lt e sa cursis eRe RA PER ae Ce Ew a 94 Scatter plot of PWV data from three co located IRMA units in Chile based on Sky calibration s a 4 44 Sb dhe e T nt ih Boe eS a a wy 94 Time series PWV data from three co located IRMA units in Lethbridge based on sky calibration over poa 444545 cda 2s Pe bebe ds See es 95 Scatter plot of PWV data from three co located IRMA units in Lethbridge based on sky calibration 2 95 LIST OF FIGURES xi 5 22 Emission spectra at 1 mm PWV 0 200022 eee 99 5 23 Curve of growth profile corresponding to Figure 5 22 99 5 24 IRMA skydip data s seie ma deti a k a a eia ee 101 5 25 IRMA skydip data lt i da tp re aae Fon e aa a a do fb aai p a 101 6 1 New IRMA mechanical optical design i 0002 es 105 6 2 Block diagram of original Queue Server functionality 107 6 3 Block diagram of proposed Queue functionality 108 List of Tables 3 1 3 2 5 1 5 2 5 3 5 4 A 2 A 3 A 4 B 1 pseudo code description of the the IRMAscript interpreter An example unit configuration file which is used to initialise instrument pa rameters Correlation coefficients for Lethbridge test data Correlation coefficients for Chile test data i Correlation Coefficients for Chile test data based on sky calibration Corr
25. each unit it shows whether or not it is connected to the units associated Queue Server what the ping latency is for the unit and the status if the shutter cooler and scan Again green indicates on or connected yellow indicates unknown status and red indicates off or not connected Units may be added or removed through GUI options under the SECTION 4 1 IRMA CONTROL INTERFACE 63 1 April 2007 2007 04 01720 06 30 End 2007 04 01721 06 30 Hour 1 2Hour 10min Voltage Figure 4 7 A screenshot showing a zoomed region of the graphical display area of the IRMA Control Interface The right hand graph shows a calibration cycle followed by sky observation Options menu 4 1 1 4 Data Monitoring The Data Monitoring section is capable of graphing any data channel from an IRMA unit Channels are selected though a dialog box which is displayed by right clicking on the graph This is the same for the smaller graphs The time range is selected by entering the start and end times in ISO 8601 format 32 in their respective text boxes Conveniently time ranges can also be set by clicking on the Hour 1 2 Hour or 10 min buttons which will select the corresponding most recent time interval For example the last 10 min The graph will be displayed in the units selected from the drop down list Selecting Spectral Power or PWV for a channel other than channel 1 the detector
26. far clockwise limit In such cases the AAC rotates the azimuth axis in the opposite direction to the target angle lying beyond the far limit In the case of low speed small distance movements that result in destinations crossing the rotational limit the AAC drives the axis to the destination angle in the opposite direction at high speed in order to eliminate the annoyance of slewing nearly 360 degrees as low speed altaz read position This command returns a three value colon delimited string containing altitude and azimuth values respectively The third field contains scan status 1 when a scan is executing and 0 when no scan is running altitude azimuth scan_status read position is the most common query request to the AAC because during scans the MC requests axis positions for each data point collected altaz read task_status In order to remain responsive to incoming commands the AAC executes axis movements separate from the main dispatcher task read task_status allows external processes such as an executing IRMA script to check up on an ongoing AAC movement and determine when the operation has completed Task status is returned as one of three codes code 0 indicates there is no axis movement task operating while code 2 indicates a task is executing Code 1 is returned when the AAC is dispatching a long duration job to one of its available tasks It is rare that this code would be encountered and should be considered simply
27. forces the IRMA MC software to reboot The statement irma read uptime returns the number of elapsed seconds since the IRMA MC was powered up or last rebooted The value returned by this command is represented in floating point seconds SUN_SENSOR The solenoid controlled shutter protecting the filter and IR detector can be controlled by the software using the sun_sensor commands The state of the sun shutter is read using sun_sensor read shutter_state A return value of 0 indicates that the shutter is closed covering the filter and detector while a value of 1 indicates the shutter is open A photo cell coupled with discrete logic automatically closes the sun shutter when IRMA s line of sight comes within 15 of the sun or any bright light source is read using the command sun_sensor read state A return value of 1 indicates that the sun sensor is detecting a bright light source in its line of sight A value of zero indicates the opposite CRYO The Stirling engine cryo cooler that cools the IR detector is controlled by the cryo family SECTION A 3 IRMASCRIPT LANGUAGE DEFINITION 125 of commands Before attempting to send commands to the cyro cooler the serial com munication channel to the cooler must be opened with the command cryo serial open Likewise the channel is closed with the command cryo serial close cryo read comp_amp Returns the compressor amplitude value as a floating point value cryo read set_point Returns
28. in the interpreter SECTION 3 2 RABBIT BASED IRMA 34 Table 3 1 pseudo code description of the the IRMAscript interpreter initialize command code hash table open irmascript file read do read line from file split line into fields put into array put array in source code hash table with key line count increment line count until reach EOF initialize program counter pc to 0 do get statement from source code hash table using key pc pattern match statement on command modifieri and modifier2 look up command code using keys command modi and mod2 make command packet send command packet to MC according to network comm protocol while pc lt total lines in program A compiler translates the verified language statements into lower level machine readable codes This code may then be executed on the target machine at a later time By contrast interpreters execute the statements on the target machine immediately after they are parsed For irmaExec pl this involves either generating a binary packet to send over the network to the MC or in the case where a statement does not command the IRMA hardware such as a variable assignment or flow control the interpreter will execute the statement directly within the irmaExec pl process The basic functionality of the IRMAscript interpreter can be described by the pseudo code in Table 3 1 SECTION 3 2 RABBIT BASED IRMA CP MC AAC Start Start Start Initialize ar Comman
29. index value this command returns the position log entry at that index A four field colon delimited array is returned DAC val rel pos theor pos error val DAC val contains the 8 bit unsigned integer that is written to the AAC s DAC which in turn controls axis speed Rel pos refers to the actual position of the axis relative to its start position and is given in optical encoder units Theor pos is the calculated theoretical axis position also given in optical encoder units It is this theoretical displacement path that the PID servo must track The last field error val contains the PID algorithm error value All four data are necessary in the servo tuning process 134 Appendix B AltAz Commands Table B 1 contains a list of the command codes and their respective aliases used in communication between the IRMA Master Controller and the Alt Az Controller Code Alias 1 ALTAZ_READ_CURRENT_POSITION ALTAZ_MOVETO ALTAZ_HALT ALTAZ_PING ALTAZ_SET_ALT_OFFSET o ADTAZSUENV STATUS 10 ALTAZINIT E ALTAZ POSLOG STATE ALTAZ REBOOT ALTAZ READ ALT OFFSET ALTAZ READ AZ OFESET Table B 1 IRMA AAC command codes sent over MC AAC serial link 135 Appendix C Configuration and Installation C 1 CP configuration file options A complete list and description of parameters that can be defined in a CP config uration file is given below Data_Port The TCP IP socket port that is used by the MC to send s
30. large difference in unit temperature the calibration routine was able to successfully remove the systematic effects due to the heating The results are shown in Figure 5 12 SECTION 5 1 CALIBRATION 88 Flux for 2006 11 22 fags Ge Ae Ge Gage vices fe ive Pe ve UP ta MYA MERE rates ht arte Pe amy e os ln ay 9 0 a nh Y 8 5 E i E x i p j a x e A eA x sor a uf 5 p A 8 z Lo c oa A amp 7 56 ft t la Ag 7 0 x Unit 10 3 MES A Unit 12 pa Cad 0 1 2 3 Hours UT Flux for 2006 11 22 E O O eee E LEE a 9 0 sl 8 5 A T A 80 A xo f A ic i E 7 5 a 7 0 x Unit 10 L A Uni 3 E 0 1 2 3 Hours UT Figure 5 12 Flux values for two co located IRMA units in Lethbridge while unit 10 was heated 10 K above ambient temperature The upper plot does not account for the effective stray light The lower plot includes a correction for stray light based on the method described in the textwhile the bottom plot does SECTION 5 1 CALIBRATION 89 Figure 5 13 A photograph showing three IRMA units deployed on one of the Chilean sites being considered as part of the Thirty Meter Telescope TMT project with myself in the foreground Mr Greg Tompkins can be seen in the background checking one of the Units 5 1 4 1 Field Tests in Chile The sensitivity of IRMA to PWV measurements is a non linear function of the altitude of the obser
31. performance of the embedded system I ported irmaExec pl which was written in Perl and designed to run on a powerful desktop PC to C and then incorporated it into irmamc The C version of the interpreter followed the same design as described in section 3 2 1 1 The scanner was written as a separate function called readScript Receiving a string containing the name of the script to be tokenized the func tion read the script from the IRMA SCRIPTS directory and removed any comments and additional whitespaces and returned a two dimensional array where one dimension was the line number and the other was the individual tokens in the line This array was then passed to the parser runScript which is identical in functionality to the parser in irmaExec pl except that instead of generating a network command packet when a command was to be issued runScript simply executes the function to perform the desired task directly One difficulty in porting irmaExec pl into C was the handling of typeless vari ables As mentioned IRMAscript is a typeless language In the original implementation this was trivial since the underlying language Perl is also typeless To handle this in C additional checking had to be performed to determine the type of the variable The under lying variables were then stored as either strings or double precision floating point numbers which is similar to the method Perl uses for typeless variables Merging irmaExec pl into irmamc dr
32. sound there may be issues with the design of the large blackbody When the large blackbody is placed on top of an IRMA unit during calibration it completely encloses the unit As a result SECTION 5 1 CALIBRATION 93 Table 5 3 Correlation coefficients for the pairwise comparison of the Chile test data shown in Figure 5 18 using the calibration coefficients calculated using the Chilean sky relative to Unit 10 as the calibration source Data Set 1 Data Set 2 Correlation Coefficient Box 10 Box 11 0 9896 Box 10 Box 12 0 9869 Box 11 Box 12 0 9953 when it is heated to temperatures up to 90 C there is significant internal heating of the IRMA unit which may cause exaggerated contributions to the signal voltage from some areas of the box resulting in errors in the calculated coefficients In an attempt to calibrate the units under normal operating conditions we de cided to use the Chilean sky as a calibration source Since the actual atmospheric flux is unknown a relative calibration was performed where Unit 10 was selected as the reference and used to determine the radiated flux from the atmosphere The internal blackbody was then calibrated using the method described in 85 1 3 except the flux from the atmosphere as determined by Unit 10 was used in place of the large blackbody A set of lid coefficients was generated for each day of the testing campaign The average value for each of the coefficients was determined and used to pr
33. the calibration coefficients calculated using the Chilean sky relative to Unit 10 as the calibration source SECTION 5 1 CALIBRATION 95 j x Unit 10 q i A Unit 11 16 a Unit 12 7 a a 142 E on E F p ag 7 z g a an J 2 3 4 5 6 7 Hours UT Figure 5 20 A time series plot of the three co located IRMA units in Lethbridge The gaps in the data correspond to calibration sequences These data were derived using the calibration coefficients calculated using the Chilean sky relative to Unit 10 as the calibration source 18 16 14 PWV mm 12 x 10 vs 11 A 10 vs 12 10 011 vs 12 7 8 S 4 LA y 1 L L L L 1 1 L 1 L L L 8 10 12 14 16 18 PWV mm Figure 5 21 Inter comparison scatter plot of PWV data from the three co located IRMA units in Lethbridge analagous to Figure 5 11 These data were derived using the calibration coefficients calculated using the Chilean sky relative to Unit 10 as the calibration source SECTION 5 2 FIELD PERFORMANCE 96 Table 5 4 Correlation coefficients for the pairwise comparison of the Lethbridge test data shown in Figure 5 21 using the calibration coefficients calculated using the Chilean sky relative to Unit 10 as the calibration source Data Set 1 Data Set 2 Correlation Coefficient Box 10 Box 11 0 9766 Box 10 Box 12 0 9918 Box 11 Box 12 0 9935 5 2 Field Performance IRMA is
34. the scan task by means of an event flag to construct and transmit the data packet when a full frame of data are collected Once the scan task has successfully transmitted the data packet it returns to wait on the data transmit event flag shown as a dotted arrow in Figure 3 5 When data collection is terminated the scan task and metronome task are both instructed to terminate themselves SECTION 3 2 RABBIT BASED IRMA 42 3 2 4 MC AAC Communications If the dispatcher task of the MC receives an Alt Az command it forwards the com mand onto the slave AAC The MC communicates with the AAC over a 2 wire serial link operating at 19 2 kbit s which travels through the umbilical cable linking the IRMA to the Alt Az base While it was originally planned to use a similar communication protocol as the CP MC Communications this was not possible due to the limited bandwidth of the serial communications Therefore the communication protocol needed to be significantly simpli fied Serial packets were transmitted as a 6 field colon delimited ASCII character string consisting of a command field four data fields and a CRC checksum field encapsulated between STX start transmission and ETX end transmission characters An example of a serial communications packet string appears in Figure 3 6 The MC AAC Communication has remained the same for the PC 104 based IRMA Figure 3 6 The description of the IRMA serial communications packet string
35. 5 A screenshot of the dialog box that the operator uses to select GUI options A and Queue Server information B 65 A screen shot of the IRMA Archive Interface GUI 68 A flowchart describing the steps involved in calculating the PWV from the TOW Gata cos aoan ee as Gs koe i AA ee ee 70 Normalised IRMA instrument response function 74 Thermal image of an internal lid blackbody 75 Potential error in atmospheric flux introduced through two point extrapolation 76 Mapping the IRMA field of view 0 0 000000 0 ee ee 77 IRMA apertures sve inde oe ea bee Bed Ree bon hed a Bd 78 External view of the large reference blackbody 79 Typical LBB and lid calibration process run in the lab 81 Three IRMA units on the roof at the University of Lethbridge 84 Raw voltage data for three co located IRMA units in Lethbridge 85 PWV values for three co located IRMA units in Lethbridge 86 Scatter plot of PWV data from three co located IRMA units in Lethbridge 86 Flux comparisson with heated IRMA unit i 88 Three co located IRMA units in Chile i 89 Correlation of Chile data using original coefficients 90 Time series PWV data from three co located IRMA units in Chile 91 Scatter plot of PWV data from three co located IRMA units in Chile 91 Correlation of Lethbridge data using the statistically
36. CP computer for the issuance of commands If the network link between the RCM2100 in the IRMA and the CP were severed the unit would cease to function Furthermore in order to upgrade the software it was necessary for an operator to reprogram the microcontroller manually on site To address these issues in the SECTION 2 4 ELECTRONICS 22 latest version of IRMA the CP and the 8 bit microcontroller have been replaced by a 32 bit PC 104 inside the IRMA unt A significant part of my thesis has involved migrating and optimising the code to run efficiently on the PC 104 2 4 2 PC 104 At the heart of the electronic system Figure 2 5 is a Winsystems PCM SC520 PC 104 running an AMD 133 MHz SC520 processo 18 PC 104s are true IBM PC com patible computers capable of running a standard desktop operating system The PC 104 is responsible for the communication control and data acquisition of the IRMA unit The PCM SC520 is robust and designed for extreme conditions and temperatures of 40 C to 85 C It contains an onboard Intel 82551ER 10 100 Ethernet controller and an onboard Compact Flash socket which is used for data storage 2 4 3 Diamond Systems Emerald MM DIO The PCM SC520 lacks sufficient I O lines to be able to control the IRMA unit A Diamond Systems Emerald MM DIO 19 connects through the 16 bit PC 104 bus and provides 48 bi directional digital I O DIO lines as well as four serial ports The Emerald MM DIO is also r
37. E O E HSR 2 3 4 5 6 8 Hours UT Figure 5 9 Raw data for the three co located IRMA units operating in Lethbridge shown in Figure 5 8 Inter spersed with measurements of the atmosphere are three calibration sequences atmospheric model as discussed in 1 3 2 5 1 4 Results To verify the performance of the improved calibration method three IRMA units were calibrated independently using the procedure described in 5 1 3 These units were then placed side by side on the roof of University Hall at the University of Lethbridge SECTION 5 1 CALIBRATION 86 14 TTT TTT TTT TTT TT TTT TTT TTT TTT TTT 13 ran gmo x Unit 10 A Unit 11 o Unit 12 12 o PWV mm parra rl rra rr rl rr rr rol proto 3 4 5 6 7 Hours UT Birrriiiii Figure 5 10 Derived PWV values for the three co located IRMA units in Lethbridge shown in Figure 5 8 VQ TTT ya aa 13 12 11 10 PWV mm j x 10 vs 11 A 10 vs 12 e O11 vs 12 x 7 8 9 10 11 12 13 PVVV mm mb E Figure 5 11 Inter comparison scatter plot of PWV data from three co located IRMA units in Lethbridge showing a pairwise comparison The solid line is the ideal unity slope reference line while dashed lines are the 10 tolerance limits SECTION 5 1 CALIBRATION 87 Table 5 1 Correlation coefficients for the pairwise comparison of
38. Figure 5 16 Inter comparison scatter plot of PWV data from the three co located IRMA units in Chile analagous to Figure 5 11 These data were derived using the calibration coefficients calculated using the statistical approach described in 5 1 3 1 SECTION 5 1 CALIBRATION 92 Table 5 2 Correlation coefficients for the pairwise comparison of the Chile test data shown in Figure 5 16 using the calibration coefficients calculated using the statistical approach described in 85 1 3 1 Data Set 1 Data Set 2 Correlation Coefficient Box 10 Box 11 0 9836 Box 10 Box 12 0 9939 Box 11 Box 12 0 9886 PWV mm x 10 vs 11 J A 10 vs 12 J a 11 vs 12 10 12 14 16 18 PWV mm Figure 5 17 Inter comparison scatter plot of PWV data from the three co located IRMA units in Lethbridge LA 8 compare to Figure 5 11 These data were derived using the calibration coefficients calculated using the statistical approach described in 5 1 3 1 It can also be seen that the systematic errors are outside our error budget target of 10 for PWV However when these new calibration coefficients were applied to the Lethbridge data an offset was present as seen in Figure 5 17 The difficulty in obtaining a single set of coefficients which would adequate cali brate both the Lethbridge and Chilean testing campaigns raised questions about the accu racy of the primary calibration method While the method is fundamentally
39. HDesigns Download Managers SHDesigns URL http shdesigns org rabbit download shtml REFERENCES 141 26 27 28 29 30 31 32 33 34 35 36 Larry Wall Tom Christiansen and Jon Orwant Programming Perl 3rd edition O Reilly Media Inc Sebastopol CA U S A 2000 European Space Agency Packet Telecommand Standard PSS 04 017 Issue 2 1992 Z World Inc Davis California MicroC OS II User s Manual 2003 URL www zworld com documentation docs manuals DC DCModules ModUcos pdf Z World Inc Davis California Dynamic C User s Manual 2004 URL www zworld com documentation docs manuals DC DCUserManual DCPUM pdf crontab The Open Group Base Specifications Issue 6 IEEE Std 1003 1 2004 URL http www opengroup org onlinepubs 009695399 utilities crontab html Andrew Tridgell Rsync remote file syncronization system URL http rsync samba org International Organization for Standardization ISO 8601 2004 Data elements and interchange formats Information interchange Representation of dates and times International Organization for Standardization Geneva Switzerland 2004 URL http www iso ch cate d26780 htm1 Ian Myles Chapman The Atmosphere Above Mauna Kea At Mid Infrared Wavelengths MSc thesis University of Lethbridge Lethbridge AB 2003 Lake Shore Cryotronics Inc 575 McCorkle Blvd Westerville Ohio 43082 U S A URL http www lakeshore
40. Lethbridge test data shown in Figure 5 11 Data Set 1 Data Set 2 Correlation Coefficient Box 10 Box 11 0 9768 Box 10 Box 12 0 9774 Box 11 Box 12 0 9652 directly above our laboratory as shown in Figure 5 8 and allowed to view the same atmo sphere The raw unprocessed voltage data from each of the units is shown in Figure 5 9 From this figure it can be noted that each unit has different offset and gain characteristics which are due mainly to differences in the gain and offset of the pre amplifier of each unit However despite these differences once the initial calibration coefficients were applied to the data the resulting PWV measurements were well correlated as can be seen in Fig ure 5 10 To better illustrate the correlation resulting PWV from each unit was plotted against the PWV from the other two Figure 5 11 The near unity linear relationship with correlation coefficients listed in Table 5 1 demonstrated that the calibration method was successful and the units were ready for testing under more optimal conditions Once in the field the units are subjected to a wide variety of environmental con ditions To demonstrate that the IRMA calibration was independent of environmental conditions one on the boxes was fitted with a heater while the units were observing on the roof in Lethbridge the lower unit in Figure 5 8 The heater caused the internal temperature of IRMA to rise 10 K above the ambient operating conditions Despite the
41. MASCRIPT LANGUAGE DEFINITION 129 before executing the scan command Additionally it is vital that the Alt Az channel be left open for the duration of the scan Closing the channel during a scan will lead to scan failure which results in the scan terminating itself scan read status Returns the value 1 if a scan is currently executing on the MC otherwise the value 0 is returned scan signal on_int Forks the data collection process task where the IR signal is sampled on the positive edge of the notch notch interrupt signal Temperature pressure and humidity channels are each sampled following one IR signal sample in a round robin fashion scan signal no_int IR signal temperature pressure and humidity are sampled but the IR signal A D conversion is not synchronized to the notch interrupt ALTAZ Movement and control of the altitude and azimuth axes is handled by the altaz family of commands Given that the AAC is connected to the MC over a serial communications link the AAC MC serial connection must be opened before any altaz command can be sent Failing to open the serial port when sending AAC commands produces subtle errors that are hard to track down Each of the altaz command groups will be examined in detail Commands destined for the AAC are sent over the MC AAC serial link using the serial packet communications protocol altaz serial open altaz serial close These commands respectively open and close the serial chan
42. Master Controller IRMA Control Interpreter Interface Queue Layer Execution Command Line Figure 6 3 Block diagram of proposed Queue functionality are queued on the TCP stack However there are some features of the Queue Server not related to conflict resolution which would still be desirable on the system These include the ability to view scripts waiting to execute removing items from the queue and modifying the priority of a queued item Rather than simply continue to use the current implementation the Queue Server should be integrated into irmamc Currently the Queue Server simply acts as a take a number box yet has no direct interaction with the IRMAscript interpreter as shown in Figure 6 2 This adds complexity to the programs which use the Queue Server such as AutoTasks and the IRMA Control Interface because they must include the communication protocols for both the Queue Server as well as the IRMAscript interpreter Additionally the current Queue Server implementation does not allow for scripts to be added through the command line and therefore these scripts are not visible from the Queue Server If the Queue Server is integrated into the Master Controller all executed scripts regardless of how they are executed would be inserted into the queue shown in Figure 6 3 Unlike the original Queue Server the queue functionality will be completely abstracted from SECTION 6 3 FUTURE SOFTWARE DEVELOPMENT 109 the u
43. O Oo PADTAZ READ POSLOGSTATE SS PADTAZ SERIAL OPEN Jo PALTAZ SERIAL OLOSE SSCS perc E DES OOo TO READ DATE SCS son SIGNAL JONINT ooo o oyo son SIGNAL STOP O SSCS SECTION A 3 IRMASCRIPT LANGUAGE DEFINITION 115 Bon READ STATE OJ IRMA emae or TJ mma eab oe Jo SUN SENSOR JREAD same o o SUNSENSOR READ SHUTTERSTATE O ooo SUN SENSOR STATE SHUTTER OPEN SUN SENSOR STATE SHUTTER OLOSE SCS sorn psam fonw SS NOTOHFETER STATE onzo SCS NOTOHFETER SAE oo SSCS NOTOHFETER STATE oz o0T oO ooo NOTOR FILT R READ omz S CSCS NOTOHFILTER READ eomz Jo PBANDPASS FILT R STATE IN BANDPASS FILTER STATE jovr Jo BANDPASS FILT R READ STADE SSCS A 3 IRMAscript Language Definition A 3 1 List Manipulation INITIALIZATION Construct initialize a list with one or more elements LENGTH Return the length of a list INDEX Reference an element of a list where the index ranges from 0 the first element to n SUBSTRING Retrieve a substring from a colon delimited data record In IRMA commands that return multiple data items such as ALTAZ INIT PING data is returned as a colon delimited string This command splits the data string into its constituent data items and returns the desired SECTION A 3 IRMASCRIPT LANGUAGE DEFINITION 116 datum based on an index value A 3 2 Utility Functions DEG2DMS Convert an Alt Az coordinate
44. OL INTERFACE IRMA Control Interface Command Received Generate eine IRMA PC 104 Queue Server Wait for Queue Items Add command to Queue Wait for Notify Queuing notification that proram when command is command is ready ji Execute ready L a EJ command via SSH Parse Output inl irmaExec pl Update Status GUI Ethernet components Figure 4 5 A flowchart showing the execution path of the IRMA Control Interface SECTION 4 1 IRMA CONTROL INTERFACE 62 Figure 4 6 A screenshot showing a zoomed region of the Unit Stat area of the GUI showing that unit 12 is connected while unit 3 is disconnected shown in Figure 4 5 Immediately to the right of the control buttons the status of the current unit is displayed Each aspect of the IRMA unit is listed with a corresponding status indicator light The indicator lights will be green for on or open yellow for unknown status and red for off or closed The status will be updated automatically at regular intervals if this option is selected in GUI options under the Options menu Otherwise the status can be updated manually at any time 4 1 1 3 Unit Status The Unit Status is located below the Queue Window Figure 4 6 It provides an at a glance overview of all the units being monitored by the Control Interface For
45. PER p BB MOTOR OVRCURNT Y NETWORK LINE 4 LINES BLACK BODY El S SHUTTER ASSEMBLY LEN NOTCH FILER oe INTERRUPT LINE TIL LOGIC LINE Figure 2 5 A Block Diagram of the new configuration of the IRMA Master Controller showing how the PC 104 interfaces with the various subcomponents of the radiometer 2 4 7 1 Rabbit RCM2010 The RCM2010 controller module is the control computer that handles motion control and communication for the AAC It is fitted onto the main Alt Az electronics board through two 2x20 pin dual in line headers These headers allow the DIO of the RCM2010 to interface with the motor controllers limit switches and optical encoders A block diagram of these connections is shown in Figure 2 7 SECTION 2 4 ELECTRONICS 25 fo po gu E 117 E Ea 5 sa Sa ma Sa ne Sa Sa Sa ne bam ori Figure 2 6 A photograph Alt Az electronics board with Rabbit RCM 2010 microprocessor attached centre left 2 4 7 2 Maxon Motor Controllers As seen in Figure 2 7 the RCM2010 does not connect directly to the Alt Az motors Instead a Maxon 1QEC50V digital motor control unit is used for each motor Using the control units allows for digital control of the 24 V motors Altitude and azimuth motor controller enable lines are mapped to output lines 6 and 7 respectively on parallel port B Motor controllers are enabled by setting these lines while clearing these lines disables the controllers A braki
46. R detector and filter design The incident infrared radiation is focused by the parabolic mirror onto a Mercury Cadmium Telluride MCT photoconductive detector supplied by Kolmar Technologies 14 which is cooled to 77 K Since the development of IRMA I improved detectors have SECTION 2 3 MECHANICAL DESIGN 16 been obtained which have better detectivity and responsivity helping to increases the observed signal to noise ratio The resistance of the detector changes as a function of infrared radiation which falls upon it When biased with a constant current source this resistance change is detected as a change in voltage across the detector This voltage signal can then be amplified and digitized by an analog to digital converter While initially cooled with liquid nitrogen to allow for remote operation the IRMA detector is now cooled by a closed cycle Stirling cooler In order to restrict the incoming radiation to the desired spectral region 450 500 cm a bandpass filter is placed in front of the detector The second generation IRMA II included a filter provided by Professor Peter Ade of Cardiff University Wales which uses resonant capacitive and inductive micro elements 15 These filters provide superior performance over commercially available filters 2 2 3 Chopper wheel A five blade reflective chopper modulates the incoming optical beam at 450 Hz The detector alternates between views of the radiation from the sky an
47. TMT units As well as the entire TMT site selection team Karim Ali Scott Jones Dan Sirbu and Amy Smedes for the summers they worked on IRMA Locke Spencer for always being willing to help out with ATEX or IDL Thanks to Drs Peter Ade and Carole Tucker University of Cardiff for supplying us with specially designed IR filters Thanks to Fluke for loaning us the Ti20 Thermal Imager Thanks also to NSERC HIA and NRC for funding the IRMA project Table of Contents Abstract Acknowledgements List of Figures List of Tables CHAPTERS 1 Introduction IA aii auia Seats Gee ho See es oe ok gate ee eee es a 1 2 Precipitable water vapour 2 1 2 1 Importance of Measuring PWV o 1 2 2 Measuring Water Vapour 0 0000 eee eee 1 3 The IRMA Solution sp p ua 24 dv Dep d We ete ME Eee ale 13 1 The Instrument saoe sari a edi h he Pe kee Re ae 1 3 2 The Atmospheric Model BTRAM LA Summary 2 sed Ge be PAG ERE ee Ee Se ee PS ee ees 2 IRMA Hardware 2 1 OVERVIEW doe we wee Reh Oe ee hee Bee eR eae Dee 2 2 Optical System eee aa 20a ew ee ee Go a Bie Ee e ee ee 2 2 1 Parabolic Mirror Design o se c corem eacee seu ua ee 0004 2 2 2 IR detector and filter design 000 2 2 3 Chopper Wheel a o a ee ee ee 2 3 Mechanical Design sera de E d ndi ae do je e o 2 3 1 IRMA Module kakao m kone r e ee 2 3 2 Alazi mount es ka mgee 6 eds Bae ee Ee ee a es 2 4 Electro
48. This thesis was typeset with IXIEX using a modified version of the University of California PhD dissertation class file ucthesis cls Unless otherwise noted all figures in this thesis were created by the author using IDLR CorelDraw or the GNU Image Manipulation Program GIMP OPTIMISATION OF THE INSTRUMENTAL PERFORMANCE OF IRMA Regan Eugene Dahl B Sc Computer Engineering University of Alberta 2004 A thesis submitted to the School of Graduate Studies of the University of Lethbridge in partial fulfilment of the requirements of the degree MASTER OF SCIENCE Department of Physics University of Lethbridge Lethbridge Alberta Canada 6 Regan Eugene Dahl 2008 To Jillian and David for your patience and support during my Master s Program iv Abstract The Infrared Radiometer for Millimetre Astronomy IRMA is a passive atmo spheric water vapour monitor developed at the University of Lethbridge It is a compact robust and autonomous instrument which is capable of being operated remotely The latest model is based on a PC 104 running an AMD 133 MHz SC520 processor which allows for more flexible control of the unit The modifications and upgrades to the software required for the transition to this new platform are discussed in this thesis In addition to software optimisation a new calibration method has been developed as the unit has become better understood This method has been verified through test campaigns carrie
49. WY is currently only used for conversion of IRMA signal voltage to PWV future plans are to expand its development into a full suite of IRMA control and data processing function thereby replacing both the IRMA Control and Archive Interfaces 72 Chapter 5 Instrument Calibration and Field Performance 5 1 Calibration IRMA detects the radiated infrared flux from the atmosphere and by use of an at mospheric model this value is converted to precipitable water vapour PWV The accuracy of the final measurement is determined by sources of error in the instrumental measurement and the accuracy of the model This section addresses the instrument calibration For more details on the atmospheric model see 2 13 and 33 5 1 1 Calibration Basics To first order the response of the IRMA infrared detector is linear with respect to infrared flux A flux to voltage relationship is established by measuring the signal voltage while observing two calibration sources at different temperatures This can subsequently SECTION 5 1 CALIBRATION 73 be used to convert any signal voltage to a measured flux The calibration method described in this section was adopted for the prototype IRMA I Although it still forms the basis of the calibration used today in the most recent IRMA design which is an enclosed system a more complex calibration procedure is required Improvements to this calibration method are discussed in 85 1 3 In the bas
50. able and finally a packet indicating the requested activity has concluded For the in Command 00006 Ack Function Begin Data Function End 1 A Figure 3 4 A summary of the IRMA network communications handshaking sequence SECTION 3 2 RABBIT BASED IRMA 40 terpreter this funtionality was encapsulated in a separate Perl module IRMA PacketComm 3 2 3 Master Controller The Master Controller MC is a real time multitasking program using the MicroC OS II real time kernel In this type of program high priority tasks perform their tasks in short bursts then sleep for a defined interval or block on an event in order to open up time in which the next lower priority tasks can execute The order of execution continues down the priority hierarchy until all the tasks have completed Task priority was assigned according to the degree to which a task can tolerate being preempted In priority based preemptive multi tasking tasks can only preempt other tasks having lower priorities than themselves It is important to determine which tasks can be preempted and for how long giving the most critical task the highest priority All tasks are subject to preemption if an interrupt service routine ISR is present This should not pose a problem however since ISRs should execute and return to the interrupted process as fast as possible The task structure of the MC is shown in Figure 3 5 The dispatcher task is the most act
51. al the test is deemed a success and a value of 1 is returned A failed read write test returns a 0 zero This command should be followed with an ADC system reset in order to clear the dummy value in the offset register adc sample sample_type channel This command initiates an A D sample on a given ADC channel The re turned value is given in ADC units thus it must be interpreted according to the channel s polarity setting bipolar or unipolar Two types of samples can be taken those synchronized to the 450 Hz chop wheel specified with the on_int parameter or samples not synchronized to the chop wheel specified with the no_int parameter When the on_int parameter is specified the A D sample commences when the chop wheel signal reports a logic level of 1 The channel parameter is mapped to 19 separate channels only 11 channels are present in the older Rabbit based implementation SCAN The scanning process involves repeatedly sampling the IR signal and temperature pressure humidity channels at some interval Scanning differs from reading an ADC channel directly in that the A D sampling process is contained separate real time task and uploads the data to a separate network port on the IRMA CP This allows the MC to service other commands while the data collection process is executing such as moving the Alt Az mount or querying the status of the cryo cooler The Alt Az serial communications channel must be opened SECTION A 3 IR
52. alculate motor speeds or slewing times Therefore it is vital that init motor be called before any axis movement is attempted altaz init servo The AAC uses a servo loop based on proportional integral derivative PID motion control algorithm to control axis movement The PID servo control algorithm has three constants P I and D which are unique to each Alt Az mount The init servo command loads the PID constants for the altitude and azimuth axes into the AAC If the servo controlled movement move_to command is going to be used servo parameters must be loaded into the AAC beforehand The slew_to non servo movement command does not require servo parameters to be set The command init servo may be called while the AAC is idle not moving as many times as required which is particularly helpful if the user is tuning the servo algorithm altaz init ping The Alt Az ping command is used to check if the AAC is on line ready to receive commands If the AAC is alive and on line it returns a three field colon delimited string of the form 987654321 123456789 uptime If the AAC is not on line or unresponsive the three fields will contain the code 999999999 The Uptime field indicates the number of elapsed CPU ticks since the IRMA MC was booted Each CPU tick is 1 64 seconds therefore to convert this value to elapsed seconds divide it by 64 Uptime can also be read using the command altaz read uptime altaz set alt_offset offset_
53. amatically reduced the complexity of both programs as the network packet handling functions could be removed Also since the Perl interpreter was no longer need the compiled C code executed much more quickly As a result of these improvements scripts now start executing over an order of magnitude faster than the previous configuration which brings an important real time response to the Operator s SECTION 3 3 PC 104 BASED SYSTEM 48 Console Moreover a reduction of 11 200 lines of code of the original 68 000 lines of code was achieved Figure 3 9 A Flowchart of PC 104 based system software SECTION 3 3 PC 104 BASED SYSTEM 49 Scripts are still executed in the same form as described in 3 2 1 3 However the new irmaExec pl simply passes the script name to irmamc and outputs the returned results 3 3 2 Autonomous Control Running embedded Linux allows for increased autonomous control of the IRMA When IRMA is operating at remote sites it is important for it to be able to recover from power failures without intervention Through the startup file etc rc d rc local the system is initialised First it verifies if the Alt Az mount is initialised and if not it performs the initialisation For the case where the PC 104 is simply rebooted the AAC will not be affected and therefore will remain initialised After initialising the mount the unit moves to the Park position while it checks the current weather conditions If favourable
54. ar in the arm of the mount which is fixed to the main IRMA box and allows rotation of 185 To prevent rotation beyond these limits and prevent the main IRMA module from contacting the Alt Az mount optical limit switches are again used The motors are disabled when the switch is interrupted by one of two metal tabs attached to the rotating housing Limit detection is independent of software in order to eliminate the risk of runaway axis movement damaging the mount if the software were to fail Attached to each of the axis are US Digital E6M optical encoders which provide positional data to a resolution of 8192 360 counts degree as discussed in 2 4 7 3 2 4 Electronics 2 4 1 Overview The IRMA electronics have evolved significantly since the original prototype The original system was very basic being controlled through the parallel port of a laptop com SECTION 2 4 ELECTRONICS 21 Figure 2 4 A photograph of the electronics side of the IRMA module The optical compartment is located immediately on the other side of the vertical wall shown in the image puter This system was first upgraded to allow commands to be sent to the laptop over the internet for remote operation With the new mechanical design the electronics were integrated into the IRMA unit This system was based on an 8 bit Rabbit Semiconductor RCM2100 microcontroller 17 which could be controlled remotely over the internet but still required a command processor
55. ask When a task is completed successfully the AAC will respond with the same packet requesting the task except with the command field populated with the MSCOMM_SUCCESS code integer value 100 and where applicable data is returned in the four data fields and a checksum value is calculated and placed in field 6 If an error occurs in the communications the command packet is populated with the MSCOMM_FAILURE code integer value 101 and each of the four data fields are populated with the associated error codes 3 3 PC 104 based system Though the Rabbit based system worked well it was limited by its requirement of the CP If the network connection between the CP and the MC was lost the MC would be unable to receive commands and would not be operational To overcome this problem the 8 bit Rabbit microcontroller of the MC was replaced by a 32 bit PC 104 The PC 104 is a true IBM PC compatible computer capable of running a standard desktop operating system IRMA uses the Slackware 10 2 Linux distribution With the increased computing SECTION 3 3 PC 104 BASED SYSTEM 46 power of the PC 104 a separate CP was no longer necessary and was merged with the MC 3 3 1 Porting the Master Controller The first step to running IRMA with a PC 104 was to rewrite the MC for execution on the PC 104 Ian Schofield started the task of porting the original code written in Dynamic CQ 29 a proprietary C variant by Rabbit Semiconductor to ANSI C Th
56. ask that the axis rotation has completed The metronome and move axis tasks then suspend themselves waiting on another event flag to signal a new request It was necessary to introduce a second mode of axis movement to handle extremely slow movement that is movement slower than 500 RPM which is the minimum speed which can be selected with the motor controller 2 4 7 2 This second movement mode is referred to as slewing meaning long range motion around the altitude or azimuth axis Slews have the ability to perform periodic steps over a long period of time thus lengthen the time to rotate from the initial to destination angle The serial communication task signals the SECTION 3 3 PC 104 BASED SYSTEM 45 job task to wake up and start the appropriate axes control tasks Each axis movement is controlled in its own task one exists for altitude movement and another for azimuth Both axis tasks can be run concurrently which requires that they each have unique priority levels Since both tasks cannot have the same priority the slew elevation task has a slightly higher priority than the azimuth slew task Given that skydip operations discussed in 85 2 which involve slewing the altitude axis are performed more often than azimuth movements preference was given to elevation movements Slew tasks and the servo move tasks have priority levels that place them below the metronome task priority but above the serial communications task and the job t
57. asurements we could still observe the atmosphere and so during this time a sequence of skydips were performed Skydipping consists of taking measurements at increasing zenith angles and in verting the measurements to retrieve the column abundance of water vapour In essence a skydip is a method of increasing the number of observed water molecules in a controlled manner Assuming a homogeneous parallel plane atmosphere with an effective height h SECTION 5 2 FIELD PERFORMANCE 98 the effective path length along the line of sight at the zenith angle 0 is given by l m 5 11 Airmass is the path length relative to the height of the atmosphere at zenith which under the same assumptions can be approximated as 1 j 0 5 12 airmass 5 sec 5 12 For example at a zenith angle of 60 the airmass 2 therefore IRMA is observing the emission of twice the amount of water molecules As the airmass increases so should the observed flux in a well defined manner For a single spectral water vapour line the peak emission increases with the num ber of molecules until the peak reaches the Planck envelope At this point the increased emission is due solely to the broadening of the line IRMA observes many lines in its spectral window 450 575 cm 1 a portion of which can be seen in Figure 5 22 plotted at vari ous airmasses From this Figure it can be seen that the amplitude of the emission spectra increases in a complex
58. at was replaced with a Stirling cycle cooler which was able to achieve liquid SECTION 2 3 MECHANICAL DESIGN 18 nitrogen temperatures without the use of cryogens 2 3 1 IRMA Module The main IRMA module can be divided into four main compartments optical electronics weather shutter and power supplies The module is designed to require only mains power and an ethernet connection The mains power is routed to the bottom of the module where the power supplies provide 24 V and 5 V DC to the unit The 24 volt supply powers the Stirling cycle cooler the lid motor and the temperature controlled calibration source while the 5 volt supply powers the IRMA computer and the other electronics Also mounted in the bottom compartment is the Stirling cooler controller which regulates the temperature of the cold finger of the cooler On one side of the module is the optical compartment which is visible in Figure 2 3 This occupies the majority of space in the IRMA module The optical system is discussed in section 2 2 The IR flux from the atmosphere enters the unit through a 117 mm diameter aperture which is directly above the parabolic mirror The 90 off axis parabola directs radiation to the detector which is mounted horizontally to the end of the cold finger of a Hymatic NAX025 01 Stirling Cycle Cooler A vacuum chamber surrounds the cold finger which is evacuated to 1 x 107 mbar This vacuum is designed to last five years and must have a leak rate
59. ated for operating temperatures of 40 C to 85 C 2 4 4 Pre Amplifier The measured signal from the MCT detector is very small having a noise floor of 10 nV and first needs to be amplified in order to make efficient use of the full range of the ADC The detector signal is connected to a low noise pre amplifier with a gain of 10 000 before passing to the IRMA motherboard SECTION 2 4 ELECTRONICS 23 2 4 5 IRMA motherboard The IRMA motherboard is a custom electronics board designed and built in house in our electronics fabrication laboratory It contains most of the electronics required to control the IRMA unit The PCM SC520 interfaces with the IRMA motherboard though the Emerald MM DIO board The 48 digital I O lines are divided between two connectors JP5 and JP6 each containing 24 I O lines which plug into the IRMA motherboard In order to simplify the configuration JP5 was used for input and JP6 was used for output The motherboard contains the electronics for the data acquisition system including signal conditioning lock in amplifier temperature sensor multiplexing weather shutter control blackbody control chopper control and GPS 2 4 6 Cryocooler Controller The cryocooler controller communicates with the PC 104 through a serial port on the Emerald MM DIO board It is a specialized controller from Hymatic designed to regulate the temperature of the NAX025 001 cryocooler 20 2 4 7 Alt Az Electronics The
60. ay the SMA phase correction data Some of the other key features of the Archive Interface include binning data taking the difference of two data sets adjusting data for elevation angle of the unit saving and printing the graph and saving the processed data to a file There are also functions to overlay other relevant data on the graphs such as weather data or command history In addition to maintaining this program I made several additions and improve ments such as the ability to overlay the weather selectable axis ranges and an auto mode which would automatically update the graph every minute with the last 10 minutes of data The improved data handling routines written for the IRMA Control Interface were also used by the Archive Interface since the module is accessed by both programs The IRMA Archive Interface also shares some funtionality with the IRMA Control Interface namely the ability to display the cryocooler data and calibrate the temperature diodes As will be discussed in Chapter 5 as the performance of the radiometer has be come better understood the calibration method for the IRMA units has been modified and improved and with it the algorithm for converting detector signal voltage to PWV has become more complex The new algorithm was not incorporated into the IRMA Archive interface but written in IDL and is discussed in 84 3 SECTION 4 2 IRMA ARCHIVE INTERFACE 68 IRMA Archive Interface Y Axis units Voltage
61. can data to Antenna The identification number of the antenna that the given IRMA unit is associated with This parameter is not always used ElevGear Reduction The gear reduction ratio of the gear box driving the elevation axis AzimGearReduction The gear reduction ratio of the gear box driving the azimuth axis BeltReduction The gear reduction ratio caused by the drive belt The total gear reduction ratio of a given gear is the sum of its gear box reduction ratio and the belt reduction ratio MaxMotorRPM This is the vendor specified maximum motor rotational rate generated when full scale voltage is applied to the motor controller unit MinMotorRPM This is the vendor specified minimum motor rotational rate generated when zero volts is applied to the motor controller unit SECTION C 2 PERL MODULE INSTALLATION 136 MaxGearRPM The maximum recommended rotational rate of the gear head not the motor This value is provided by the motor vendor elev_kProp elev_kInteg elev_kDeriv Servo constants for the elevation axis motor The three constants refer to the proportional integration and derivative constants PID which must be deter mined by the user by tuning the servo algorithm azim_kProp azim_kInteg azim_kDeriv Servo constants for the azimuth axis motor The three constants refer to the proportional integration and derivative PID constants which must be deter mined by the user by tuning the servo algorithm Loca
62. com Data Translation Inc Marlboro MA DT9800 Series User s Manual 2007 URL ftp ftp datx com Public DataAcq DT9800Series Manuals um9800 pdf Regan Dahl Richard Querel and Matthias Sch ck Proc IR THz and MMW Appli cations in Astronomy Atmospheric and Environmental Science In edited by editor 2007 G R Davis D A Naylor M J Griffin T A Clark and W S Holland Broadband Submillimeter Spectroscopy of HCN NH 3 and PH 3 in the Troposphere of Jupiter Icarus 130 387 403 1997 Cirrus Logic Inc Austin Texas CS5531 32 33 34 Product Data Sheet 2004 URL www cirrus com en pubs proDatasheet CS5531323334 F1 pdf
63. curately represented by the internal temperature sensors The smaller error of the ambient measurement causes this point to act as a fulcrum for the extrapolation The 3 K results in an error equivalent to 5 5 K at typical sky temperatures which is the key driver for the 0 1 K accuracy requirement for the heated blackbody During rigorous laboratory testing it was determined that a difference in lid gra SECTION 5 1 CALIBRATION 78 Figure 5 5 A photograph of the additional aperture which was attached to the cryo cooler window to better define the field of view dients was not the only source of the offset observed between IRMA units when viewing the same atmosphere To explore this further it was decided to view a dewar of liquid nitrogen LNo since at this temperature there is essentially zero emission at 20 um The IRMA unit was carefully turned upside down and held above an LN dewar A signal indicative of a stray radiation component was detected prompting further investigation A hot soldering iron was slowly moved around the expected field of view of the detector to further inves tigate this phenomenon Figure 5 4 These tests confirmed that the detector was viewing past the primary mirror To correct for this over illumination an additional aperture was mounted onto the cooler window as shown in Figure 5 5 Though reducing the aperture size lessened the effect of the stray internal box radiation it did not elimina
64. d Initialize Initialize System Alt Az Codes Wait for Wait for in Scanner incoming coming Alt packet Az packet 0 Decode p packet packet Execute command Pattern match Command Code translation Wait for return Alt Az packet Make Command Packet Execute Command Wait for Return Packet Send return packet Ethernet Serial Figure 3 2 A flowchart of Rabbit based system software 39 SECTION 3 2 RABBIT BASED IRMA 36 Y i IRMA gt Config b gy HelperProgs gt gy IRMA gt li SCRIPTS gt Tasks Figure 3 3 An overview of the IRMA directory structure 3 2 1 2 Directory Structure and Configuration Files The software requires the defined directory structure shown in Figure 3 3 The base directory for the software is named IRMA and located in the users home directory usually denoted Within the base IRMA directory there are four subdirectories IRMA IRMA where the custom IRMA Perl modules are located IRMA Config where configuration files are stored IRMA SCRIPTS scripts to be executed with irmaExec pl1 must be located in this directory and finally IRMA HelperProgs where irmaExec pl is located The interpreter is initialized via configuration files The first is the command set which i
65. d by BTRAM 9 These spectra correspond to 1 mm PWV and span an airmass range from 1 to 5 16 NO E Radiance VV m sr Air Mass Figure 5 23 A theoretical curve of growth calculated by integrating flux underneath the spectra shown in Figure 5 22 as a function of airmass The symbols represent the 5 discrete integrals and the solid line represents the continuous curve of growth SECTION 5 2 FIELD PERFORMANCE 100 the unit was powered down and covered to await repair The data from the 149 skydips were compared with the CSO tau data The Teso values were first converted to PWV using a relationship which has been published in the literature for Mauna Kea given by 37 A curve of growth similar to the one shown in Figure 5 23 was then generated by BTRAM using the meteorological data and the PWV as determined from Toso This curve of growth was then fitted to the IRMA skydip data Although as mentioned above it was not possible to calibrate IRMA due to the weather shutter calibration source failed in the open position it is instructive to compare the IRMA skydip data with the that derived from CSO tau During these complemetary observations a storm front was approaching the Hawaiian islands from the North West direction An example of an IRMA skydip and corresponding curve of growth for Toso 0 032 PWV 0 74 mm taken when the storm front is distance is shown in Figure 5 24 together with a satellite
66. d from an unfo cused view of itself which produces a chopped signal of high stability over a wide range of temperature variations This signal is then detected using a standard lock in amplifier 2 3 Mechanical Design As mentioned earlier the prototype IRMA I was constructed on a three legged instrument platform This platform contained all of the optical components two blackbody references and an LNo cooled infrared detector dewar assembly This configuration worked SECTION 2 3 MECHANICAL DESIGN 17 Input 7 Beam Chopper a Blackbody Shutter SSY y Y Cryocooler Power Supplies Umbilical Cryocooler Controller Alt Az Mount Figure 2 3 A rendered model of IRMA in its Altitude Azimuth fork mount well for initial testing and proof of concept but a more compact and robust design was required for autonomous remote operation The IRMA unit was designed to fit into a compact box such that it could be mounted on the side of a radio antenna when used in a phase correcting role see Figure 1 5 An Alt Az mount was also designed to allow IRMA to be pointed in any direction when not attached to an antenna The wet cryostat was a major stumbling block to remote operation since it required an operator to refill the dewar every 4 to 6 hours Moreover the wet cryostat could not be attached to the side of an antenna because had to be kept vertical To address this problem and allow for remote operation the wet cryost
67. d out in Lethbridge and Chile The results of these tests are included in this thesis Acknowledgements This thesis is the result of contributions by many people It would never have been completed without the assistance of numerous others Some have helped me directly while others I have stood on the shoulders of their pervious work in order to reach a little higher I would especially like to thank my supervisor David Naylor for welcoming me into the Astronomical Instrumentation Group to complete my Masters degree I would like to acknowledge those who have made IRMA what it is today Graeme Smith for starting the ball rolling with the construction of the IRMA prototype Ian Chapman for his work on atmospheric modelling allowing us to take IRMA anywhere in the world lan Schofield for all his work on the control and communication software Robin Philips for his work as IRMA Project Manager during the first year of my Masters and for all the help and guidance he gave me Richard Querel for working with me on the IRMA project for these past two years Greg Tompkins for all the insight and entertainment he brought to the group as well as his electronics expertise Frank Klassen for his machining work supporting the IRMA project Brad Gom for his mechanical work as well as the additional time he took from his other projects to help with IRMA Matthias Schock for acting as Project Manager for a month helping to calibrate the
68. del has been performed The scale height of water vapour was found to be the parameter most affecting PWV sensitivity While scale height of water vapour is critical to the accuracy of the model it remains the most difficult to measure in real time 13 1 4 Summary Water vapour is an important component of the atmosphere Due to its inter ference with astronomical signals understanding the quantity of water vapour in the at mosphere is especially important for astronomy IRMA provides a method of measuring the columnar abundance of water vapour PWV for astronomical applications This thesis describes the work that I have done in optimising the instrumental performance of IRMA specifically improving the performance of the control and data acquisition software and refining the calibration schema of the radiometer Chapter 2 gives a brief history as well as an overview of the current optical me chanical and electronics design of the IRMA Hardware Having a knowledge of the physical instrument is required for an understanding of the system software Chapter 3 presents an overview of the previous version of the IRMA software My work involved implementing and improving the previous distributed version of IRMA onto a new platform Chapter 4 describes the control and data analysis software which allow the unit to be controlled re SECTION 1 4 SUMMARY 11 motely and the calibrated data products to be reduced Chapter 5 introduces an improved
69. design provided an opportunity to find solutions to some of the problems with the shutter design The current weather shutter has been the source of some failures such as the incident mentioned in 85 2 In the new design rather than retracting inside the unit the weather shutter rotates out of the beam which will result in less heating of the optical cavity The blackbody on the underside of the new weather shutter will contain an array of sixteen temperature sensors rather than only SECTION 6 3 FUTURE SOFTWARE DEVELOPMENT 107 IRMA Control P S T AutoTasks IRMA Master Controller Command Line irmaExec pl Figure 6 2 Block diagram of original Queue Server functionality two which will allow for a more accurate mapping of any temperature gradients accress the blackbody surface and a more accurate calculation of the effective temperature of the blackbody calibration source 6 3 Future Software Development The IRMA related software will continue to evolve This includes enhancements to both the system software resident on the PC 104 discussed in Chapter 3 and the front end software discussed in Chapter 4 6 3 1 System Software Enhancements 6 3 1 1 Queue Server Integration As mentioned in 83 4 the Queue Server is not required to prevent two scripts from attempting to access irmamc on the PC 104 based IRMA since the TCP IP requests SECTION 6 3 FUTURE SOFTWARE DEVELOPMENT 108 AutoTasks IRMA
70. dpass_filter state in out is used to enable or disable the 455 Hz bandpass filter whose job is to filter out all frequencies above and below the 455 Hz chopper wheel frequency The filter is enabled with the in parameter and disabled with the out parameter The state of the bandpass filter can be read with the bandpass_filter read state command A return value of 0 zero indicates that the filter is not enabled while a return value of 1 indicates that the filter is enabled SHUTTER The command shutter state open close signals the shutter control circuitry to respec tively open or close the shutter Once this command is issued it cannot be aborted The shutter will open or close until it has reached its destination position Shutter condition during actuation can be read with the shutter read limit command The following integer codes are returned 3 shutter is in the process of moving during shutter movement 2 shutter is closed covering the optical aperture and 1 shutter is in the open position optical aperture is exposed Shutter jams can be detected by looking for an increase in the amount of current going to the shutter motor The shutter overcurrent bit is set when this condition occurs Calling the shutter read overcurrent statement returns the value of the overcurrent bit 1 when the overcurrent condition exists and 0 when it does not When the overcurrent condition bit has been set it must be reset to zero by calling the shut
71. e code was organized by having a separate file encapsulate the functions for each command family These functions were called by the dispatcher function which was responsible for interpreting the binary packets sent by the CP The main function simply initialized the system and opened a socket to listen for incoming command packets and forward them to the dispatcher This code was compiled into a single executable named irmamc Originally the PC 104 based MC was a direct port of the Rabbit based MC and the CP software was simply installed on the PC 104 This meant that the IRMAscript interpreter was still encapsulating the commands in binary packets and sending them to the MC which was now located on the same machine Also the same communication protocol was still used even though the packets were not being sent over the network The complexity of the distributed system was being carried over to the PC 104 based system Since the CP and MC were merged into one processor it was no longer necessary to encode commands with the interpreter send them over a TCP socket and decode the commands and execute them on the MC Additionally PC 104 based systems do not have the same resources of a standard desktop PC so the additional overhead caused delays on the order of 10 seconds in executing commands SECTION 3 3 PC 104 BASED SYSTEM AT Running a distributed system on a single processor is inefficient as shown by the ex ecution delay To optimise the
72. e an input span of 22 38 The CS5534 s default offset setting is 0 adc set gain channel value Similar to the set offset command set gain allows the user to manually set a gain value ranging from 64 to 272 to channels 1 through 4 When a channel s gain register is set the offset is subtracted from the A D sample value after which this result is multiplied with the gain value IRMA currently does not use custom gain settings Instead gain and offset are applied to the data in post processing The C 5534 s default gain value is 1 adc read gain channel adc read offset channel These two commands read the current gain and offset values from the CS5534 ADC adc set csr channel gain word rate polarity Each of the CS5534 s four input channels can be configured in terms of signal gain accuracy word rate and input span polarity Channel settings are SECTION A 3 IRMASCRIPT LANGUAGE DEFINITION 127 stored in the CS5534 s four channel setup registers CSR Gain as defined in the CSR is separate from the gain contained in the channel gain registers described earlier Gain values can be defined with the IRMAscript constants or their respective numeric values as shown in Table A 2 Table A 2 CS5534 ADC gain settings in IRMAscript CS5534_GAIN_1 CS5534_GAIN_2 CS5534_GAIN 4 CS 5534_GAIN_ CS5534_GAIN_1 CS5534 GAIN3 CS5534_GAIN_6 64 Resolution refers to the number of noise free bits contained in the A D sample
73. e program the Perl interpreter must be installed A complete list of Perl modules required by the IRMA Control Interface along with installation instructions is found in Appendix C 2 As with the IRMA software installed on the PC 104 the IRMA software running on other machines must follow the same directory structure listed in 3 2 1 3 The IRMA Control Interface is executed from the command line from the IRMA directory with the IRMA pl command 4 2 IRMA Archive Interface The IRMA Archive Interface shown in Figure 4 10 is used for viewing data from the IRMA units It is similar to the data monitoring portion of the IRMA Control Interface In fact through modular programming the graphing functions written by Ms Amy Smedes for the IRMA Control Interface are re used in the IRMA Archive Interface It also has many features which are not included in the Control Interface These additional features relate to data processing The IRMA Archive Interface has no menu items and all options and controls are selected through the checkboxes and buttons on the GUI The IRMA Archive Interface has been a continually evolving program As different data reduction SECTION 4 2 IRMA ARCHIVE INTERFACE 67 fuctions were needed they have been added to the interface For example after completing a testing campaign at the Smithsonian Millimeter Array SMA where IRMA was used in a phase correction role see Figure 1 5 a button was added which would overl
74. e time when the parameters immediately following took effect The parameters within a block include all lines following the time stamp up to but not including the next block delimiter line A parameter line consists of a label followed by a value and is terminated with a carriage return A whitespace separates the label from the value SECTION 3 2 RABBIT BASED IRMA 38 When irmaExec pl is executed it reads in the box file specified by the box number command line parameter accepting parameter fields whose time stamp is closest to the current time date For example using the configuration file in Table 3 2 if irmaExec pl were executed on 2006 09 01 it would accept the IP address values from the command block dated 2006 08 09T15 00 00 because this is the most recent entry For a complete list and description of parameters that can be defined in a CP configuration file see Appendix C 1 If parameters are not defined default parameters are assigned 3 2 1 3 CP software requirements and execution As described earlier the system is designed to run on a desktop or server PC running Linux All CP software has been designed to run with Perl 5 8 6 and later In addition to having the Perl interpreter installed additional Perl modules are also required Appendix C 2 lists the modules and installation instructions Scripts are executed on an IRMA unit by issuing a command following the form HelperProgs irmaExec pl lt box number gt lt
75. el MB102 The profile is the end to end instrument response the convolution of the filter transmission profile transmission of the anti reflection coated ZnSe window and the photodetector response over the given spectral range Vi Pa 1 do Pi VE Vi W 5 5 Once calibrated these parameters can subsequently be applied to the measured signal voltage during sky observations using Equation 5 3 to determine the infrared flux radiated by the atmosphere Finally the atmospheric flux can be related to PWV by using an appropriate atmospheric model as described in 1 3 2 5 1 2 Calibration Issues In the IRMA I prototype the lower temperature blackbody source consisted of a large dewar of liquid nitrogen LN2 77 K 73 K at the summit of Mauna Kea The second calibration blackbody was a large aluminum plate coated with carbon black doped SECTION 5 1 CALIBRATION 75 Figure 5 2 A thermal image of an internal lid blackbody taken with a Fluke Ti20 7 14 left A cross sectional temperature profile of the image taken through the middle um camera thermally conductive epoxy which tracked the ambient temperature Since the effective sky temperature falls between these two extremes it allowed for a direct interpolation of the measured atmospheric flux from the signal voltage As discussed in Chapter 2 the IRMA unit was redesigned to provide robust remote and autonomous operation This redesign necessitated remov
76. elation Coefficients for Lethbridge test data based on sky calibration CS5534 ADC gain settings in IRMAscript o o CS5534 ADC sample resolution settings in IRMAscript CS5534 ADC polarity settings in IRMAscript IRMA AAC command codes sent over MC AAC serial link xii Chapter 1 Introduction 1 1 Overview The Infrared Radiometer for Millimetre Astronomy IRMA is an instrument used to measure the amount of water vapour in the atmosphere usually expressed as precip itable water vapour PWV It was developed as a collaboration between the University of Lethbridge and the Herzberg Institute of Astrophysics to be a fast relatively simple robust radiometer which could be used to determine PWV at high altitude sites around the world Central to IRMA is a sensitive infrared photoconductive detector used to measure the atmospheric emission of a carefully selected spectral band centred at 20 um The total radiant infrared flux received from the atmosphere is converted into PWV by use of an atmospheric model BTRAM Blue Sky Transmittance and Radiance Atmospheric Model 2 which has been developed by our research group SECTION 1 2 PRECIPITABLE WATER VAPOUR 2 1 2 Precipitable water vapour Water vapour is simply water in the gaseous phase Precipitable water vapour PWV is a measure of the columnar abundance of water vapour in the atmosphere PWV refers to the depth of liquid water present upon condensing a column of at
77. en the subject of 4 theses from the Astronomical Instrumentation Group at the University of Lethbridge 1 33 24 13 Two of these theses focused on the instrumentation and software design I joined the group as the IRMA project was entering its latest design change moving from a distributed Rabbit based system 3 2 to the current PC 104 based system 3 3 To support all of the existing systems it was necessary to become familiar with both IRMA designs in addition to the graphical interfaces which supported them In all this involved familiarising myself SECTION 6 4 FINAL THOUGHTS 111 with over 68 000 lines of code As I gained greater understanding of the system I was able to make improvements to the operation from minor bug fixes to the design changes mentioned in this thesis The most significant update was the porting of the IRMAscript interpreter from Perl to C and incorporating it into the Master Controller 3 3 1 This allowed for over an order of magnitude improvement in the execution time and reduced the complexity of the system by eliminating the need for a special communication protocol between the Command Processor and Master Controller Approximately 2 800 lines of code were added for this change while 14 000 lines of code were eliminated During this time we increased our understanding of the IRMA instrument which lead to the development of an improved calibration method 5 1 3 IRMA has now been deployed ar
78. enges an additional aperture was attached to the window of the cryocooler which helped reduce though not eliminate the sensitivity to stray radiation A large reference blackbody was constructed to act as a primary calibration source A new calibration method was developed which associated the internal blackbody secondary to the reference blackbody primary and determined the signal dependence on stray radiation Once calibrated three units were tested in Lethbridge side by side viewing the same atmosphere It was shown using the new calibration method that there was very good agreement between the units In order to evaluate the performance of the units under the operating condition for which IRMA was designed the units were sent to Chile where they again observed the same atmosphere while on a mountain at an altitude of 3000 m As expected the units showed an even better correlation with measured inter unit correlation coefficients of 0 99 under these conditions The second section of this chapter demonstrates the flexibility of the IRMA unit SECTION 5 3 SUMMARY 103 During a time of instrumental failure it was possible to modify the operation of the unit remotely such that useful data could be obtained With the Alt Az mount damaged and the weather shutter stuck open skydips were performed in the same direction as the Cal tech Submillimeter Observatory CSO taumeter an independent measure of water vapour These results were c
79. es the same patch of sky as a 10 m antenna as described in 2 2 Importantly IRMA is a passive radiometer and therefore does not require any radio frequency RF components which may interfere with the RF instrumentation of a radio telescope Figure 1 5 shows IRMA mounted to one of the antennas at the Smithsonian Millimeter Array SMA on Mauna Kea in Hawaii during phase correction testing in June 2004 1 2 2 Measuring Water Vapour Many methods have been described in literature to measure the water vapour content of the atmosphere These include 183 GHz radiometer 5 225 GHz Radiometer Caltech Submillimeter Observatory CSO Taumeter 6 tropospheric delay measurements using GPS signals 7 and infrared radiometer 8 IRMA is the only method to employ a passive infrared radiometer Moreover it is also compact robust and able to be operated remotely SECTION 1 3 THE IRMA SOLUTION 7 Figure 1 5 A photograph showing IRMA attached to the side of one of the antennas of the Smithsonian Millimeter Array telescope on Mauna Kea Hawaii 1 3 The IRMA Solution 1 3 1 The Instrument The IRMA instrument will be the main focus of my thesis IRMA is essentially a very sensitive infrared thermometer which measures the emission from the atmosphere in a narrow spectral band where only water vapour contributes to that emission The IRMA device consists of an optical system which focuses the emission onto a photoconductive detector wh
80. expressed as floating point degrees into degree minute second DMS format The degrees minutes and seconds must be variables because the deg2dms function places values in these variables They are not input variables STARTPROG SOCKET OPEN ENDPROG SOCKET CLOSE Open and close a TCP IP stream socket connection to the IRMA master controller If a script contains instructions to execute on the IRMA master controller a network socket must be established to the IRMA MC as low level IRMA system commands and data flow over this connection If a script does not contain IRMA hardware control commands it is not necessary to wrap a script with these statements LOCALHOST This command handles system functions performed by the host computer s operating sys tem localhost log open Open the log file A log file name must be created using the new log filename command before logging can commence localhost log close Close the log file NEW The new family of functions creates new data items of various types such as filenames and time stamps new log filename Automatically generate and return a filename and create a directory path for the new file Filenames generated by this function follow the ISO time format SECTION A 3 IRMASCRIPT LANGUAGE DEFINITION 117 YYYY MM DDTHHmmSS dat and end with the dat extension File paths follow the structure IRMAdata IRMA_ lt boxNumber gt YY YY YYYY MM DD where IRMAdata is a link or filesystem
81. g 18 RCM 2010 agi SLAVE gt gt gt 26 24V GND ALT AZ DAC REF Figure 2 7 A Block Diagram of the IRMA Alt Az Controller showing how the Rabbit RCM2010 interfaces with motor control system on parallel port A Setting either of these two lines sets the corresponding axis into clockwise CW rotation while clearing puts the corresponding axis into counterclockwise CCW rotation Altitude CW and CCW limits are respectively mapped to input DIO lines 2 and 0 on parallel port B Azimuth CW and CCW limits are respectively mapped to input DIO lines 1 and 3 on parallel port B When a limit line is set a limit has been encountered otherwise the axis angle is within safe rotational limits The speed of the motor is set by supplying an analog voltage of 0 to 2 5 V to the controller An 8 bit 2 channel Maxim 5223 serial digital to analog converter DAC is SECTION 2 4 ELECTRONICS DATA used to generate the required analog voltage The 5223 has a 3 wire serial communications interface involving a chip select line CS a serial clock line SCLK and a data input line DIN All the communication lines are mapped to Rabbit parallel port A CS active low SCLK and DIN are mapped to pins 1 2 and 3 respectively Voltage is individually adjustable on each of the two analog outputs of the 5223 A and B Voltage can be set between 0 V to full scale the input reference voltage in 256 equal steps Analog output c
82. g to execute For example in Figure 4 2 the executing SECTION 4 1 IRMA CONTROL INTERFACE GMT 21 06 12 Chopper Shutter O Unit 12 Refresh Chop Off ome wo on 4 Shutter in Motion Black Body Pro 60Hz Notch e fi x gt gt enn v 455Hz BP mawa oo 60Hz Notch Off __ se eron _Sun Shutter Sync RTCIGPS Cooler 0k serre o FERRO RTC 21 06 09 Stop Scan Manual Cooler On Set Temperature GPS 00 00 00 59 Figure 4 3 A screenshot of the Current unit status area of the IRMA Control Interface showing the current status unit showing the chopper on the shutter moving and the blackbody off script is bbOn and Refresh is waiting to execute The list of scripts waiting to execute includes scripts which were initiated by the current operator an operator at another location controlling the same unit or scripts executed automatically by AutoTasks IRMAscripts executed directly from the command line however will not be listed The operator is able to remove any queued items by right clicking on the item and selecting remove 4 1 1 2 Current Unit Control and Status The Current Unit Control and Status is located to the right of the Queue Monitor as shown in Figure 4 3 The unit is selected from the drop down list located at the top of this section This selection also affects the Queue Status and Data Monitoring sections Control of the un
83. h the atmosphere increases by 6 mm 4 More over the distribution of water vapour in the atmosphere is neither spatially nor temporally homogeneous inside the column projected from the antenna s receiving dish For example with a wind speed of 20 m s which is not uncommon at the altitude of many observatories the mass of air in the column above the telescope changes every 0 5 s Thus in general the amount of water vapour along the line of site of each receiving antenna will be different The effect of atmospheric phase distortion is illustrated in Figure 1 4 which shows an interferometric array with two antennas The antenna pair observe the same object whose wavefront appears planar in the upper atmosphere The wavefront above the left SECTION 1 2 PRECIPITABLE WATER VAPOUR 6 hand antenna passes through a region of water vapour which adds excess optical path length d to the incoming signal Interferometry requires the precise measurement of the time the wavefront was received at each antenna The apparent direction of the observed object is perpendicular to the planar wavefront A slight phase error manifests itself as a slight change in the immediate apparent angle of the astronomical source decreasing the spatial resolution of the interferometer Real time knowledge of the PWV above each antenna would allow for a partial correction of this phase delay IRMA is well suited for phase correction It has been designed such that it sampl
84. hain taken at 08 00 UT April 24 2007 three hour after Figure 5 24 left The IRMA skydip asterisks and the modeled curve of growth derived from the A GO independent CSO tau data taken at the same time right It can be seen that as the storm has advanced there are now significant differences between the IRMA and CSO tau data different parts of the sky and have different beam sizes which become more pronounced as the storm front approaches Mauna Kea Although the uncalibrated IRMA data are of limited scientific value Figures 5 24 and 5 25 shows the effectiveness of IRMA in sensing impending inclement weather SECTION 5 3 SUMMARY 102 5 3 Summary The prototype IRMA used a simple two point linear calibration method How ever the latest mechanical designs posed some problems for this method With the elim ination of wet cryogens the two calibration points moved from liquid nitrogen LNo and ambient temperatures to ambient and 20 K above ambient temperature As a result the calibrated sky measurements need to be extrapolated rather than interpolated from the calibration points increasing the required precision of the thermometry A gradient was found to be present across the internal blackbody calibration source when heated making it more difficult to accurately determine its effective temperature In addition it was found that the signal voltage was also sensitive to stray internal radiation To address these chall
85. hannel A is mapped to the azimuth motor controller while analog output B is mapped to the altitude motor controller Using dip switches on the motor controllers the speed is limited to a range of 500 to 12 500 RPM However the maximum manufacturer recommended rotational speed of the gear box is 8000 RPM a limit which is set within the software 2 4 7 3 Optical Encoders The AAC determines the position of the Alt Az mount through the use of two US Digital E6M optical encoders 22 which are interfaced with the RCM2010 through a US Digital LS7266R1 encoder chip 23 The optical encoders employ 2048 lines per revolution optical encoder wheels However the LS7266R1 can obtain 8192 lines of resolution per revolution or 2 arc seconds per encoder step by operating in quadrature mode To operate in quadrature mode a filter clock with a frequency in the range 10 to 35 MHz must be connected to the LS7266 LS7266 pin 2 The filter clock is supplied by a 10 MHz oscillator The LS7266R1 detects and counts ticks from the encoders in either mode where the count is relative to a software determined location SECTION 2 4 ELECTRONICS 28 The LS7266 optical encoder chip is capable of reading the position from two en coders and interfaces to the RCM2010 using 12 DIO lines Data sent to and received from the LS7266 is carried over an 8 bit bidirectional data bus mapped to pins 0 through 7 on parallel port D The software must set the appropriate da
86. he command is decoded and executed For a complete list of the possible AAC commands and their associated integer codes see Appendix B Most AAC commands are short duration such as querying SECTION 3 2 RABBIT BASED IRMA 44 the current position and therefore are executed within the communication task Long duration tasks such as axis movements must be run in their own task otherwise serial communication would not be possible until the task was complete The job task waits for an event flag to be set by the serial communication task Once the flag is received it carries out the desired task in parallel with the communication task such that short duration task can still be handled If axis movement is requested the job task starts either the single axis move task or the dual axis move task depending on the request To ensure accurate speed control a proportional integral derivative PID servo tracking loop is used The axis move task waits on a 10 Hz servo tick event flag set by the metronome task the highest priority task of the AAC to ensure accurate timing which signals the move task to update the servo loop calculations The servo loop must be updated exactly at this rate for the servo control algorithm to function correctly The ability to meet deadlines in a multi tasking environment is the defining attribute of real time programming 28 Once the movement has completed the move task sets an event flag which signals to the job t
87. he sixteen sensors were calibrated by placing the entire large blackbody in a convection oven and referencing each sensor to a calibrated Lakeshore temperature diode 34 Initially and creatively the temperature sensors of the LBB were read using a PC 104 and a spare IRMA motherboard The IRMA software required only minor modifications to acquire the temperature data from the 16 sensors This was later replaced with a dedicated Data Translation DT9803 35 high performance USB data ac quisition module A wooden housing was built to allow for repeatable positioning atop the IRMA viewing port shown in Figure 5 6 Several of the internal IRMA temperatures were also incorporated into the calibration in an attempt to identify and account for the effects of stray internal radiation measured by the detector The calibration was performed by taking a series of LBB measurements inter spersed with internal lid blackbody measurements as shown in Figure 5 7 A typical cali bration scheme consists of viewing the LBB at four distinct temperatures of approximately 22 C 35 C 60 C and 90 C Once the calibration data were obtained a linear least squares fit was performed between the effective LBB temperature being viewed by IRMA the internal IRMA tem peratures to account for stray radiation and the photodetector signal voltage This fit was then used to calibrate the internal blackbody temperature against the LBB such that it could be used as a seconda
88. ic calibration method the detector observes two blackbody calibration sources of different temperature For a blackbody of temperature T in Kelvin the spectral radiance L o T is given by the well known Planck equation 2hc 1004 05 hco exp kp z 1 where is the wavenumber in cm his Planck s constant in Js kg is Boltzmann constant Ea T W m sr cm 5 1 in J K and cis the speed of light in m s Calculating the radiant flux detected by IRMA required knowledge of the instrumental throughput AQ and the spectral response function of the detecting system Fy The flux is calculated by integrating over the normalised spectral bandpass of the detection system 450 to 575 cm equivalent to 17 to 22 um shown in Figure 5 1 and is given by 575 2 4 3 2hc 1 00 7 EE od M 5 2 4 hco 50 1 ex 72 The measured signal voltage is related to the calculated flux by an equation of the form D AV4S Wh 5 3 After observing the two calibration sources the gain A and offset Po of the voltage to flux relationship are established by the following formulas _ Oo Oy Ame SA V V 5 4 SECTION 5 1 CALIBRATION 74 Wavelength um 5 25 20 3 1 0 u 0 8 D 0 5 0 6 o 0 E 0 4 E N 0 2 T Sn a 400 450 500 550 600 650 Wavenumber cm Figure 5 1 Normalised IRMA instrument response function as measured at 70 K using an ABB Bomem FTS mod
89. ich will be discussed in 82 2 Finally this flux can then be converted into PWV using an atmospheric model BTRAM which has been developed by the Astronomical Instrumentation Group at the University of Lethbridge and is briefly described in the next SECTION 1 3 THE IRMA SOLUTION 8 section A mechanical system described in 2 3 has been designed such that IRMA can be pointed in any direction and gives it the ability to protect itself from the elements with a weather shutter which doubles as a calibration source The electronics discussed in 82 4 facilitate the control of the instrument through software which will be discussed in Chapter 3 Other software discussed in Chapter 4 has also been designed which allows for remote access and control of the IRMA unit in addition to data reduction routines 1 3 2 The Atmospheric Model BTRAM To determine the relationship between the detected flux and PWV IRMA em ploys the atmospheric model BTRAM Blue Sky Transmittance and Radiance Atmospheric Model BTRAM 9 originally known as ULTRAM 2 is a line by line layer by layer ra diative transfer model used to simulate transmission emission or opacity of a user definable atmosphere The impetus for developing BTRAM was that IRMA could be deployed in lo cations where standard models do not apply The main goal in developing BTRAM was to provide the user with the flexibility to model radiative transfer under local conditions It was orig
90. inally designed as a customisable GUI with a simplified subset of geometries available in Fast Atmospheric Signature Code FASCODE 10 FASCODE was difficult to customise due to the volume of code causing modifications to one aspect of the model would have unpredictable consequences in other parts of the code BTRAM has been de veloped into an independent easy to use fully customisable atmospheric model using the HITRAN 2004 database of spectral lines 11 Within BTRAM there are nine pre built atmospheric profiles Antarctic Summer SECTION 1 3 THE IRMA SOLUTION 9 Chajnantor Winter Mauna Kea Mid Latitude Summer Mid Latitude Winter Sub Arctic Summer Sub Arctic Winter Tropical and U S Standard Atmosphere 1976 A customised atmospheric model is built by either modifying one of the pre built profiles or creating one based on radiosonde data when available A radiosonde is a suite of meteorological instruments carried by a weather balloon which makes in situ measurements of pressure temperature wind speed and dew point or relative humidity These data are then used to determine inputs for the creation of a site specific atmospheric model such as the scale height and the adiabatic lapse rate Astronomical observations must account for the variable transmission through the earths atmosphere While IRMA measures emission from of the Earth s atmosphere by application of Kirchoff s Laws the transmission can be derived and a suitable cor
91. ing Sampled patch of sky of solid angle Q at detector area Subtending a solid Detector area A angle Q at paraboloid Focal length Range to sampled patch of sky of paraboloid Figure 2 2 A schematic of the equivalent optical system of IRMA 1 Referring to the schematic of the optical system in Figure 2 2 conservation of the SECTION 2 2 OPTICAL SYSTEM 15 throughput of the optical system requires that Aggy A 2 1 where Ag is the area of the detector Qg is the solid angle viewed by the detector Ap is the area of the paraboloid and Q is the solid angle subtended by the 10 m diameter patch of atmosphere viewed at a distance of 1 km The solid angle subtended by the sampling area A at the parabolic mirror at a distance R can be approximated as Q As I and similarly the solid angle subtended by the paraboloid at the detector is Qq Ap f Equation 2 1 can then be rearranged to specify a focal length _p Ad _ pf where dy is the diameter of the detector element and d is the diameter of the source area The focal length of the paraboloid is the determined by dy ds and R which have been previously determined and is independent of the diameter of the paraboloid Inserting the values into equation 2 2 gives a focal length of f 1000 Ahi siim 2 3 Thus a 10 cm diameter f 1 off axis parabolic mirror was chosen as a good compromise between collecting area and compact design 2 2 2 I
92. ing the dependence on LN which is not available at remote sites both in the wet cryostat and the calibration blackbody As has been discussed 82 3 1 the wet cryostat has been replaced by a Stirling cryocooler The original calibration sources were replaced by a blackbody attached to the underside of the weather shutter 2 3 1 which could be heated to a temperature of 20 K above ambient The two point calibration was then accomplished by observing the blackbody at ambient and elevated temperatures Unfortunately in this method the temperature of both calibration points lies above the effective temperature of the sky and an extrapolation of the calibration data is needed to determine the atmospheric flux Modeling has shown that in order to measure PWV to an accuracy of 10 the effective temperature must be know to a precision of 0 1 K 13 In the original design two temperature sensors were embedded in the blackbody SECTION 5 1 CALIBRATION 76 Voltage LI x Warm 6 C 1 200 H Ambient e EJ A Sky e i S F 2 1 150 F ee 8 F gt 1 100 a e 1 050 ee an Ls PSE RI A Figure 5 3 Potential error in atmospheric flux introduced through a two point extrapolation The flux calculation for the ambient reading is taken to have small error bars and acts as a fulcrum point Warm blackbody flux variance equivalent to 3 K results in an extrapolated error equivalent to 5 5 K at typical
93. ipulation 4 c st 2 44 0 cos Bee a A ASD Utility Functions sesse a8 6 ee a Se we ceed j Het ei A 3 3 Variable Manipulation Se esine h a e 0200050 A 3 4 Delays A 3 5 Flow Control coca aia i Ee ke ke T a a a a ae A 3 6 Input Output Commands ooa dia e s s A 3 7 System Commands oo 02000002 eee eee B AltAz Commands C Configuration and Installation C 1 CP configuration file options oie ee eae e a ES C 2 Perl Module Installation 2 52 45 e288 24 4 e459 e C 2 1 System Software Perl Modules 202004 C 2 2 Control and Visualisation Software Perl Modules C 3 Sample Daily Tasks File 2 2 0 0 0 0 0000000000008 C 4 AutoTasks conf References viii 112 112 113 115 115 116 117 118 118 119 120 134 135 135 136 137 137 138 138 139 ix List of Figures 1 1 Atmospheric windows 00 0000 eee ee ee 2 1 2 A photograph showing an IRMA unit deployed in front of the Gemini South telescope in Chile concer a aR RS a a eS 4 1 3 A photograph showing three IRMA units deployed on one of the Chilean sites being considered as part of the Thirty Meter Telescope TMT project 4 1 4 A schematic of the cause of atmospheric phase distortion of a celestial signal 1 5 1 5 A photograph showing IRMA attached to the side of one of the antennas of the Smithsonian Millimeter Array telescope on Mauna Kea Hawaii 7 2 1 A photograph of the original IRMA inst
94. it is accomplished by use of a column of command buttons one for each function The function of the button is determined by the status of the Unit For example SECTION 4 1 IRMA CONTROL INTERFACE 60 Move Altaz Mount Move Altaz for Unit 3 Elevation Refresh Azimuth Dual Axis Elevation Azimuth Elevation lo deg fo min Jo sec Azimuth fo deg o min Jo sec Speed fsec 0 2 Siow Move To Figure 4 4 A screenshot of the dialog box that the operator uses to point IRMA if the chopper wheel is currently turned on the button will display Chop Off indicating that the chopper wheel can be turned off by clicking the button If the status of the unit is unknown the button will be disabled In order to enable it the user simply clicks the Refresh button to force a query of the unit status Since Alt Az motion cannot be accomplished through a simple control button Alt Az functions are accessed though a menu option under Engineering The Move Alt Az Mount dialog box is shown in figure 4 4 The Alt Az mount can also be initialised by selecting the appropriate menu item The sun shutter is also controlled through the Engineering menu Under normal operating conditions it will only be controlled by the hardware when the sun sensor is illuminated however provision is made for manual control during testing To send a command to the IRMA unit the Control Interface executes the process SECTION 4 1 IRMA CONTR
95. itch not responding the initialisation routine would continue to drive the motor seeking the CW limit Using the relative movement commands it was possible to verify that the elevation drive still had a full range of motion 190 and it was pointed to zenith using the new commands However the weather protection routine uses absolute movements and therefore requires a fully initialised unit After a standard elevation initialisation the unit stops in the CW limit and sets its position to 0 degrees elevation even though it is actually pointing below the horizon An offset is applied to correct this Before a unit is shipped the offset between the limit and the true 0 degrees is calculated and stored in a file on the unit It was determined that we could simulate an initialisation by using an offset value to compensate for the uninitialised value 90 000 optical encoder OE counts By applying an offset of 90 000 plus the original offset the unit would appear initialised This solution corrected the Alt Az problems until a repair could be scheduled During this same period and before the unit could be repaired the motor on the weather shutter failed in the open position With the weather shutter not functioning IRMA was seriously compromised as it was no longer possible to calibrate the unit Had the shutter failed in the closed position it would have been impossible to view the atmosphere at all Although we were not able to obtain calibration me
96. ive task within the MC and is primarily concerned with receiving commands from the CP Since the majority of commands are classed as short duration meaning they take less than a second to execute they are allowed to execute within the dispatcher task Long duration commands such as scans however must be executed outside the context of the dispatcher task otherwise the dispatcher would be unable to accomplish its primary duty of listening for incoming commands until the task was completed When a scan is requested the dispatcher task forks the scan task as shown in Figure 3 5 then returns to wait for SECTION 3 2 RABBIT BASED IRMA 41 Boot Metronome Heartbeat Task Task T i Flag send data I Dispatcher gt Scan Task Task Data Collection Interrupt Service Routine Figure 3 5 A schematic of the IRMA master control software task structure during scanning incoming commands This allows short duration tasks to run concurrently with the long duration scan task The scan task is responsible for the data collection process It also handles the construction of data packets and their transmission to the CP The scan task starts the metronome task and enables the ISR to trigger on the external interrupt provided by the chopper blade The metronome whose priority is greater than the scan task counts the data points collected by the ISR fetches the current Alt Az coordinate from the AAC for each data point and signals
97. l Time Piece IRMA CCIT16_CRC IRMA commander IRMA packetComm C 2 2 Control and Visualisation Software Perl Modules Time Piece Time Seconds List Util qw max min Tk Tk Balloon Math Round qw al1 Math Trig Expect IRMA Data IRMA scripter IRMA queue IRMA BoxInfo IRMA gui_graph SECTION C 3 SAMPLE DAILY TASKS FILE 138 C 3 Sample Daily Tasks File Shown below is a typical daily tasks file which monitors the status of the unit every 10 minutes and performs a calibration routine every two hours 1 accuracy allowance function id label repeat reptime time At 5 script At 5 script At 5 script At 5 script At 5 script status check irma 5699 status 1 00 10 00 00 04 00 shutterOpen irmal 57001 shutterOpenl11102 00 00100 40 00 shutterClose irmal5703 shutterClose 1 02 00 00 00 00 00 bb0n irma 5704 bb0n 1102 00 00 100 10 00 bbOff irmal5706 bb0ff 1102 00 00100 41 00 C 4 AutoTasks conf Shown is a typical autoTasks conf file It must be located in the IRMA Config directory daily_tasks daily_tasks daily_tasks daily_tasks skymap on O on 1 delay 30 boxes 1 unkhumidrun 1 skymap delay 20 skymap boxes 1 skymap unkhumidrun 1 humidity on 1 humidity humid 70 humidity delay 60 humidity boxes 1 cooler on 1 cooler comp_amp 1 cooler temp 1 cooler delay 30 cooler boxes 1 cooler osc_freq 0 139 References 1 S 10 Graeme J Smith
98. lso key factor in deciding the location of future ground based observatories IRMA has been used as both an opacity meter for existing observatory sites Figure 1 2 as well as in a site testing role Figure 1 3 Water vapour is also a cause of phase delay in submillimeter interferometric arrays These submillimeter arrays are ground based observatories which are able to synthesize a massive receiving antenna whose diameter equals the length of the maximum baseline of the array The maximum baseline is the distance between the two farthest separated antennas in the array SECTION 1 2 PRECIPITABLE WATER VAPOUR 4 Figure 1 3 A photograph showing three IRMA units deployed on one of the Chilean sites being considered as part of the Thirty Meter Telescope TMT project SECTION 1 2 PRECIPITABLE WATER VAPOUR 5 actual direction of source instantaneous apparent angle of astronomical source water vapor features additional electro e tates aid e magnetic path length interferometer baseline length Figure 1 4 A schematic of the cause of atmospheric phase distortion of a celestial signal 1 At millimeter wavelengths water vapour found in the Earth s atmosphere is present in sufficient amounts to slow down the incoming wavefront of a celestial signal Studies have shown that for every 1 mm of precipitable water vapour along the line of sight of a millimeter telescope the optical path length throug
99. ltimately get sent through irmaExec pl IRMAscript is a custom scripting language used to control the unit It provides a highly flexible fine grained control mecha nism for the instrument A comprehensive description of the IRMAscript language can be SECTION 3 2 RABBIT BASED IRMA 32 Command Processor Master Controller RCM2100 Alt Az Controller RCM2010 Ethernet Serial Figure 3 1 A System diagram for the Rabbit based IRMA found in Appendix A 3 2 1 1 irmaExec Software Structure irmaExec pl written in Perl 26 a popular systems programming language is the interpreter for IRMAscript Interpreters and compilers translate instructions from one form to another Interpreters typically translate statements into actions as they are encountered while compilers translate source code into machine language instructions to be executed on the target machine at a later time Compilers and interpreters are built using similar mechanisms A typical compiler consists of a scanner a parser a scope checker and code generation An interpreter such as irmaExec pl is also built using these components except that it will typically execute the statements rather than generating code SECTION 3 2 RABBIT BASED IRMA 33 The scanner is responsible for recognising tokens or strings of a language statement irmaExec pl accomplishes this by reading in the source file ignoring white space and comments and
100. manner Integrating the spectra with respect to wavenumber and then plotting the radiant flux against the respective airmass gives the curve shown in Figure 5 23 which is known as a curve of growth The atmospheric opacity 7 can be calculated from this curve and is proportional to the precipitable water vapour in the atmosphere On the summit of Mauna Kea the Caltech Submillimeter Observatory CSO Tau Meter also measures PWV by performing skydips with a 225 GHz radiometer 6 at a fixed azimuth of 316 degrees every 10 minutes to determine the atmospheric opacity Teso IRMA was pointed to approximately the same azimuth as the CSO Tau Meter while the skydips were performed 149 skydips were performed over the course of 19 hours Following this SECTION 5 2 FIELD PERFORMANCE 99 012 e pe 0 10 I E O A 0 08 0 N i E 0 06 y J 2 0 04 All IK LIME z a 3 8 0 02 Y 0 00 480 490 500 510 520 Wavenumber cm Figure 5 22 Theoretical emission spectra of the atmosphere in the 20 u spectral region produce
101. med Include In Refresh when a unit is selected W Cooler Query Unit Status Every 5 mins User jirmauser _ Gonnect GPS Time 3 Load Data Plots 4 Load Data Plots Wm Pass Pet New Server Altaz Mount EASON pS Every 5 mins Save Changes Delete Server W Sun Shutter Cancel Last 10 mins A Figure 4 9 A screenshot of the dialog box that the operator uses to select GUl options A and Queue Server information B Tasks may either be setup to run on a specific day of as the default task file Clicking Apply will upload the task file as well as any associated IRMAscript files 4 1 1 7 IRMA Control Interface Setup Additional dialog boxes are provided under the Options menu to allow the op erator to customize the IRMA Control Interface The GUI Options dialog Figure 4 9A allows the operator to select which units to control as well as other options such as automatic status queries and graphing The Queue Server Options dialog Figure 4 9 B is used to associate unit num SECTION 4 2 IRMA ARCHIVE INTERFACE 66 bers to the Queue Servers This dialog is also used to setup the Queue Server information which includes the IP address username and password for remote login and SSH port The standard SSH port is 22 however some IRMA units are setup with non standard ports 4 1 2 Requirements and Execution As mentioned earlier the control GUI is written in Perl Tk Therefore in order to use th
102. mosphere PWV is a linear parameter referred to in units of mm For example 1 mm PWV condensed over an area of 1 m would correspond to 3 34 x 10 molecules of water in the atmosphere 1 2 1 Importance of Measuring PWV Atmospheric Opacity inm 10nm 100nm tum 10um 100um 1mm 1 cm 10 cm 1m 10m 100m 1km Wavelength Figure 1 1 The atmospheric opacity as a function of wavelength The graphs show with regions of the electro magnetic spectrum are visible from ground based sites Courtesy of NASA JPL Caltech Astronomical objects emit light at all wavelengths of the electromagnetic spectrum By and large this light travels vast distances through space relatively unimpeded until it SECTION 1 2 PRECIPITABLE WATER VAPOUR 3 reaches the Earth s atmosphere While our atmosphere is transparent to some forms of radiation it is opaque to many wavelengths as shown in Figure 1 1 3 The dominant source of opacity for infrared wavelengths is due to atmospheric water vapour Therefore the best infrared observing sites require low column abundances of water vapour This is why the world s best observatories are located at high and dry mountain tops which place them above most of the atmospheric water vapour Since water vapour is the primary source of atmospheric opacity quantifying its abundance usually expressed in terms of PWV is critical for the calibration of astronomical data at observatories around the world For this reason it is a
103. n operated co located observing the same atmosphere While on site the correlation between the units was verified However a small offset was evident between the units shown in Figure 5 14 It was assumed that this was because a thorough study of the internal box temperatures had not yet been completed Upon my return to Lethbridge the statistical study of the contributing temperature sensors was used to determine new coefficients The results of the test campaign in Chile using the new coefficients are shown in Figure 5 15 36 Figure 5 16 with corresponding correlation coefficients listed in Table 5 2 shows that the data points are more closely correlated than with the Lethbridge data SECTION 5 1 CALIBRATION 91 CI i nO a SO NS a CR a DR DD E ERE x Box 10 5 A Box 11 o Box 12 PWV mm A LE Al o o ee ete jojo 2 454118 105 2 454119 105 2 454120 105 2 454121 105 Julian Day Figure 5 15 A time series plot of the three co located IRMA units in Chile The gaps in the data correspond to calibration sequences These data vvere derived using the calibration coefficients calculated using the statistical approach described in 5 1 3 1 Si a ESET TEST L Et TA al N PWV mm A Lit tf ff tf AO TE MA RES SI SE Fido LILIJIT LI Lit TROTTTTTTP TTT TTT rr ry rrr rrr rrr pr rrr rrr rr yp rrr x 10 vs 11 A 10 vs 12 o 11 vs 12 3 2 3 4 5 6 PWV mm
104. nel from the MC to the AAC altaz init altaz Once the Alt Az serial channel has been opened the first command that should be sent to the AAC is the init altaz command This command has the effect of initializing the AAC s two channel optical encoder chip that is responsible for digitizing axis encoder positions Upon initializing the optical encoder chip altitude and azimuth axis positions are set to 90 000 encoder units There are 8192 encoder units in one revolution SECTION A 3 IRMASCRIPT LANGUAGE DEFINITION 130 altaz init axes axis Upon using the AAC for the first time or where re initialization is required the axes should be sent to their default positions The init axes command performs a homing operation whereby it determines the clockwise and counter clockwise optical limits on both axes When axis homing has completed altitude and azimuth positions are set to position 0 in encoder units The axis parameter can be defined in three ways altitude elevation or azimuth altaz init motor The gearboxes and motor controllers used in the IRMA AAC differ from unit to unit In order to deal with these variations gear ratios motor RPM values and other configuration information unique to the given IRMA unit is contained in a configuration file By issuing this command the CP uploads motor configuration information stored in the particular IRMA unit s configuration file into the AAC Without this information the AAC cannot c
105. ng system is also available and used to hold the axes stationary Braking is applied by setting bits 4 and 5 for altitude and azimuth respectively on parallel port A Altitude and azimuth motor direction is controlled by DIO output lines 6 and 7 respectively SECTION 2 4 ELECTRONICS GND 5V PB2 PBI AZ ENABLE PB 6 y ALT ENABLE PB 7 y to from PR A ALT AZ RCM 2100 ALT BRAKE M MOT 1 CTRL ALT MASTER AZ BRAKE E 5 otors MAXON 1 Q EC ALT DIR _PA 6 123 DIR m 4 GND Ippen 1 2 DAC ALT AZ V2l6 eshe cs p r SCLK _PA 2 4 y scLK MOT 2 CTRL AZ S E vilo SPEED o DIN PA 8 p DIN Pal h aT CCW E gt 2 al E MAX5223 VDD 3 a voD 2 A A gt Maxon 1 E c pa COW Q L o LS7266R1 uy RS485 p rg les ALT sa SI DRIVER dl 00lg Poo apio o CH AL 25 ana Fee 2 oll Pol D1 CH B lg 24 SJCH B Jgs 25 S dal 202 bp 02 gt STR pl 2 INDEX ESS lt PRELD El Ss E 03 lg 03 Lp os Bee 3 VoD 4 HES D4 lg Pos 8p D4 DE li lq po ae E 05 p05 Op 05 E6M N D D 06 q PD6 103106 TT 07 Lg Por Lp 07 AZ Y in CH A g20 SICH A a E 2 Di 1RK el DRIVER z Wla WR PiL em S CTRL ha PE2 13 CTRL GND 1 S OE X wy XYSEL Lg PES 1p XYSEL cs 15 E A gt GND L a
106. nge of observable zenith angles from 0 to 70 38 which represent an airmass range from 1 to 3 The prototype was unable to move in the azimuth direction and therefore was fixed when the instrument was deployed The initial design was improved by obtaining a detector with higher responsivity and using a bandpass filter that more closely matched the desired spectral range Fur thermore a complete mechanical and electrical redesign was undertaken to make the unit more compact autonomous and robust As part of this redesign the scanning mirror was removed and the parabolic mirror was mounted in a compact box such that the system SECTION 2 2 OPTICAL SYSTEM 14 could be pointed in any arbitrary direction by use of an Alt Az mount as shown in Figure 2 3 By replacing the original 45 off axis parabolic mirror with a 90 off axis parabolic mirror the optical system became more compact 2 2 1 Parabolic Mirror Design A 90 off axis paraboloid was chosen as the focusing optic to avoid obscuring part of the optical beam by a secondary mirror assembly and therefore maximize the optical throughput The focal length of the optical system was determined by the size of the detector element 1 x 1 mm and the atmospheric sampling criterion which was chosen to sample a 10 m patch of sky at an altitude of 1 km thus defining a field of view of about 1 100 radian Primary mirror paraboloid represented here as a lens with area A pa subtend
107. nics s soss sr agoa sogon apg RE ee ie ee eR ne wees DAL OverviedW 2 24 64 8464 ne Ha he be wee pe eh aa Ho PCNA 2 e ek ode eee eR eee EEE EEE eae 2 4 3 Diamond Systems Emerald MM DIO 2 4 4 Pre Amplifier 22 0584 484 4 0b Bed S eee aR HE EGS 2 4 5 IRMA motherboard 000000022 2 4 6 Cryocooler Controller 2 0 0002 eee eee 24 15 Alt Az Eletronics is daa hes Boe Boats eee a ee a 2 0 SUMMaAy roate oe A ee eS e ee ee eee A vi iv ix xii Oo NS So NR RA TABLE OF CONTENTS 3 IRMA System Software 3l BISTO Lee heme fo Bey eee Ges dy ae ee ee pe eine eden ae Sete we ee Ae 3 2 Rabbit based IRMA 2 a S32 Command Processors x tree da Na we we 3 2 2 CP MC Communication A 2 e osa os Bee ka a eR Ask ai eed as 3 2 3 Master Controller ssar rad aree et 3 2 4 MC AAC Communications 0 00000 eee 32 0 VS e 3 3 1 Porting the Master Controller 0 2 200504 3 3 2 Autonomous Control i a a a 3 3 3 AAC Updates es ra q s oaia a pte R a 34 Qu e Server sse ss sacda goan gb maaana a A a E SG 319 Aatto Laske ope a a aS dos Sn He a ee oe Oe a 3 5 1 Software Structure as aa 3 6 Housekeeping Tasks 2 2 ee 3 1 Summary se seca de febaee ed ethe ke bee PASE a KRW S IRMA Control and Data Analysis Software 4 1 IRMA Control Interface o e ALL Software Structure ea s mesma bee ee pe a a ee 4 1 2 Requirements and Execution o e 4
108. no greater than 1 x 10715 mbar cm s in order to allow the cryo cooler to operate at its target temperature of 70 K Also mounted inside the vacuum chamber on the detector block are the bandpass filter and a cold stop The window of the vacuum chamber is made of anti reflection coated ZnSe 16 The cold stop was designed to restrict the viewing angle of the detector to the size of the mirror However during tests in the SECTION 2 3 MECHANICAL DESIGN 19 lab it was found that there was some spill over of the beam 5 1 2 which was corrected by adding a second aperture to the exterior of the ZnSe window The optical chopper is mounted such that the blades pass in front of the exterior aperture Due to the fast optics of the IRMA if the unit is pointed directly at the Sun the focused light is sufficiently intense to burn a hole in the infrared filter and damage the detector Unfortunately this occurred twice on previous IRMA models To prevent this from recurring a solenoid operated shutter was incorporated into the design such that if the unit moves within 15 of the Sun it blocks the beam This Sun shutter is activated independent of software by a phototransistor mounted in a small hole on top of the unit on axis with the IRMA field of view it may also be controlled via software The electronics are mounted on the side opposite the optical system and include the PC 104 the IRMA motherboard and the pre amplifier The electronics are de
109. ocess the data A time series plot of the Chilean data is shown in Figure 5 18 while Figure 5 19 shows the correlation of the boxes with the correlation coefficients listed in Table 5 3 The correlation coefficients do not show a signif icant improvement over those calculated for the other calibrations because the coefficients do not account for an absolute calibration The coefficients were then applied to the data obtained in Lethbridge and are shown in Figures 5 20 and 5 21 Reassuringly there was found to be excellent correlation between the boxes Table 5 4 and virtually no offset SECTION 5 1 CALIBRATION 94 6h x Unit 10 A Unit 11 J o Unit 12 a Sh ci i E E J gt 3 4 A a J 31 2 454118 10 2 454119 10 2 454120 10 2 454121 105 Julian Day Figure 5 18 A time series plot of the three co located IRMA units in Chile The gaps in the data correspond to calibration sequences These data were derived using the calibration coefficients calculated using the Chilean sky relative to Unit 10 as the calibration source E a RR R R 7 A 55 7 er J E t z C J a 40 7 C x 10 vs 11 E A 10 vs 12 J E o 11 vs 12 J 35 7 Y 24 As o o o o a lp ro dp 2 3 5 6 4 PWV mm Figure 5 19 Inter comparison scatter plot of PWV data from the three co located IRMA units in Chile analagous to Figure 5 11 These data were derived using
110. ock The dual opposed pistons minimize the vibration and acoustic noise from the cooler while doubling the cooling power SECTION 6 2 FUTURE MECHANICAL AND OPTICAL DEVELOPMENT 106 Reducing the vibrations should also improve the signal to noise ratio of the detector The SAX101 cryocooler has the advantage of using the same cooler controller as the current version of IRMA so the hardware and software interfaces to the cooler will be identical to the present model of IRMA Unfortunately however the SAX101 cryocooler is larger than the previous model which necessitates mounting the cooler such that the detector is at an 45 angle to the mirror as shown in Figure 6 1 Mounting the cooler in this way may pose challenges to aligning the detector 6 2 2 Larger Primary Mirror As mentioned fitting the new cooler into the restricted dimensions of the IRMA module necessitated installing it at an angle While mounting the cooler in this way poses some challenges it allows for the installation of a larger 12 5 cm diameter primary mirror Having a larger primary mirror increases the collection area and therefore increases the amount of flux incident on the detector The increase in flux will increase the signal from the detector and thereby increase the signal to noise ratio 6 2 3 New Weather Shutter Design The 12 5 cm diameter mirror requires a larger viewport for IRMA Thus the design of the weather shutter needed to be modified The re
111. often deployed at remote sites where there is little physical interaction with the unit For this reason the units must be robust but it is possible that a mechanical failure may occur in the field IRMA is designed to minimise single point failures such that when such mechanical failures occur it may still be possible to operate the unit and obtain meaningful data In previous versions it was only possible to move the Alt Az mount to an absolute position i e Elevation 70 Azimuth 350 West of North However thanks to recent updates to the Alt Az Controller AAC 3 3 3 commands were added which allowed the Alt Az mount to make relative movements i e move clockwise 30 in azimuth In April 2007 while a unit was deployed atop Mauna Kea in Hawaii failures oc curred which previously would have made normal operation of the unit impossible First a problem with the elevation drive was discovered which was traced to a failure in an optical switch During an initialisation routine the unit would not recognize one of the elevation limits This meant that the elevation initialisation routine could not be completed and therefore the standard move commands could not me used At this time the initialisa tion involved moving one direction assigned counter clockwise CCW until the limit was reached then rotating in the other direction assigned clockwise CW to the CW limit SECTION 5 2 FIELD PERFORMANCE 97 With the CW limit sw
112. oing so will result in the call to set the RTC to timeout and fail Once the serial channel has been opened the command rte set date_time will read the current date time from the GPS receiver convert it to 1980 epoch format and write it to the RTC The GPS emits a time synchronization signal every second Date time is written to the RTC as soon as this time synch signal goes high One concludes the RTC setting session by closing the serial channel to the GPS GPS The GPS family of commands involves the reading of time date and location information from the IRMA MC s GPS receiver board The GPS is interfaced to the MC by means of a 4800 bps serial channel Consequently all commands to the GPS must be preceded by issuing the command to open the GPS serial channel gps serial open After the transaction with the GPS has been completed the GPS serial channel should be closed SECTION A 3 IRMASCRIPT LANGUAGE DEFINITION 124 using gps serial close Datetime is read from the GPS receiver using the command gps read date_time The data returned is contained in a colon delimited string year month day hour minute second Epoch time returned in 1980 epoch format is read by calling gps read epoch_time Latitude longitude data is read with the command gps read lat_lon Data is returned as a colon delimited string IRMA This family of commands is used to perform system level activities on the IRMA system as a whole The statement irma state off
113. ompared and showed a good correlation 104 Chapter 6 Future Directions 6 1 Overview Since the development of the prototype evolution has occurred in all aspects of the IRMA design The latest improvements have been discussed in my thesis This chapter outlines some of the areas where further improvements will allow IRMA to better perform its task of measuring atmospheric water vapour Improvements in software will make the unit more autonomous reliable and robust Modifications to the mechanical and optical design will improve the signal quality reduce noise and also contribute to the robustness of the instrument 6 2 Future Mechanical and Optical Development Work is currently underway to build the next generation of IRMA which features a redesign of the mechanical and optical systems Figure 6 1 Fortunately the electronics and control software are largely unaffected by these changes SECTION 6 2 FUTURE MECHANICAL AND OPTICAL DEVELOPMENT 105 Figure 6 1 New IRMA mechanical optical design Courtesy of Brad Gom 6 2 1 New Cryocooler The redesign was driven by the the decision of Hymatic the makers of the cooler to discontinue the cryocooler model NAX025 001 used in the previous design The new cryocooler from Hymatic model SAX101 was chosen for the next generation IRMA The new cryocooler provides many improvements over the previous model It made use of two dual opposed linear compressors mounted on a central bl
114. on of a block of statements This structure is equivalent to a for loop that increments from 0 to n GOTO The most basic flow control mechanism is the goto statement When the IRMAscript inter preter executes goto label statement program control jumps to the IRMAscript statement immediately following the label labelName statement Using GOTOs as a form of pro gram flow control can lead to unstructured unmanageable code However in the context of IRMAscript whose scripts tend to be quite short less than a printed page long the issue of structured GOTO less programming is not important Given the relatively prim itive flow control mechanisms available in IRMAscript GOTO allows the programmer to develop sophisticated flow control within an IRMA script With labels the use of a colon after the label name is optional A 3 6 Input Output Commands PRINT Feedback from an executing IRMAscript can be directed to the console or shell by means of the print command The argument to the print command can be a literal or a variable In its most simple form print can accept bare literals either text or numeric which is inconsequential to IRMAscript as it is a typeless language If a string literal enclosed in double quotes is passed as the parameter the user can format the output mixing variables SECTION A 3 IRMASCRIPT LANGUAGE DEFINITION 120 and literals together The only stipulation is that each item in the string whether literals
115. or access to the system The MC based on an 8 bit Rabbit Semiconductor RCM2100 17 was located inside the instrument itself It handled the communication with the CP and interfaced directly with the hardware to control the instrument The Alt Az tasks however were off loaded to another Rabbit processor RCM2010 the AAC which acted as a slave to the MC It was located in the Alt Az mount and was connected to the MC via a serial link through the umbilical cable which can be seen in Figure 2 3 This system was programmed by a previous graduate student Ian Schofield and formed the basis of his Master s thesis 24 Though most of my work involved the PC 104 based system it was necessary for me to understand the Rabbit based system in order to maintain the Rabbit based IRMA units currently deployed I was also able to make some improvements to this system the most significant of which was to add a software package 25 which allowed for remote reprogramming of the RCM2100 The Rabbit based IRMA is briefly described below to aid the reader in understanding the current design 3 2 1 Command Processor The Command Processor was built around a standard Linux installation At the core of the CP was the JRMAscript interpreter irmaExec pl Whether commands are sent to the IRMA unit directly through the command line 3 2 1 3 through the Control In terface GUI 84 1 or programmed to execute automatically using AutoTasks 83 5 all commands u
116. ound the world and will continue to improve and evolve Fin 112 Appendix A IRMAscript A 1 Overview An IRMAscript statement is structured simply For commands that directly con trol IRMA a command statement consists of a command type followed by two modifiers and zero to fifteen arguments The command type and its two modifiers define a unique command The arguments are provided in order to pass information pertinent to the com mand to IRMA Most command statements with the exception of the Alt Az moveto slewto commands have zero or one argument In addition to IRMA commands IRMAscript pro vides variables data assignment arithmetic system commands such as reading system time looping mechanisms lists and flow control and console I O They do not follow the same command structure described above Whitespace is used to delimit or separate each of the elements command type modifiers and arguments that make up an IRMAscript statement Whitespace can consist of spaces or tabs Each statement must terminate with a carriage return Only one state SECTION A 2 IRMASCRIPT LANGUAGE SUMMARY 113 ment can appear on one line which precludes IRMAscript from being a free form language such as C or C IRMAscript is caseless it does not matter whether IRMAscript state ments are written in upper or lower case letters Within the interpreter all statements are converted to uppercase A 2 IRMAscript Language Summary
117. rection applied to the astronomical observation Previous studies have shown that at a wavelength of 20 um water vapour is the primary source of emission in the atmosphere 12 Thus by measuring the emission at these wavelengths the detected flux can be converted to water vapour abundance through the use of an atmospheric model From any site there are natural variations of temperature and pressure which must be taken into account when generating the model Using the range of local temperatures and pressures common to the desired site BTRAM is run in a batch mode to generate a series of model files These files are combined to form a lookup table relating flux to PWV for the given range of local meteorological conditions The operator uses the atmospheric flux values obtained by IRMA along with the concurrent temperature and pressure to determine the corresponding PWV value from the lookup table The final accuracy of the retrieved PWV value depends not only on instrumental uncertainties but also on uncertainies in the SECTION 1 4 SUMMARY 10 atmospheric model For this reason where possible nearby radiosonde data are used to generate the models used by IRMA Error analysis studies have also been performed on the models and show that the contribution to the total error budget is driven by the accuracy of the scale height and the adiabatic lapse rate A detailed analysis of how the key parameters affect the resulting PWV output from the mo
118. referred SECTION 2 5 SUMMARY 29 2 5 Summary The IRMA hardware can be divided into three main components optics mechan ics and electronics Each of these areas have evolved significantly since the first prototype The optics have been fitted into a compact box which can either be mounted to the side of a radio antenna or in a specially designed Alt Az mount Through the use of custom electronics embedded computer systems now allow for remote autonomous control of the IRMA unit 30 Chapter 3 IRMA System Software 3 1 History As the IRMA hardware has evolved so has its software In the beginning the software consisted of a simple C program running on a laptop which controlled IRMA through its parallel port Subsequently a Common Gateway Interface CGI was incor porated into the original program so that it could be operated over the Internet with a web browser Though it was still controlled through the parallel port this enabled remote operation However with the mechanical redesign IRMA became a more complex system necessitating a complete software redesign 3 2 Rabbit based IRMA With the mechanical redesign the IRMA became a distributed system running over three processors As shown in Figure 3 1 these were the Command Processor CP the Master Controller MC and the Alt Az Controller AAC The CP was a standard SECTION 3 2 RABBIT BASED IRMA 31 desktop PC running Linux which provided the operat
119. rence between the fits and the measured Tiss were computed The combination of temperature sensors resulting in the lowest o was chosen Later a more SECTION 5 1 CALIBRATION 83 general approach was taken An initial fit was performed using all available temperature sensors The temperature sensors with the largest coefficients from the fit were selected In each case the dominant sensor was the same one selected using the statistical method However the secondary sensor differed in each case 5 1 3 2 Secondary calibration A secondary calibration is performed to relate the internal blackbody to the LBB Calibrating the internal blackbody to the LBB allows for periodic verification of the calibra tion while the unit is operating remotely For this secondary calibration the data from the internal blackbody during the calibration sequence are used shown in green in Figure 5 7 Once the primary calibration has been completed the infrared flux of the lid is determined similar to Equation 5 7 by n V Vo gt Ci Dd i 0 CLBB W 5 8 Dostip The correlation between the flux associated with readings from the embedded temperature sensors and the effective flux of the internal blackbody is determined by fitting to the equation Peso Po ta bi e W 5 9 where Po is an offset term and Po are the calculated flux associated with the lid diode temperatures and cj and co are their respective fit coefficients
120. reset using the reset command adc init resynch The last step involves configuring each of the ADC s four channels with the command adc set csr arguments The following list describes each of the CS5534 IRMAscript functions in depth adc init resynch Calling this command puts the ADC s serial port into a known state When using the ADC for the first time it is recommended that this command be called in order to ensure that the ADC will successfully accept serial commands At low level this command serially writes 15 bytes of the value OxFF followed by 1 single byte valued OxFE adc init reset This command resets the ADC and sets its fundamental parameters At low level the reset command sets the RS bit in the C 5534 s configuration register which has the effect of forcing a system reset adc set offset channel value Set offset command allows the user to configure each of the CS5534 s four input channels offset registers Channels 1 through 4 can be specified while the value field can accept offset values ranging between 2 and 274 The offset value represents the fraction of the input span that must be applied to the output value of the ADC to shift it up or down Offset values must be defined in ADC units For example an offset of 255 refers to a positive offset of 255 274 of the ADC s input span ADC channels configured for taking unipolar samples have an input span of 274 while channels configured for bipolar mode hav
121. rface Options Engineering Queue for Unit 12 Unit 12 Refresh GMT 21 06 12 Status Chopper O Connected to BOX12 Chop Off Shutter Runnin ai Black Body bbOn 2 Refresh 3 Sat 60Hz Notch Off scan Y Sun Sensor 455Hz BP Off sun Shutter Shutter Sync RTC GPS Cooler OK e DK Set RTC ATC 21 06 09 Stop Scan Manual GPS 00 00 00 iai Pause Azimuth U mg 5 1 April 2007 aes noe gt Elevation uy 2007 04 01720 06 40 2007 04 01721 06 30 Hour 1 2Hour 10 min Set Temperature Humidity 7 unit ping Sh C Se E Gd 1000 ALT_OFFO x d AzZ_OFFO x d rj 000 06 12 CODLERIS Ve d TEMPO x d SET_POINTO x d RIC_TIME2008 5 1 21 6 9 x d Ending program x d BOX 12 x d 1 06 12 21 06 10 chop 1 I shutter 3 bb 0 notchl Figure 4 1 A screen shot of the IRMA Control Interface GUI IRMAscript and execute it from the command line 3 2 1 3 or through Auto Tasks 83 5 After Ms Smedes finished her work term I learned the code and was responsible for maintaining it I subsequently improved the performance such as improving the data handling routines for the graphs and allowing for more flexible cryocooler monitoring 4 1 1 Software Structure The IRMA Control Interface is a high level control and monitoring program writ ten in Perl Tk Tk i
122. rument 13 2 2 A schematic of the equivalent optical system of IRMA 1 14 2 3 A rendered model of IRMA in its Altitude Azimuth fork mount 17 2 4 A photograph of the electronics side of the IRMA module The optical com partment is located immediately on the other side of the vertical wall shown I The IMAGE eme sea ee a RRR a ES g 21 2 5 A Block Diagram of the new configuration of the IRMA Master Controller showing how the PC 104 interfaces with the various subcomponents of the En s eas ea Ro ewe ae ee Hed Owe See Es 24 2 6 A photograph Alt Az electronics board with Rabbit RCM 2010 microproces sor attached centre left 22044 ac 2 534 264808566498 2 weds 25 2 7 A Block Diagram of the IRMA Alt Az Controller showing how the Rabbit RCM2010 interfaces with motor control system 26 3 1 A System diagram for the Rabbit based IRMA 32 3 2 A flowchart of Rabbit based system software o 35 3 3 An overview of the IRMA directory structure 36 3 4 A summary of the IRMA network communications handshaking sequence 39 3 5 A schematic of the IRMA master control software task structure during scan NPs pee eo we ba Soa a ea V Bow ee Gd a eh we Sew Re S n q d a 41 3 6 The description of the IRMA serial communications packet string 42 3 7 IRMA serial communications protocol between the Master Controller and the slave Alt Az Controller i a
123. ry calibration source in the field The complete calibration procedure was done in three steps SECTION 5 1 CALIBRATION 81 Typical calibration routine 1 170 Primary 1 165 Secondary e 1 160 g gt J155 O 8 1 150 A 1 145 1 140 1 2 3 4 5 Hours Figure 5 7 A typical calibration sequence performed in the laboratory The black line is detector voltage the red square highlighted sections correspond to measurements of the primary calibrator LBB the green asterisk highlighted sections correspond to the measurements of secondary calibrator lid blackbody 5 1 3 1 Primary calibration The primary calibration is responsible for determining the influence of internal unit temperatures on the signal voltage Originally a relationship involving the effective temperature of potentially contributing components of the signal voltage was used in the analysis However the signal voltage is linear with respect to radiant flux not effective tem perature Therefore a special scaling function was required to linearise the signal voltage to temperature relationship Once the effective temperature of the sky was calculated this value could be converted to flux using Equation 5 2 The scaling function was potentially SECTION 5 1 CALIBRATION 82 a source of error in the flux calculation because it had to be pre calculated for a range of temperatures If the actual sky temperature was found to be outside the pre calc
124. s Previously the AAC was designed to accept movement commands only after the Alt Az mount was initialised This was due to the fact that the input parameters to these commands were required to be an absolute position i e an elevation of 30 However if a limit switch fails the IRMA unit will be unable to complete the initialisation For this rea son functionality was added which would allow the operator to move to a relative position i e increase elevation 20 These additional commands provided invaluable flexibility and proved useful when problems were encountered in the field see 85 2 3 3 3 4 Offset handling The AAC is capable of applying offsets to each direction to allow the unit to adjust the value to North in the azimuth direction and the horizon for elevation The position data are then returned in accurate altitude and azimuth coordinates In the Rabbit based system SECTION 3 4 QUEUE SERVER 51 the values were set in the AAC but could not be queried A copy of the values were stored in the MC and it was this copy that was returned when the value was queried In the PC 104 based unit the MC is not as closely coupled to the AAC Since irmamc is simply a program running under Linux on the MC it may restarted or the computer may be rebooted without affecting the AAC which would cause incorrect offset values to be returned The MC was modified to forward Alt Az offset queries to the AAC which would respond with the c
125. s a powerful graphical toolkit which has been ported to Perl a popular systems scripting language for the rapid development of graphical interfaces Various pre programmed graphical elements called widgets are included such as control buttons text boxes radio buttons etc Like most GUI programs Perl TK programs are event driven meaning a main loop waits for events such as clicking on a button to occur and then SECTION 4 1 IRMA CONTROL INTERFACE 58 Queue for Unit 12 Status Connected to BOX12 Running bbOn 2 Waiting Refresh 3 Figure 4 2 A screenshot of the Queue Status area of the IRMA Control Interface showing the blackbody ON command running with the Refresh command waiting in the Queue executes associated funtions called callbacks Some events are handled automatically within the Tk core while others are associated with custom written callback routines 26 The primary purpose of the IRMA Control Interface is to provide remote real time status monitoring and control of the IRMA unit The IRMA Control Interface can also be used to monitor the data from the unit and setup and control AutoTasks remotely The functionality of the Control Interface is described in this section 4 1 1 1 Queue Status The Queue Status area is shown in Figure 4 2 Here the operator is able to view the connection status of the selected unit Also shown is the script currently executing as well as a list of scripts waitin
126. s stored as an Excel file and shows the association of each command with is corresponding numeric code By default is is named IRMAscript x1s and is located in the IRMA HelperProgs directory In addition each IRMA unit has a unique configuration file box_ lt box number gt cfg stored in the IRMA Config directory The configuration file shown in Table 3 2 provides SECTION 3 2 RABBIT BASED IRMA 37 Table 3 2 An example unit configuration file which is used to initialise instrument parameters FOO A A RK RK a 2k 2 2k 2k k kkk k 2006 01 01T00 00 00 IPaddress 128 171 116 72 Data_port 10072 Cooler TR282 Board 1 Dummy calibration data for this time period CalibrateLow 77_7 74e6 CalibrateHigh 319_6 82e6 FO I I E A a E a 2K 2K 36 2k kk kk kk kkk kkk kk 2006 08 09T15 00 00 Unit returned from Hawaii IPaddress 142 66 41 40 ElevGearReduction 128 AzimGearReduction 128 BeltReduction 8 MinMotorRPM 500 MaxMotorRPM 25000 MaxGearRPM 8000 elev_kProp 10 0 elev_kInteg 1 0 elev_kDeriv 1 0 azim_kProp 1 0 azim_kInteg 1 0 azim_kDeriv 1 0 FORO I I A RK kk a 2k kk kk kk 26 2k 2k k 2k 2k kk the CP with the IP address and data port of the master controller and supplies the gear reduction ratios servo parameters and detector calibration constants to the MC The file is broken into parameter blocks which are delimited by lines of repeating asterisks A time stamp appears at the head of the block which establishes the dat
127. scribed in more detail in 2 4 A narrow compartment at the top of the IRMA unit contains a weather shutter The shutter can be positioned via a lead screw to cover the main aperture in case of inclement weather The shutter serves a dual purpose as a blackbody calibration source is mounted on its underside When the shutter is closed the detector views the blackbody which is used to provide radiometric calibration 2 3 2 Alt Az mount An Alt Az mount is provided to allow full range of motion when the IRMA is not attached to the side of a radio antenna In this case power and Ethernet are connected to the bottom of the Alt Az mount and then provided to the main IRMA box through an SECTION 2 4 ELECTRONICS 20 umbilical cable Alt Az articulation is powered by two Maxon EC167129 low noise 50 W brushless DC motors each coupled with a Maxon 1QEC50V digital motor control unit Axis rotation for each direction is geared down substantially by a 1621 1 gear head An additional 8 1 gear reduction is provided by belts connecting the motors to their respective axes For the azimuth direction the motor is mounted in one arm of the mount and connected via a belt to a fixed central gear about which the unit rotates The unit is capable of rotating 370 about its azimuth axis this range being defined by optical limit switches restricts the rotation to prevent damage to the wiring connecting the articulating parts The altitude motor turns a ge
128. separates each line into tokens Tokens may be either literals strings variables constants or reserved words words which are commands of the language The statements of each line are then stored in an associative array or hash table using the generated line number as the key to access the statements The parser is responsible for determining if a sequence of tokens conform to a language statement Since IRMAscript uses a finite automaton or finite state machine and does not allow for nested statements the parser in irmaExec pl is greatly simplified Basic syntax for hardware commands consists of 3 tokens Command Modifier field 1 Modifier field 2 These are followed by n parameters as needed This structure helps make the language more readable and also allows the commands to be divided into families For example the commands relating to the chopper wheel are shown below CHOP READ STATE CHOP STATE ON CHOP STATE OFF The three token hardware commands are converted into 3 digit command codes and embed ded in network command packets The command set is stored in an Excel spreadsheet which is parsed using the Perl module Spreadsheet ParseExcel The parser stores variables in a hash table and also handles nested loops Scope checking and type checking are generally combined in a compiler Since IRMAscript is typeless and all variables are considered global that is visible throughout the program this stage is not included
129. ser or programs sending commands to irmamc Whenever a IRMAscript is received it will automatically be entered into the queue Additional commands will be added to query the status or modify the queue Having the Queue Server tightly integrated into irmamc will greatly simplify the design of programs which will interact with irmamc 6 3 1 2 Improved Autonomous Operation Currently irmamc is only capable of executing IRMA scripts Autonomous control is carried out by additional programs which generate scripts and send them to irmamc such as Autotasks 3 5 With the adoption of the PC 104 based platform all of the system software is executed on the same processor so these programs can be integrated into irmamc This would reduce overhead and allow for more efficient operation In addition with the integration of the Queue Server and Auto Tasks into the Master Controller the system would be written almost entirely in C perl which runs slowly on the PC 104 due to the increased overhead of the Perl interpreter would be used only for very simple housekeeping tasks 6 3 1 3 IRMAscript Modification IRMAscript allows for flexible control of the radiometer However in most cases instrument control is handled by short scripts which handle an individual task This is due to the fact that only one script can be executed at a time if a large complex script were executing all other scripts would be required to wait The subsequent delay could be
130. shortcut to some directory where IRMA data is stored lt boxNumber gt is the IRMA unit s identifier number YY YY is the year in which the data log file was created and YY YY MM DD is a year month day time stamp This directory format organizes data files chronologically according to the particular unit new iso timestamp Create a time stamp string conforming to the ISO date time format YYYY MM DDTHH mm SS sss Where Y Y Y Y refers to year MM to month 1 12 DD to day 1 31 HH to hour 0 23 mm to minute 0 59 SS to second 0 59 and sss to milliseconds 0 999 The symbols T and are delimitation symbols A 3 3 Variable Manipulation ASSIGN Assign a value to a variable The source of the assignment can be literal or another variable Literal values can be numeric or strings Strings can be defined with or without enclosing double quotes When quotes are used 1t is permitted to include whitespace in the string INCR DECR Increment or decrement a value contained in a variable This operation does not work with literals as literals cannot have values assigned to them EVAL Perform arithmetic operations and assign results to a variable This command precedes a simple arithmetic statement involving two operands and one operator The operations SECTION A 3 IRMASCRIPT LANGUAGE DEFINITION 118 available are addition subtraction multiplication division modular division and expo nentiation The operands can be li
131. sk is performed is given in seconds by the delay parameter A sample configuration file is shown in Appendix C 4 3 6 Housekeeping Tasks There are various other smaller scripts which assist IRMA As mentioned in 1 3 2 to determine PWV the local temperature and pressure are required Using the same Perl SECTION 3 7 SUMMARY 55 module used by the weather monitoring task IRMA Weather a script logger pl was written to generate a weather file which records the local temperature pressure and hu midity This file is stored with the data files The default directory is IRMAdata IRMA_ lt Box_number gt lt year gt lt year gt lt month gt lt day gt Due to the limited amount of storage space available on the Compact Flash drive 1 GB data must be copied to a remote server and removed from the IRMA unit before the available storage space is exhausted The program data_sync pl was written to handle this task It synchronizes the data to a remote server using rsync 31 Then if the rsync was successful it removes any IRMA data more than one day old Both of these tasks are executed on a regular basis using the cron 30 scheduler 3 7 Summary The IRMA control and communication software has evolved significantly over the course of the project From its humble beginnings when IRMA was controlled through the parallel port of a laptop the system has evolved into a complex distributed system involving three processors a PC and
132. sky temperature to determine the effective temperature one sensor in the center and the other located near one of the edges Under ambient conditions the blackbody temperature is uniform across its surface but when heated a temperature gradient was found to exist as shown in Figure 5 2 which shows a thermal image obtained with a Fluke Ti20 infrared camera kindly made available by Fluke Electronics Canada The temperature profile shows that the centre of the blackbody was measured at 50 0 2 C 323 K while the edge was 46 0 2 C 319 Initially it was unclear how the two temperature readings from the heated black body related to the effective temperature of the surface Since the blackbodies on individual IRMA units are custom made each has unique gradients which lead to errors in their re spective calibration This results in different flux values being reported by two IRMA units SECTION 5 1 CALIBRATION 77 Figure 5 4 A photograph showing how the IRMA field of view was mapped using a soldering iron observing the same atmosphere which is clearly unacceptable Moreover this error is fur ther amplified by the extrapolation process described above To illustrate this point Figure 5 3 shows the resultant extrapolation error due to a 3 K error in the effective tempera ture of the hot calibration source At ambient temperatures the blackbody is assumed to be in a state of thermal equilibrium and therefore ac
133. ta direction depending on whether data are being written or read Since the chip select active low line is connected directly to ground there are four pins used for controlling the LS7266 read write con trol data and X Y axis select connected to pins 0 to 3 on parallel port E respectively The control data line LS7266 pin 13 selects whether data registers low or control registers high are selected Read and write LS7266 pins 16 and 14 respectively are active low and are used for enabling reading or writing to the chip Similarly the X Y axis line LS7266 pin 13 determines whether the azimuth axis counter low or altitude axis counter high is selected The LS7266 requires three input lines from each encoder two analog inputs la beled A and B and an index marker The azimuth axis encoder is connected to the X axis inputs of the LS7266 while the altitude axis encoder is connected to the Y axis inputs The A B inputs for both axes are enabled by setting the A B input enable lines LS7266 pins 18 and 28 high These lines are permanently tied high on the AAC main board Each encoder contains a unique index mark which generates a pulse signal when detected This signal from the altitude and azimuth encoders are connected to pins 1 and 19 of the LS7266 respectively and can be used to either reset the counter or load a preset value While the index mark could be used for initialisation as the mechanical limit switches are currently p
134. te considerably slower than the minimum motor RPM due to the motor s gear box and drive belt Be aware that it is impossible to perform movements slower than the minimum motor RPM For performing movements slower than the minimum achievable speed there is the slew_to command SECTION A 3 IRMASCRIPT LANGUAGE DEFINITION 132 altaz slew_to dms axis alt_d alt_m alt_s az_h az_m az_s speed Usage of the slew_to command is identical to move_to What differs is the range of speeds available and the fact that movement is not servo controlled When a speed less than the minimum achievable speed is selected slew_to goes into stepping mode where the given slew path is broken up into sub degree one encoder unit steps The axis or axes rotate for the duration calculated from the slew path length and the requested speed Mention should be made about the relationship between offset angles and axis moves Offset angles for each axis are measured from the counterclockwise limit switch in the clockwise direction The AAC rotates to the requested angle to which is added the currently defined offset angle The offset angle should be considered as zero degrees Altitude angles less than the offset angle are reported as negative angles while azimuth angles less than the offset wrap at 360 degrees because the azimuth axis has the ability to rotate a full 360 degrees Full rotational movement allows for the possibility of destination angles that lie beyond the
135. te it SECTION 5 1 CALIBRATION 79 Figure 5 6 A photograph of the the large reference blackbody LBB which is mounted in a wooden frame that can be accurately positioned atop an IRMA unit Inset in the image is a representative mapping of the embedded temperature sensors In the next phase of the investigation 11 of the 16 temperature sensors were placed throughout the optical cavity to study the temperature distribution inside the IRMA unit in an attempt to determine the cause of the stray radiation At the same time a new calibration procedure was developed 5 1 3 Improved Calibration Method To address the issues arising from the box to box variance in blackbody emis sion a new process was adopted which involved the development of a large external black body LBB The LBB would serve as a primary reference to which all of the individual IRMA blackbody sources would be calibrated To overcome the limitations of the internal calibration sources which by virtue of the available space were small found to exhibit un expectedly large temperature gradients and only contained two temperature sensors the reference blackbody was designed with a higher power heater and a much larger surface SECTION 5 1 CALIBRATION 80 to avoid edge effects The large blackbody had 16 temperature sensors embedded into its surface arranged in the pattern shown in Figure 5 6 to allow for the accurate mapping of its surface temperature profile T
136. ter set oc_reset command BB The bb state setting command enables or disables the blackbody shutter heater The heater is turned on by calling this command with the argument on while off turns the blackbody heater off The state of the heater can be read by calling the command bb read SECTION A 3 IRMASCRIPT LANGUAGE DEFINITION 122 state The return value 1 indicates that the blackbody heater is on while a return value of 0 zero indicates that it is off CHOP_MOTOR The 450 Hz chop wheel is controlled and monitored by means of the chop_motor family of commands The chop wheel is turned on or off by the chop_motor state setting command where setting can be set to on or off chop_motor read state reads the chop wheel status returning the value 1 if the chop wheel motor is on and 0 if it is off To read the chop wheel s angular speed in revolutions per minute RPM IRMA counts the number of interrupt pulses from the chop wheel over a selected period of time Consequently one cannot simultaneously perform a data collection scan and measure chop wheel angular speed To perform a measurement one turns the chop wheel on then issues the command chop_motor state measure_rpm_on The next step is to wait for a period of time using the wait seconds command The longer the time spent in angular speed measurement mode the more accurate the average angular speed value will be Wait periods ranging between 30 and 60 seconds provide adequate resul
137. terals or variables but the result must be assigned to a variable A 3 4 Delays WAIT Delay execution of the script by N seconds N can be a real value ranging from 0 to some arbitrary value A 3 5 Flow Control DO WHILE While loops repeatedly execute a block of statements while some arbitrary condition is logically evaluated to be true With do while statements the condition is tested at the end of the block as opposed to the beginning of the block which occurs in while loops A do while loop in IRMAscript opens with a do statement and closes with a while condition statement Any number of IRMAscript statements including other do while loops can be included in this block There is no limit to the number of do while loops that can be nested within one another The condition can take two forms a simple comparison involving two operands or a compound conditional statement that logically ANDs or ORs two comparisons For exam ple a simple conditional statement takes the form x lt y while a compound conditional is structured x lt y or a b Four kinds of comparison are available less than lt greater than gt equality and inequality Logical ANDing and ORing can be specified in a compound SECTION A 3 IRMASCRIPT LANGUAGE DEFINITION 119 conditional using the symbols and and or Do not use the symbols amp amp or to perform logical evaluations REPEAT ENDLOOP Repeat executi
138. the cryo cooler s set point temperature in Kelvin The return value is a floating point number cryo read mode Returns an integer code representing the operational mode of the cryo cooler controller cryo read curr_temp Returns the current temperature in Kelvin of the cryo cooler s cold finger The return value is a floating point number cryo read osc_freq Returns the cryo cooler s oscillation frequency which is the frequency of the piston inside the cold finger The oscillation frequency is expressed in cycles per second Hz cryo set manual_mode This command sets the cryo cooler into manual mode which powers the cryo cooler down cryo set set_point temperature This command sets the desired temperature of the cryo cooler s cold finger This command will successfully execute only when the cryo cooler is in man ual_mode cryo set auto_mode The cryo cooler begins to cool when this command is received Cooling is a gradual process taking roughly 30 minutes according to the cryo cooler con troller s internal configuration settings When target set point temperature is reached the controller will maintain this temperature as long as it is in auto mode ADC Control of the Cirrus CS5534 Delta Sigma ADC is handled by the ade family of commands Before A D conversions can be performed the ADC must be first initialized using the SECTION A 3 IRMASCRIPT LANGUAGE DEFINITION 126 resynch command adc init resynch then
139. tion This refers to the name of the site where this given IRMA unit is located Cooler The model number of the cryo cooler associated with this given IRMA unit Board An ID number which identifies the IRMA motherboard associated with this given IRMA unit CalibrateLow This is the ADC count when the IR channel measures the unpowered shutter blackbody calibration source cold Calibration of the calibration target in hot and cold states powered and unpowered relates the IR measurement with a temperature reading from the same target CalibrateHigh This is the ADC count when the IR channel measures the powered up hot shutter calibration source See the description of CalibrateLow for calibration details C 2 Perl Module Installation Additional Perl modules must be installed for the IRMA software to run properly Perl modules can be installed using the Comprehensive Perl Archive Network CPAN using the following command sudo perl MCPAN e shell This will bring up a cpan gt prompt At this prompt simply enter the command cpan gt install lt module_name gt SECTION C 2 PERL MODULE INSTALLATION 137 To install custom IRMA modules the files simply need to be copied into the IRMA IRMA directory C 2 1 System Software Perl Modules The following modules must be installed on the PC 104 for the system software to operate properly 10 Handle 10 Socket Net Ping Socket 10 Spreadsheet ParseExcel Time Loca
140. tive parameters within autoTasks conf The cryocooler logging function is executed at a time interval specified in autoTasks conf and the information is be stored in a cooler statistics file 3 5 1 3 Daily Tasks AutoTasks also allows for the scheduling of scripts This provides functionality similar to cron 30 on linux systems At a regular time interval set by daily_tasks delay SECTION 3 6 HOUSEKEEPING TASKS 54 in autoTasks conf AutoTasks will check if a scheduled task is ready to run Task file must be placed in the IRMA Tasks box_ lt box no gt directory AutoTasks will first check if a special task file for the current day is present which is determined by the filename i e 2007 05 12 task If no such file is present the file default task will be used Any changes to the task file while 4utoTasks is running will be detected automatically and the updated scheduled tasks will be loaded A sample daily task file is included in Appendix C 3 3 5 1 4 Sky Maps Sky maps are obtained by scanning the entire sky vvith the IRMA unit to provide an all sky map of the PWV Though skymaps have been performed manually in the lab the process has not yet been automated in AutoTasks It is present in Auto Tasks as future development 3 5 1 5 autoTasks conf The main configuration file for Auto Tasks is autoTasks conf The different tasks within AutoTasks can be activated by setting their respective on parameter to 1 How often the ta
141. ts After the wait period has passed one takes IRMA out of measurement mode with the command chop_motor state measure_rpm_off then reads the resulting value with the command chop_motor read rpm RTC Current time on the MC s RTC is read with the command rtc read date_time Returned is a colon delimited string containing the current date time year month day hour minute second As an example February 12 2005 at 3 37 49 PM would be returned as 2005 2 12 15 37 49 Months range from 1 to 12 days range from 1 to 31 hours range SECTION A 3 IRMASCRIPT LANGUAGE DEFINITION 123 from 0 to 23 and minutes and seconds range from 0 to 59 Current time can be read using the rtc read epoch_time command This com mand is convenient for timing events within an IRMA script as it returns a 32 bit unsigned integer number representing the current time as the number of elapsed seconds since mid night of January 1 1980 User defined date time can be set using the the command rtc set arbitrary_time ISOtimeString An ISO formatted date time string has the following format YYYY MM DDThh mm ss where Y Y Y Y is a four digit year MM is month 1 12 DD is day of month 1 31 hh is hour 24 hour format mm is minute and ss is second The punctuation contained in this format the T and dashes must be left as shown The second method of setting date time using the GPS receiver requires that the serial channel to the GPS board be opened Not d
142. two 8 bit Rabbit Semiconductor microprocessors I joined the project and was able to successfully deploy the latest generation of IRMA based on a PC 104 This latest system added significant computing power and flexibility to the unit by integrating a 32 bit processor capable of running a standard Linux operating system At the same time the complexity of the overall system was reduced by eliminating the need for an external PC Improvements were also made to the software to make it more robust and autonomous 56 Chapter 4 IRMA Control and Data Analysis Software The IRMA control and data analysis software resides on a remote computer This software allows the operator to control any IRMA unit and to view and interpret the collected data 4 1 IRMA Control Interface In order to control the IRMA unit the operator has the option to either connect to the unit remotely via ssh and run desired commands from the command line as described in 3 2 1 3 or use the IRMA Control Interface The Control Interface shown in Figure 4 1 provides a simple intuitive graphical user interface GUI to the unit The GUI was written by a previous undergraduate student Ms Amy Smedes and is able to provide real time status information for multiple IRMA units It also allows for basic control of the unit For more complex controls however the operater must write the desired functionality in SECTION 4 1 IRMA CONTROL INTERFACE 57 IRMA Gontrol Inte
143. ulated range extrapolation errors would be present when the scaling function was applied The primary calibration is accomplished by observing the LBB shown in red in Figure 5 7 Due to stray radiation within IRMA the detector voltage is not simply a result of the flux from the LBB P pp but also contains some additional flux The detector voltage can in general be expressed as a linear combination of all possible sources of emission V Vo ctes Ores gt ci Di IVI gt 5 6 1 0 where V is photodetector voltage Vo is the offset term cep is the coefficient associated to the flux emitted by the LBB and c are the n coefficients associated with the flux emitted from the n areas inside the IRMA unit used to account for stray internal radiation This equation can be rearranged to solve for Pipi V Vo gt ci Dy im W 5 7 CLBB The offset and coefficients were calculated by performing a linear least squares fit to the flux emitted by the LBB Selecting the internal temperature sensors to use in the fit was a non trivial task Any of the 11 temperature sensors located in the optical compartment can potentially be used to account for the stray internal radiation Originally a statistical approach was taken to determine which of the 11 sensors should be used All relevant combinations of sensors were fitted to a set of primary calibration data using Equation 5 6 The standard deviation of the diffe
144. urrent value 3 4 Queue Server The Rabbit based IRMA could only handle one TCP connection at a time If a second command was sent while the first was executing the second command would receive a Connection Refused error This problem was solved by writing a non proirty queue server Programs which wanted to send a command to an IRMA unit would first submit the request to the associated queue server The queue server would then notify the program when it was free to execute the RM Ascript This was only of limited use however because any commands executed directly from the command line would bypass the queue server and potentially cause Connection Refused errors With the increased flexibility of the PC 104 irmamc automatically queues up to 10 simultaneous commands in the TCP stack making the functionality of the queue server redundant At present the queue server is still included on the PC 104 based system because it was very tightly integrated into some of the software such as Auto Tasks which will be discussed in the next section SECTION 3 5 AUTOTASKS 52 3 5 AutoTasks AutoTasks helps the unit run more autonomously by monitoring the system and executing tasks 3 5 1 Software Structure AutoTasks is written entirely in the perl scripting language It is designed to run in the background as a service As shown in Figure 3 10 AutoTasks is simply a loop which performs various tasks at defined intervals These tasks include
145. value altaz set az_offset offset_value The set offset commands are provided in order to allow the user to define vir tual fiducial points thus avoiding the necessity of physically orienting IRMA s SECTION A 3 IRMASCRIPT LANGUAGE DEFINITION 131 fiducial the axis limits to external physical references such as zenith for el evation or North for azimuth By providing an offset value defined in optical encoder units IRMA s AAC calculates all axis moves relative to the offset po sition instead of the default physical limit Axis offsets is the angle between the physical limit and the position where the physical reference is determined to be The default offset value for both axes is zero altaz state poslog action AAC position logging is controlled using this command Three separate activ ities can be performed log initialization log enabling and log disabling Upon AAC start up the position log is allocated zero filled and its index pointer is pointed to the first element in the position log array This action should be explicitly called before using the position log by using the log_clear con stant in the action parameter One begins logging an axis movement by calling this command using the log_enable constant Logging is stopped by using the log_disable constant altaz state halt Stop movement immediately in both axis altaz state reboot Perform a soft reset or reboot of the AAC software running on the Alt Az controller
146. vered that the output voltage of the detector is also dependent on other factors such as internal box temperature There fore changes were made to the calibration method which modified the conversion algo rithm and therefore new calibration software was required IRMA PWV was written in IDL Interactive Data Language because it is well suited for large arrays of data and its integration of complex mathematical functions such as multivariable linear regression and interpolation functions Built in plotting functions greatly simplify the visualisation of IRMA data IRMA PWV is still in the early stages of developement It is currently a simple program which performs the specific task of converting the IRMA signal voltage to PWV The conversion process performed by the software is described in Figure 4 11 It is envisioned that this software will form the basis of the next generation of data reduction and control software replacing both the IRMA Control Interface and the IRMA Archive interface The proposed future development is discussed in 86 3 2 4 4 Summary The front end and visualisation software allows the user to control the IRMA unit and analyse data remotely The IRMA Control Interface provides an intuitive GUI for real time unit status and data monitoring in addition to remote AutoTasks configuration The IRMA Archive Interface is used for viewing and analysing the IRMA data While 70 SECTION 4 4 SUMMARY
147. ving site and the amount of water vapour in the atmosphere IRMA is designed to have maximum sensitivity at high altitude sites In fact it is not expected to perform well at the lower altitude and thus wetter site above our laboratory For this reason in Lethbridge it is not possible to evaluate the performance of IRMA in the operating conditions for which it is designed Therefore to test the units under more optimal conditions the three units were shipped to Chile In January 2007 I had the opportunity to travel to Chile where we carefully drove the IRMA units across the SECTION 5 1 CALIBRATION 90 es ey as AS S SD Ta Cs E a Sl hi 5 7 E E J E E J S yr J a 47 mi E x 10 vs 11 7 E A 10 vs 12 E 011 vs 12 a 3H 7 Y a DA A i AT AA A A a a 2 3 4 5 6 PWV mm Figure 5 14 Inter comparison scatter plot of PWV data from the three co located IRMA units in Chile analagous to Figure 5 11 These data were derived using the original calibration coefficients calculated during the original test campaign in Lethbridge As can be seen and as expected this site is significantly drier than Lethbridge It can also be seen that the systematic errors are outside our error budget target of 10 for PWV desert in 4x4 trucks to a 3000 m high mountain The units were unpacked and set up side by side as shown in Figure 5 13 Here the testing procedure was repeated and the three units were once agai

Download Pdf Manuals

Related Search

Related Contents

CC-168 online_180813  Manual - Lynx Pro Audio  24016 ガーデンアーチM型 組立・取扱説明書  NA2800,NA2820 NAKAMICHI 2014-9  YES Plus Twelve-channel Air Quality Monitor    owners manuaL CD PLaYers enGLIsH DeuTsCH FranÇaIs  Apparecchiature macchine utensili ed attrezzature, strumenti  

Copyright © All rights reserved. Failed to retrieve file