Version 3.0 03 July, 2003 CanariCam JCI Startup Manual Version 1.0 22 July 2005 Prepared by: James Radomski , Tom Hayward, Chris Packham, and Margaret Moerchen CanariCam Team: Charles Telesco, Chris Packham, Francisco Reyes, James Radomski, Jim French, Kevin Hanna, David Hon, Roger Julian, Jeff Julian, Frank Varosi, Naibi Mariñas, and Margaret Moerchen. Preface -1- Version 3.0 03 July, 2003 The purpose of this manual subdocument is to provide information regarding the use of JEI in association with the GTC CanariCam. CanariCam is a mid-infrared (8-25 micron) camera, slit spectrograph, and polarimeter built by the University of Florida Infrared Astrophysics Group for facility use at the Gran Telescopio Canarias. This document explains the startup procedures for CanariCam and JCI at the telescope, and is for use by experienced personnel. -2- Version 3.0 03 July, 2003 TABLE OF CONTENTS 1. 2. 3. 3.1 3.2 3.3 3.4 3.5 3.6 4. 5. 5.1 5.2 5.3 5.4 5.5 6. 6.1 6.2 1 Instrument Safety x 2 Powering Up Thermal Enclosure Electronics 4 3 Starting JEI 9 Master Panel .......................................................................................................... 9 Bias Panel ............................................................................................................ 10 Temperature Panel ............................................................................................... 11 Motor High-Level Panel ...................................................................................... 12 Detector Panel...................................................................................................... 13 Standard System Test .......................................................................................... 15 4 Shutting Down Thermal Enclosure Electronics 19 5 Troubleshooting 20 Starting a tip Session ........................................................................................... 22 Restarting individual subsystems and agents ...................................................... 20 Power Cycling remotely via Baytech .................................................................. 22 Rebooting EPICS ................................................................................................. 21 Serial-Port Testing of Rackmount Electronics .................................................... 22 6 T-ReCS Safety Issues 24 Instrument Monitoring:........................................................................................ 25 Starting the JEI .................................................................................................... 25 -3- Version 3.0 03 July, 2003 1 Introduction and Instrument Safety This manual outlines the general procedures involved in operation CanariCam on the telescope and is for use by experienced personnel. Section 2 discusses the initial procedures for powering up CanariCam on the telescope at the beginning of an observing run. Section 3 discusses the typical startup for an observer at the start of every night using the CanariCam Java Engineering Interface (JEI) software. Section 5 covers the shutting down of the T-ReCS electronics at the end of an observing run or when necessary. Section 5 highlights general trouble shooting procedures. And finally Section 6 covers TReCS safety issues. In order to protect the instrument when not in use and while JEI is operational, check to make sure: 1. Motor High Level Panel: Window Changer is set to Block 2. Motor High Level Panel: Lyot Wheel is set to Block 3. Bias Panel: Power is Off and Power light is Red 4. Temperature Panel: Heater is Off Refer to Section 3 for more information on individual JEI Panels. Further safety information is discussed in Section 6. 2 Powering Up T-ReCS 2003 May 12, TLH. Updated 2003 June 12. Updated 2003 July 03 (JTR) This section describes how the rack-mounted electronics in the thermal enclosures are powered on and JEI is started. These procedures are generally only applicable at the beginning of an observing run. For nightly startup procedures after T-ReCS has been powered up refer to Section 3. The two racks are named “Instrument Control System” (ICS) and “Analog Signal Processor” (ASP). All figures of JEI panels below display a black dotted ellipse to highlight regions to watch and solid black lines over buttons which may need to be used. In order to make changes on any panel make sure the Panel Lock box in the lower right corner is unchecked. 1. 2. 3. 4. Connect orange power cable to 3-pin connector on ICS rack. Connect power cable between ICS and ASP racks. Connect fibers between racks. Power-on ICS and ASC racks w/ switches at base. The rackmount computers trecsnewton and trecskepler should power-up and boot automatically. If the computers were not shut down properly they may have to be switched on manually. If so seek help from local Gemini staff. 5. Front-panel power switches on desired components must be on. 6. Start a terminal window using a machine and account with network access to the TReCS system: e.g. the trecs account on a Gemini workstation. If network access is not available, the T-ReCS components may be accessed via serial ports by following the procedure described in Section 5.3. Standard network access is assumed here. -4- Version 3.0 03 July, 2003 7. At xterm type ping trecs-2924 – Ping Cisco switch to check connection. 8. Type ping trecsbaytech – Ping Baytech. 9. Type telnet trecsbaytech – connect to Baytech power switch. 10. You will be prompted to Enter Selection, type 1 Enter Selection> 1 – (outlet control) 11. The prompt will change to RPC-3> at this prompt type on 1 RPC-3> on 1 – turn on port 1 12. Then hit carriage return (ENTER) to update display with current amps RPC-3> <cr> – updates display 13. Repeat for ports 2, 3, 4, 5, 8. Don’t turn on port 7, PPCVME, or port 6, the coldhead, which is no longer connected to the Baytech. 14. Wait ~ 2 minutes for the Perle to boot. 15. At an xterm log on to kepler by typing the alias irkepler or type rlogin trecskepler-ctl 16. At the kepler prompt type ping trecsperle to verify Perle has booted before continuing. 17. Turn on port 7; must be done last so PPCVME finds other components upon booting. You may run a tip session (see Section 5.3) on trecsnewton-ctl to monitor the PPCVME boot. 18. Start another terminal window on the Gemini workstation and type ufstartjei to start the JEI. 19. Upon starting JEI the Master Panel will be displayed. Figure 1 Master Panel (19) -5- Version 3.0 03 July, 2003 20. When the PPC boot finishes, the JEI should indicate that the MCE-4 detector parameters are being queried. 21. On the Master Panel confirm that the trecs:heartbeat increases incrementally.Check control data fields and confirm that all read IDLE and that there are no error messages. If there is no trecs:heartbeat or error messages occur, get help or refer to Troubleshooting Section. 22. Exit the Baytech control program with by typing menu and then typing 6 RPC-3> menu Enter Selection> 6 Figure 2 Temperature panel (21) 23. When T-ReCS is first powered up, the Lakeshore 340 agent must be stopped, then restarted with a special argument in order to load the temperature calibration curves. If this is not done, an error message will be displayed on the 340 front panel, the Detector Temperature (JEI Temperature panel, see Figure below) will read 0.0, and turning on the detector power (JEI Bias panel) will not be permitted. Details on starting and stopping agents are given in Section 5.1. If no kepler xterm is running type rlogin trecskepler-ctl (or simply use the rkepler alias) to bring up an xterm on kepler At the kepler xterm type ufstop ufgls340d Then type ufstart ufgls340d –setcrv A few minutes are required to load the curves before the Detector Temperature and trecs heartbeat to begin updating. 24. In the JEI Master Panel, execute Clear-Init-Apply to connect to the Motor, Environment, Observation, and Acquisition agents. The four indicators in the center -6- Version 3.0 03 July, 2003 of the panel should go to BUSY, then IDLE. The success of the connections can be monitored in the tip session. Figure 3 Master Panel (22) (23) 25. If EPICS has just been started or rebooted, a Datum-Apply is required in the master panel. This will Datum all the motors and Apply, but not Latch, the software default bias voltages. 26. Select the JEI Motor High-Level panel and verify that all the mechanisms have stopped at the Datum position. The Window Changer should be at the Block position, which protects the windows. If the window is not at Block, resolve the problem before continuing in order to protect the windows from dust and water contamination. -7- Version 3.0 03 July, 2003 Figure 4 Motor High-Level Panel (24) (25) 27. Set Filter Wheel 1 to Block. 28. See below for JEI nightly startup guide for further instructions on using the instrument for observations. -8- Version 3.0 03 July, 2003 3 Starting JCI This section discusses the typical startup for an observer using JCI at the start of every night. For initial powering up of the CanariCam system refer back to Section 2. All figures of JCI panels below display a black dotted ellipse to highlight regions to watch and solid black lines over buttons which may need to be used. In order to make changes on any panel make sure the Panel Lock box in the lower right corner is unchecked. In order to monitor image frames select Quick Look tool of choice at a Gemini workstation terminal. 3.1 Master Panel 1. At a workstation (e.g. ufnewton), start a terminal window and type ufjci to start the JCI interface. Wait for the text to scroll by until the JCI is displayed. (see Figure 1) N.B. Do not run JCI or JDD on kepler (reserved for agents and data acquisition.) Figure 5 Master Panel (2) 7. 2. On the Master Panel confirm that the trecs:heartbeat increases incrementally. Check control data fields and confirm that all read IDLE and that there are no error messages. If there is no trecs:heartbeat or error messages occur, get help or refer to Troubleshooting Section. 3. Click on clear (to clear the EPICS database), then init (to initialize the instrument (i.e. connect the JEI to the agents to EPICS) and then the apply button. Watch for the control data fields to read IDLE. -9- Version 3.0 3.2 03 July, 2003 Bias Panel 4. Now change to the Bias Panel by clicking on the bias tab (see Figure 2) Figure 6 Bias Panel (5) (8) (6) (7) 5. Click the connect button (even if it reports connected) and then the clear EPICS button to clear outstanding EPICS commands. Confirm connect button is green. 6. Set the array well depth to deep or shallow depending on the particular observation. Then set the detector bias level to medium and then click the apply button. Confirm that Well Depth = 0 for shallow or Well Depth = 1 for deep and Bias Level = 2 for medium. If the Well Depth or Bias Level numbers do not match the current settings, try changing the current settings, (i.e. deep to shallow) clicking the apply button, then changing back and hitting apply. Watch the bottom display for the CAR to read IDLE. 7. Click on the power ON button and ensure that the power button is illuminated green. 8. Finally, verify bias settings by clicking Read All in lower left corner and confirming V-Gate is aproximately equal to -4. (If V_Gate reads ~ -6 power is OFF.) Also confirm Well Depth =1 or X and Bias Level =2 from step 7. - 10 - Version 3.0 3.3 03 July, 2003 Temperature Panel 9. Now check the array temperature (Detector) is below 15K by clicking on the temperature tab to display the Temperature Panel (see Figure 3). 10. Confirm Connected to Agent button is green Figure 7 Temperature Panel 11. Check that the other values are near normal. Nominal values for when the detector is ON/OFF are displayed in Table 1. Table 1 Detector ON/OFF Temperatures Sensor Name Detector Cold Finger Cold Plate Cold Head Grating Motor Pressure Detector Power On ~ 7.3 K ~ 6.5 K ~ 7.0 K ~ 42.8 K ~ 71.1 K ~ 7.5e-8 Torr - 11 - Detector Power Off ~ 6.1 K ~ 5.0 K ~ 6.8 K ~ 42.6 K ~ 71.1 K ~ 7.5e-8 Torr Version 3.0 3.4 03 July, 2003 Motor High-Level Panel 12. Select the motor high-level tab to change to the Motor High-Level Panel (see Figure 4) Figure 8 Motor High-Level Panel 13. To move to any wheel to a named position, click on the drop down menu of the wheel of interest and select the named position. The named location column will change to a yellow background to show a position change has been requested but not yet applied. 14. To move to the requested position, click the move button and ensure that the status column (far right column) stabilizes and reports the requested position. You can also select the Query Status button in the lower right to check the status of the motor positions at any time. - 12 - Version 3.0 3.5 03 July, 2003 Detector Panel 15. Next change to the detector panel by clicking on the detector tab (see Figure 5). Figure 9 Detector Panel (18) (17) (19) 16. Click on the connect button (even if it reports connected) and ensure it is illuminated green. 17. Next select the desired observing mode (e.g. stare), readout mode (e.g. S1R3 (correlated quadruple sampling )), nod-handshake mode (e.g. false) and data mode (save or discard). See the T-ReCS User Manual for a more complete discussion of observing modes. 18. The input parameters required for the observation should then be input to the text dialogue boxes on the left. Parameters will change to a yellow background to show a change has been requested but not yet applied. (if no parameters are input, sensible defaults are provided on the first compute click). 19. Next click the compute button to obtain the adjusted parameters (based on array characteristics) and wait for the control CAR to change from BUSY to IDLE. Finally, click on the configure button to prepare the array for observations and wait for the control CAR to change from BUSY to IDLE. 20. Start the observation by clicking the start button. Ensure that the Frame # in lower left corner starts to increase incrementally (may take a few seconds…). Use Quick Look tool of choice to monitor frames. 21. To abort an observation, click on the abort button. - 13 - Version 3.0 03 July, 2003 22. Note that if any detector input parameters need to be changed (e.g. frametime), execute the following procedure: a. Click abort b. Enter the new input parameters into the dialogue boxes c. Click the compute button to obtain the adjusted parameters d. Click on the configure button to prepare the array for observations e. Wait for the control CAR to change from BUSY to IDLE. f. Click on the start button - 14 - Version 3.0 03 July, 2003 3.6 Standard System Test The following procedures describe a brief test of the system after startup that can be used to confirm all systems are operating nominally. In addition this configuration can be used for characterizing the detector responsivity and noise levels, so that these quantities can be tracked over time. The general procedure involves taking two images and comparing the response of the detector when dark and when illuminated. 1. Go to Motor-High Level Panel Figure 10 Motor High-Level Panel 2. Configure the motors as follows: Table 2 Motor Configuration 1. Sector Wheel 2. Window Changer 3. Aperture Wheel 4. Filter Wheel 1 5. Lyot Wheel 6. Filter Wheel 2 7. Pupil Imaging Wheel 8. Slit Wheel 9. Grating Wheel Black Plate ZnSe Matched PAH-8.6um Block Open Open-1 Open Mirror 3. The Lyot Wheel should be set to Block for the initial Dark frame. - 15 - Version 3.0 03 July, 2003 4. Go to Bias Panel Figure 11 Bias Panel 5. Configure as follows: Table 3 Bias Configuration Detector Well Depth V_DetGrv Bias Level Deep Medium Well Depth =1 Bias Level =2 6. Confirm Well Depth =1 and Bias Level =2 and in the first bias column confirm V_DetGrv = -6.5. If the Well Depth or Bias Level numbers do not match the current settings, try changing the current settings, (i.e. deep to shallow) clicking the apply button, then changing back and hitting apply. - 16 - Version 3.0 03 July, 2003 7. Go to Detector Panel. Figure 12 Detector Panel 8. Set the following Modes: Table 4 Observing Mode Readout Mode Nod-Handshake Data Mode stare S1 FALSE save 9. Configure the Input Parameters as follows: Table 5 Frame Time (ms) 50 On Source Time (min) 0.2 Save Frequency (Hz) 1.0 Max SCS Duty Cycle (%) * Chop Frequency (Hz) * Nod Time (sec) * Nod Settle (sec) * Pre-valid Chop Time (sec) * Post-valid Chop Time (sec) * * Since these inputs are not used at all in stare mode one need not change from the default values. 10. Click Compute, Configure, and Start - 17 - Version 3.0 03 July, 2003 11. Using your Quick Look tool of choice confirm a single 50 ms Dark frame results in ~ 12,000 ADU’s. 12. Repeat the preceding steps with the Lyot Wheel set to Open to take an image of the detector when illuminated 13. Using your Quick Look tool of choice confirm a single 50 ms Illuminated frame results in a total of ~ 36,000 ADU’s. Subtracting the dark frame should result in a total of ~ 24,000 ADU’s. See Table below. Table 6 Detector Response Image Dark 50 ms frame (Lyot = Block) Illuminated 50 ms frame (Lyot = Open) Illuminated – Dark ADU’s ~12,000 ~36,000 ~24,000 Miscellaneous Information: Filter λ =8.6 μm, Δλ = 0.37, transmission = 0.76 Optical transmission = 0.94 ηG = 0.4 718 e- /ADU photon flux at detector = 1.13 x 109 per pixel per sec. 14. At completion of integrations confirm the data is written to disk properly via the Data Handling System (DHS). This done by starting a terminal window using a machine and account with network access to the T-ReCS system: e.g. the trecs account on a Gemini workstation. a. % cd /net/reggie/staging/perm/ b. % ls c. example filename should be S20030604S0025.0.0.fits UT Date File # d. the date shown will be the UT date at the end of the observing night. 3.7 End of Night/Observing Session In order to protect the instrument when not in use and while JEI is operational (e.g. at the end of a night of observing), do the following: 1. Abort any running observation 2. At JEI Motor High Level Panel: set Window Changer to Block 3. At JEI Motor High Level Panel: set Lyot Wheel to Block 4. At JEI Bias Panel: set Power to Off and confirm Power light is Red 5. At JEI Temperature Panel: confirm Heater is Off - 18 - Version 3.0 03 July, 2003 4 Shutting Down T-ReCS 2003 May 12, TLH This is the safe procedure to shut down the rackmount computers trecskepler and trecsnewton and the other electronics when they need to be powered down such as at the end of an observing run. 1. 2. 3. 4. Start a terminal session on a workstation using the trecs account. % rlogin trecskepler-ctl (may need argument “–l trecs”) % su root % shutdown –g0 –i0 (those are zeroes) Stops OS, does not power off rackmount 5. Wait for rlogin to exit. 6. % rlogin trecsnewton-ctl 7. % su root 8. % shutdown –g0 –i0 – same as above. 9. Wait 30 sec for OS to shut down. This procedure leaves rackmounts powered up, but in a safe state for manual power-down, and they will automatically boot when power is reapplied. 10. % telnet trecsbaytech 11. Enter selection> 1 – Select power control menu. 12. RPC-3> off 1 – Power-off port 1. 13. Repeat for all components. Note that the Coldhead (Unit 6) is no longer powered by the Baytech. 14. Power-off both racks (switches on back of racks). - 19 - Version 3.0 03 July, 2003 5 Troubleshooting If problems are encountered with subsystems such as the MCE-4 detector controller, the subsystem may be power-cycled and/or its agent restarted without requiring a full EPICS reboot from the JEI System panel. If these procedures don’t resolve the problem, an EPICS reboot is probably necessary. All of these processes are highlighted below in addition to miscellaneous helpful information. The general process to follow in case of problems is: Go to JEI Master Panel and click clear, init, apply. (see Section 3.1 # 3) If that does not work make sure all the agents are running properly, if not stop/start them as necessary (see Section 5.1) If restarting the agents does not work you may need to reboot EPICS (see Section 5.2) If rebooting EPICS does not solve the problems you may need to power cycle the system (see Section 5.3) NOTE: These represent only brief descriptions of procedures which should only be performed by those very familiar with these operations. 5.1 Restarting individual subsystems and agents 2003 May 9, TLH Updated 2003 July 3, (JTR) To begin, Abort any observation that may be running and confirm that no motors are moving. Then rlogin to the machine running the agents, usually trecskepler-ctl. 1. At xterm type rlogin trecskepler – add -l trecs if not already in a trecs account 2. To list all agents able to be started at kepler xterm type ufstart –l 3. To list all agents able to be stopped at kepler xterm type ufstop –l 8. To rest all agents at once 1. To stop all agents type ufstop –all or ufstop –9 2. To start all agents type ufstart 3. Go to JEI 4. At JEI Master Panel click Clear-Init-Apply to connect Motor, Environment, Obs, and Acq. agents. 5. Go to JEI Detector, Bias, and Preamp Panels and click Connect to connect detector agents After such a stop/start, no datum should be needed as long as EPICS has not been rebooted and the motor indexer has not been power-cycled since the last datum. To recover from MCE-4 problems (e.g. some channels bad): To Power-cycle MCE-4 1. At kepler xterm type ufstop ufgmce4d to stop MCE4 agent only 2. Then type ufstart ufgmce4d to start MCE4 agent only 3. Go to JEI 4. At JEI Detector panel click Connect - 20 - Version 3.0 03 July, 2003 To reset Motor Agent 1. At kepler xterm type ufstop ufgmotord 2. At kepler xterm type ufstart ufgmotord 3. Go to JEI 4. At JEI Motor High-Level Panel click Connect. Again, no motor datum should be required here if the motors were successfully datumed after the last EPICS reboot or Indexer power-cycle. The agent log files are in /var/tmp and can be listed (e.g. for MCE-4) by typing ls – ltr /var/tmp/ufmce4d 5.2 Rebooting EPICS 1. Click on the system tab to access the system window (see Figure 13). Figure 13 System Panel 2. Click on the REBOOT Epics and all Device Agents button to reboot EPICS and the agents. A confirmation window appears, click on the yes button, and then click on the apply button on the system panel. 3. On the Master Panel (see Figure 1), wait until the heartbeat increases incrementally (this will take a couple of minutes…). 4. On the Master Panel, click on clear (to clear the EPICS database), then init (to initialize the instrument (i.e. connect the JEI to the agents to EPICS) and then the apply button. Watch for the control data fields to read idle. - 21 - Version 3.0 03 July, 2003 5.3 Power Cycling remotely via Baytech 2003 May 1, TLH revised June 12 Telnet to the Baytech power control module from a GS workstation. % telnet trecsbaytech Enter selection > 1 – displays status of outlets RPC-3> off 1 – e.g. to turn off TReCSPerle RPC-3> on 1 – e.g. to turn on TReCSPerle (note that the Perle and MCE-4 require ~ 2 minutes to boot after they are powered on.) 5. RPC-3> menu – back to main menu 6. Enter Selection> 6 – exit 1. 2. 3. 4. 5.4 Starting a tip Session 2003 May 1, TLH Run a tip session from a GS workstation to monitor activity on the trecsppc VME crate. 1. At xterm type rlogin trecsnewton-ctl (-l trecs if not in trecs account) 2. Type tip coma – trecsnewton com A port is connected to trecsppc crate 3. monitor session 4. Hit carriage returen (ENTER) for prompt – if no prompt, VXWorks kernel may not be operating correctly, seek help from Gemini staff 5. At xterm type reboot – reboots trecsppc and EPICS 5.5 Serial-Port Testing of Rackmount Electronics 2003 Apr 19, TLH. Revised May 12. The rack-mounted electronics in the thermal enclosures may be connected and tested with a serial-port interface from a PC on which the uf2003 system has been installed. These components are normally checked through their EPICS interfaces, but the serial port procedures may be useful in the event that the EPICS system or network is not operating. The two racks are named “Instrument Systems Controller” (ISC) and “ “ (ASP). 1. Racks should be powered up as described in Section 2. 2. Connect Cisco switch to a PC via a hub. PC should have IP addresses of the rack components in /etc/hosts. 3. Ping Cisco switch “trecs-2924” to check connection. Serial Port Test of Motor Indexer 1. Disconnect indexer serial cable in ISC rack and reconnect it to a PC serial port. - 22 - Version 3.0 03 July, 2003 2. Make sure Indexer is powered up: front panel power switch and “on 5” command to Baytech. 3. Use the program uf2003/annexApp/uftty0. a. Compile if necessary: make uftty b. Set /dev/ttyS0 serial port permission: # chmod a+rw ttyS0 c. Run: uftty0 4. AZ <cr> => 0.00 Indexer A position 5. Valid ports are AZ – JZ; all should read 0.0. Serial Port test of MCE-4 1. Connect MCE-4 serial cable in ISC rack to PC. 2. Power-up MCE-4 (power switch and Baytech as described above) and make sure it boots properly. 3. % uftty0 –cr – Go slow to avoid crashing MCE-4 4. % <cr> – show status 5. % ? – help 6. % status – prints system information. Look for Cycle Type = 0, the default. - 23 - Version 3.0 03 July, 2003 6 T-ReCS Safety Issues T. Hayward, 2003 June 12 Updated 2003 July 3, (JTR) The following describes T-ReCS safety issues of which all Gemini staff involved in maintaining and operating the instrument should be aware. Please note that at this time the procedures are intended to be especially conservative as we are still learning about the instrument. Contacts: Name Tom Hayward Manuel Lazo Charles Telesco Chris Packham Cerro Pachon 205 701 205-718/731 Office 205 619 205 608 352-392-2052 x265 352-392-2052 x259 Home 205-266 Cell 2187045 Web Page: More up-to-date information may be found on the Gemini internal web page www.internal.gemini.edu/Instruments/trecs. Serious Situations: The following situations may place the instrument in some danger or potentially degrade its performance, and should be addressed as soon as possible. One of the above contacts should be informed of such problems at any time, day or night. Interruption of Cryocooler (compressor or coldhead) due to power failure or other problem in the helium system. o Interruptions of up to ~ 30 minute duration are not serious; the dewar will probably recool on its own when the helium system is restarted. Longer interruptions will likely require vacuum pumping of the dewar to recover from outgassing of the getters as they warm up. (Temperature and pressure should be carefully monitored) Interruptions of electrical power to the rackmount computers and electronics. o The rackmount systems should be shut down properly by one of the above contacts in the event of anticipated power outages. Exposure of the hygroscopic KBr dewar windows to ambient air, because either the Window Changer wheel cannot be set to Block or Datum, or the window air purge is inoperative. - 24 - Version 3.0 03 July, 2003 6.1 Instrument Monitoring: The status of the instrument may be monitored as follows. 1. Check the internal dewar temperatures and pressure in the JEI Temperature panel. The temperatures may also be read directly from the Lakeshore temperature controllers (which face down) in the ICS rack. Typical values are: Sensor Name Detector Cold Finger Cold Plate Cold Head Grating Motor Pressure Detector Power On ~ 7.3 K ~ 6.5 K ~ 7.0 K ~ 42.8 K ~ 71.1 K ~ 7.5e-8 Torr Detector Power Off ~ 6.1 K ~ 5.0 K ~ 6.8 K ~ 42.6 K ~ 71.1 K ~ 7.5e-8 Torr 2. Check the Window Changer position in the JEI Motor High-Level panel – it should be set to Block when the instrument is idle. 3. If the JEI is not operating, the Window Changer datum switch and actuator on top of the instrument may be viewed directly. The actuator engages the switch at the datum position, which is also the Block position. If the wheel fails to datum, the position may be set by moving the wheel from the JEI Motor Low-Level menu or even by turning it manually. 4. Check that the ball in the flow meter on the T-ReCS air purge line is displaced several mm from the bottom of its track, indicating that air is flowing normally. 6.2 Starting the JEI See Section 3 for more detail information. 1. At a Gemini workstation (e.g. Chameleon) start a terminal window using the trecs account The password is GStrecs. Type ufstartjei to start the JEI interface. The Master Panel will be displayed. 2. On the Master Panel confirm that the trecs:heartbeat increases incrementally. Check control data fields and confirm that all read IDLE and that there are no error messages. If there is no trecs:heartbeat or error messages occur, get help or refer to Troubleshooting Section. 3. Click on clear (to clear the EPICS database), then init (to initialize the instrument (i.e. connect the JEI to the agents to EPICS) and then the apply button. Watch for the control data fields to read IDLE. 4. Check that the temperatures in the Temperature panel are nominal. 5. Check the motor positions in the Motor High-Level panel - 25 - Version 3.0 03 July, 2003 6.3 Instrument Shutdown Configurations: At the end of every observing session: 1. Abort any running observation 2. At JEI Motor High Level Panel: set Window Changer to Block 3. At JEI Motor High Level Panel: set Lyot Wheel to Block 4. At JEI Bias Panel: set Power to Off and confirm Power light is Red 5. At JEI Temperature Panel: confirm Heater is Off - 26 -