Last Updated:   9/24/2000 (PRL)

Please also consult the Observing Handbook for the Palomar Testbed Interferometer.
See also the new autoaligner writeup.
There are also writeups on the pivot monitor and acquisition in crowded fields.

Right now, before we forget:
How to change things back from dual-star mode to single-star mode

• Check with the lead dual star observer, and if requested, remove the metrology corner cubes from their posts in front of the siderostats.  Frequently, the corner cubes will be left in place between dual-star runs.
• Remove dual-star beamsplitters on the DSMs.  Store them carefully on the DSM breadboard on the side away from the air pipes.
• Remove kinematic base with extra ST filters in front of primary star tracker.
• Remove the CT extraction dichroic on the primary table - also on a kinematic mount.
• Take polarizers out of way of starlight beams in primary arms.  Store next to the small delay line
• Remove the paper masks taped in front of the primary star tracker telescope (they to block metrology light).
• Power off all secondary APDs, please (disconnect power supply cable to distribution box under the table).
• Put primary filter-wheel controller in computer mode.
• Put CT filter-wheel controller in manual mode, put in switch position1 to block laser.
• Reinstall the fiber if removed: insert the alignment corner cube, stop down the boresight beam with the iris, and adjust the coupler so that the spot on the spectrometer side hits the template on the H prism.  Carefully reinstall the the first parabola.  Open the iris, and trim the coupler to get laser light into the fiber (hints: you can move the boresight corner cube to a location right in front of the coupler to get more light.  Depending on how long your arms are, you may find it convenient to look at the fiber output with a mirror while adjusting, or else to adjust from the opposite side of the table where you can see the fiber output directly).  Carefully insert the second parabola, and adjust the spectrometer lens to put the spectrum on the proper area on the array (hints:  with the room light on, you can see an image of part of the secondary fiber chuck, from which you can estimate where the fiber spot will fall.  Also, you can look near the half-mirror for some indication if things are way off: the height of the spectrometer spot should match that of the white light spot, and the proper left-right location has the laser spot from the secondary just grazing the right edge of the mirror). You should also check and adjust the focus of the spectrometer lens with the 2.2 um filter in place (only if it was adjusted when the fiber was removed.
• Turn off oscilloscopes for CT and small dl, CT video monitor, secondary "ftFull" Monitor, and monitors next to autoalignment computer. Leave Wyse terminal on.
• Do a fidSet on small delay line; on completion, turn off motor power on back of delay-line rack in the computer room.
• Be sure Trek stroke  amp (bottom of fringe-tracker rack in computer room) is on for channel A (LDL stroke), off for all other channels.
• To be nice to the next single-star observer, please trim the fold mirrors out to the huts. This often walks off during dual-star observations, and is easier to get back with two people.
• Type NS and 10MS, or as approriate, and reboot everything. Note that, if you have not already done so, you will want to redo fold mirror alignment to optimize pupil (rather than CT), and that you will need to redo the chop calibrate to account for different star tracker filter wedges.

DUAL STAR MODE

Configuring for dual star mode (once per run)

• Install CT corner cubes on their kinematic bases on the posts in front of the siderostats. They are labeled N and S. A lock screw goes in the center to secure. Snug up this screw, but not too tight.
• Install dualstar beamsplitters on the DSMs and make sure they are seated properly! After you've put one in place, try nudging its base to verify that it has seated itself.
• Check that the Primary and Secondary CT injection mirrors are inserted in the beam. They're mounted on translation stages. To reinsert, hold the stage, remove the block that's holding the stage open, and gently let the spring pull it back into position. Don't adjust the micometer. (You may want to wait on this until you do the first step in the next section; /home/pti/writeup/figures/fig10-beamcombiner.gif may be helpful).
• Turn on the small dl motor driver (at back of center rack in computer room).
• Insert polarizers in the primary beams against their stops; they are labeled North and South.  You'll position them more carefully later on.  Secondary table already has these in place.
• Put primary filter-wheel controller into manual mode.
• Place kinematic base with additional ST filters in front of the primary star tracker (secondary table already has filters in place).
• Insert the metrology extraction dichroic on its kinematic mount on the primary table between the main beamsplitter and the white-light lens.
• As appropriate, remove the parabolas for the single mode fiber, and realign without it.  First insert and put a mask on the H prism, insert the boresight corner cube, stop down the boresight beam, and mark where the beam from the coupler beamspslitter falls on the mask to aid in realignment.  Put the K prism back and reopen the iris.  Adjust the coupler to center the beam on the K prism.  Tweak the lens to put the spot in the right location on the camera.You may need to adjust focus, too.

Aligning for dual star mode (once per run)

Verify that the star tracker corner cubes are well aligned on the boresight beam (look at the image of the corner cube vertices on the chop-calibrate beams in front of the star-tracker telescope).

Insert the boresight corner cube, and adjust the metrology injection mirror to retroreflect the metrology beam to the same spot as the laser spot.

Verify that the polarizers are properly centered on the CT beams on primary and secondary, and lock them in place.

When the polarizers are in place, place a small mirror after the polarizers, retroreflect the beam, and then place a paper finger in front of the star tracker telescopes to block light passing through the polarizer from getting into the telescope.

Verify that the s and p polarizations are coaligned for the primary and secondary tables by alternately blocking the individual polarizations on the source plate and observing any centroid motion of the CT image on the monitor. If they are misaligned, it will be necessary to adjust the source plate.

Center two small corner cubes on the metrology beams right after they pass through the polarizers.  Carefully adjust the cardboard mask that is in front of the metrology detector to be centered on the beam to block light leaking around the polarizer.

Using the return from the small corner cubes, adjust the angle of the polarizers to minimize self interference.

Check the left/right translation of the lens feeding the CT detectors.

Verify that the reference signals on the source plate are OK (look at channels 3 and 4 on the scope).

Important: verify system colllimation.  Specifically, 1) home the SSSMs, 2) align the fold mirrors carefully using the alignment cubes on the DSMs, 3) verify that the obscurations overlap on the FSM using the template (see the DSM alignment writeup for more info), 4) block the secondary at the DSM, and adjust the fold mirror behind the telescope to center the primary pupil on the telescope secondary, 5) block the primary at the DSM, and adjust the tilt of the dual-star beamsplitter to center the secondary pupil on the telescope secondary.

Align the small delay line metrology

Turn on the CT laser

Using the signal on the third scope, align the small delay line metrology (this uses a fraction of the primary CT source).  You normally just need to trim the lower mirror.  Use the filter wheel on the CT to attenuate the signal such the signal on the scope is linear.  If the signal is not strong with ND1, align like you're aligning a big dl: align bs cube to go straight down the delay line, align top mirror to overlap spots on detector, align bottom mirror for signal.

Turn off the CT.

NIGHTLY ALIGNMENTS

IMPORTANT:  Because of the problems with secondaryAPD failures, please depower them at the end of each night, and repower them only when it is dark - we have no more spares!

Align CT metrology

• (Every few days)  Check alignment of the pupils w.r.t. spiders of compressors.  Using the iris at the exit of the primary and secondary boresight laser plates, reduce the diameter of the boresight beam to be just slightly larger than the compressor secondary. Adjust the fold mirrors feeding the small delay lines (primary) or feeding the compressors (secondary) to center the annulus leaking around the compressor secondary (you can slide out the injection mirror if necessary to get a better spot). Reopen the irises when done.
• Block the CT.
• Using the corner cube, retroreflect the boresight laser, insert the ND filter (on the primary table setting, "5" on Oriel filter wheel works well), and adjust the lens in front of the microscope to center the spot on the monitor.
• Unblock the CT laser. Note that the filter wheel controller has closed, open, and ND 1, 2, and 3 positions. There is also an additional ND 0.5 flipper in front of the laser. With ND 3 + flipper, adjust the CT injection mirror to center the CT image on the boresight location on the monitor.
• (Every few days) Check the location of the CT spots in the pupils in front of the beam compressors. Use the metal block that has labeled spots (primary S, secondary S), and place it against the provided stops in front of the south compressors. Adjust the spot position using the fold mirror at the exit of the CT metrology plate. Then adjust the tilt of the CT injection mirror as described above to recenter the image, and iterate as necessary.

Adjust the compressors so that the beams go straight down the delay line tracks

• Do this with the CT beam after you have boresighted it.
• Mark the CT location with the delay lines at the front.  Use the tall templates: place one on the North side,  just before the pipes leading out.  Place the south one at the delay line exit; you normally need to take a panel out to get access. Mark where the spots fall with a marker or post-it.
• Move the delay lines to the back, and trim the white knobs on the compressors so the spots are back where they started. Iterate if you had to adjust a lot.

dlSetMode 0,4 / dlSetMode 0,4 will move the passive / active delay line to the front
dlSetMode 1,5/ dlSetMode 1,5 will move the passive / active delay line to the back
dlSetMode 0,1 / dlSetMode 1,1 will stop the passive / active delay line

Alignment of CT metrology to huts with autoaligner

• Boresights off. Alignment Corner cubes out (primary and secondary table).
• CT on.
• Have someone stay at the autoaligner console in the lab and someone else go out to the North Siderostat.
• Check that the alignment corner cubes are also out at siderostats.
• Put the picomotor controller in manual mode:
  • Turn the picomotor driver off.
  • Unplug the gray ribbon cable (lying on the ground).
  • Turn the picomotor driver back on.
• With the picomotors now in manual mode, and using the intercom, adjust the pupils of the primary and secondary metrology beams so that they are centered on the CT corner cube at the north siderostat.
• Using the picomotors, adjust the south beams so that the CT spots are centered in the targets on the monitor in the lab. If you've done a careful alignment job, there's usually a metrology signal after you adjust picomotors 1-4 to center the N spots on the monitor.  If there's no signal on the scope, or just self interference, don't start adjusting yet.
• Then inside the lab adjust the south (primary and secondary) fold mirrors to maximize the signal: Increase ND as the signal level increases.  When you're done, you should ideally have some signal at an ND of 1.5 to 2.0.
• Remove the flipper with ND 0.5 (if you used it)
• Go to the console of the Autoaligner and hit Set Zeropoint for each of the four beams.
• Turn off the CT.

Coarse alignment of boresight beams to huts (old way: using corner cubes)

• Turn off the CT, and turn on the boresight lasers.
• At the huts, insert the primary and secondary alignment corner cubes (CCs) against their stops; the primary ones go to the right.
• Place the north alignment template before the big delay line at the position marked "north primary align" in front of the compressors (white side toward the delay lines).
• Align return spot to be centered on the hole by tweaking the north primary fold mirrors with knobs or picomotors.
• Repeat with the same template in the north secondary position and adjust the north secondary fold mirrors.
• Repeat on the south side
• Take out CC s at siderostats and install the air tubes.  On the south, install the mask around the pipe entry to the building.



NB:  Go slowly here; if you don't get a signal with minor tweaking, you should probably go outside and see where the spot is.  If you misalign so that the spot doesn't make it down the pipe, it's very difficult to get back with one person.

Align the big delay line metrology

• Move the big delay lines back and peak their metrology (see dlAlign.txt for details).

Align primary and secondary tables

• Insert corner cube and co-align WL source and boresight laser.
• Align WL pixel [0.47 volts] (ftFull, arrayDoWL) and spectrometer [1.02 volts] (arrayDoSpec).
• If the secondary fringe tracker is in ftABCD mode, turn it off (ftOff) so it doesn't interfere with the next step - tweaking the fiber.
• Tweak primary fiber if installed in ftABCD mode. Rotate the switch on the voltmeter off then to DC volts.   Set the light level to zero and note the voltage.  Increase the light level so you have a good signal to work with (if you turn up the level too much, the output will wrap and give negative values).  Click on the yellow button on the voltmeter to lock the range (if desired), and adjust for maximum signal.
• Special notes for secondary table.

The white light on the secondary table is aligned like for the primary table.

However, the secondary dewar now has the prism internal to the dewar, and the spectrometer light gets sent through the slit behind the dewar window. The fixed slit position establishes the left-right spectrum location. The alignment process is to use the up/down knob on the spectrometer lens to adjust the up/down location of the spectrum.

Use the left/right knob on the lens to maximize the spectrometer signal - note that you do not use the (x,y) values displayed on the screeen. Use the value typed out for the spectrometer in ftFull / arrayDoSpec mode.  When you're done, put in the 2.2 um line filter, and obseve where the spot falls. If it's more than 1/2 pixel off of center either left or right, you will need to update the SPEC_PIXEL_START, SPEC_PIXEL_STOP values in the section of the code for the secondary spectrometer, K-band, in ft.ini.master (in /home/pti/config/engineering). For example, if the spot is falling to the right by 1/2 to 1 1/2 pixels, you should increment these values by 1; then type 10MS or 20MS, depending on whether you're in 10 or 20 ms mode; reboot the secondary fringe tracker; and redo the left/right spectrometer alignment.

Computer stuff

• Probably best to a full reboot each night
• Select NS mode by typing NS /home/pti/bin before rebooting.
• Select 10ms or 20ms mode by type 10MS or 20MS in /home/pti/bin before rebooting.
• Reboot the computer system (see reboot.txt for more info):
  • hibernate (in the Sequencer window on the GUI click on the hibernate box).
  • Exit the gui and turn off the three crates.
  • Turn on the three crates (right to left) keeping the reset pressed in as you do each one.
  • Quickly, reset the remote card cages using the "plug 'n power" switches on the wall next to horatio: on-then-off. Also, power cycle the acquisition cameras off-then-on using the remote switches. (The pattern of keys you have to press is shown on the wall above the SOUTH/WEST set of switches.)
  • Make sure the green lights flash afterwards on the GreenSpring card (second from left in leftmost crate) before you bring up the gui; see reboot.txt if problems.
• On harbinger, run "beginNight" from /home/pti/bin after 5:00 pm (0:00 UT) to set up a directory for the night's observation.
• In /home/pti/bin type "dataDisks" to check for available space.
• If you run low on space during the night (a CD holds about 650MB), quit the gui, run beginNight again to create a 98XXXb directory; enter y when it asks if that's what you want to do.
• Bring up the gui on harbinger. Go to /home/pti/bin and type "g &"
• Read the OBSERVER_NOTES ; these will undoubtedly save you headaches later.
• Common_problems.txt (same as common_problems.html on http://horatio) is also a useful reference.

Calibrations

• Stow then Home the primary and secondary siderostats (see troubleshooting, below, if problems.).
• FidSet all three delay lines; wait for them to complete before going to the next step.  When done, put the small delay line into off mode.
• On the Long Active Delay Line set the stroke amplitude and frequency to the corresponding amplitude and frequency (frame rate) displayed in the primary fringe tracker window (usually 3260, 100 Hz for 10 ms mode; 2933, 50 Hz for 20 ms mode), then Send stroke and StrokeSet.
• Do ftCalLow and ftCalHigh for fringe trackers on both tables
• WL off, boresight lasers off, alignment CC out, star tracker CCs covered.
• ftCalLow, primary and secondary.
• WL on, alignment CC in.
• ftCalHigh, primary and secondary.
• Remove alignment corner cubes.
• In the North and South Secondary Star Trackers: set Prefilter Size 5, X Gain 0.05, and Y Gain 0.05.
• Plug in power to the dying APDs on the secondary table.
• Chop calibrate the star trackers on both tables
• You need to do a "Dark Calib" on the secondary star trackers, as we are using the high dark current (ie. dying) ones; do this with CT, boresight lasers blocked, WL off, star tracker corner cubes covered, boresight corner cube out.
• Make sure chopper mirrors are on (bzzzz).
• Remove alignment CC
• CT and boresight lasers off!
• Uncover the star tracker corner cubes
• Turn on WL source and adjust for 50k-100k counts/sec (3.0 volts on Primary, 2.0 volts on Secondary)
• All lab lights out.
• Chop calibrate Primary North and South simultaneously; repeat
• Chop calibrate Secondary North and South simultaneously; repeat
• WL off, cover star-tracker CCs

Boresight the primary and secondary lasers

Do Boresights for both north and south simultaneously.

Primary Boresight

• CT metrology off;  leave filter wheel controller in manual mode.
• WL off, alignment corner cube out.
• Secondary boresight laser off.
• Primary boresight laser on
• In the North and South Acquisition System windows click  "reset" to clear the CCD.
• Wait one iteration and then turn "Off"
• Click "Laser Boresight" on each
• Look at the monitor to see that it is working correctly.
• Wait for it to complete: State: PBS: Done.
• The correct answer should be something near (300, 200).
• (See single-star documentation if you have problems here.  Sometimes you need to use the hardware reset gui button, and worse.)
• Turn the primary boresight laser off.

Secondary Boresight

• CT metrology still off.
• Primary boresight laser off.
• Secondary boresight laser on.
• On SSSMs (North and South Secondary Siderostat windows) click off, retro, and then track. This puts the SSSMs in a mode where they accept commands from the acquisition system.
• In the Acquisition windows hit Secondary Boresight.
• Wait for it to complete: State: SBS: Done.
• Clear SSSM counters: In the Secondary Siderostat windows click Show Detailed Status, and Clear Motor Counters.
• Secondary boresight laser off.

Update some constants

• Primary N and S star tracker windows: Change Acq2St Counts Scale from 8 to 2 (needed for star tracker because of presence of beamsplitter).
• Primary and secondary Fringe Tracker window: set VisCalTime from 25.0 to 5.0 seconds.
• Secondary Fringe Tracker window: if you're in 10 ms mode, set Track Coadd from 1 to 2. If the seeing turns out to be awful, return the value to 1. If you're in 20 ms mode, leave it alone.
• Acquisition rlogin window (rlogin acq): type "starBoresightMaxStep=25"; otherwise, star boresights after the first will be limited to a change of 2 from the previous one (default for single-star observing).
• Observing window: Set the zenith angle limit in the observing window to 45 deg.
• If you haven't rebooted, you may want to put the search step size on primary and secondary back to 4.4 um.
• Note: currently, setting sum low limit on the secondary star trackers to 30 works better.
• Note that these values do not persist if you reboot the system

Activate the Autoalignment System

We now adjust the fold mirrors so that the CT beams accurately hit the CT corner cubes.  This will result in a slight shear of the starlight beam, but it's more important to get the CT right.

There is a filter wheel on the CT metrology table. It has an open, shut, ND 1, ND 2, and ND 3 setting; there's an additional ND 0.5 flipper in front of the laser. For initial alignment use the open setting, and add ND as the signal improves to keep the detector in its linear region.  The oscilloscope displays the CT metrology signals: channels 1, 2, 3, and 4 are primary and secondary unknowns and primary and secondary references.  Normally, you only need to look at channels 1 and 2; leave scope on alternate.  Channel 1 is at 110 kHz, channel 2 at 440 kHz. Triggering on the references, channels 3 and 4,  lets you see the phase motion in the measurement (if the phase doesn't change when you do this, you're only seeing self interference).  Triggering on the signals themselves is convenient as long as you know you aren't seeing self interference

• Make sure you have homed the SSSMs as requested above; otherwise secondary CT can vignet.
• Alignment corner cubes out
• Block boresight lasers (primary and secondary).
• Put CT filter wheel in manual and select the open position.
• Check for vignetting.
• You have already set the zeropoints of the autoaligner, so it should come back when you turn things on.
  • Check that the picomotor driver is off.
  • Plug in the grey ribbon cable on the floor.
  • Turn the picomotor driver back on.
  • Turn the loop control ON.
  • Wait for the beams to recenter.
  • Check the metrology again using ND filters, if you wish.
• Set the thresholds of the Autoalignment system to about 50 (?).
• Put the CT filter wheel to blocked and in computer mode.

If you are paranoid and want to double-check that the CT pupils are still nicely superimposed on the CT cornercube. You might also check (sometimes works) using a small CC after the polarizer to retroreflect say, North, back to the detector. If South is not too badly misaligned, the extra gain should give you an adequate signal to peak up the South fold mirror. Now move the inside corner corner cube to retroreflect South and peak up the North fold mirror.

Baseline stars

For fall 2000 we have /home/pti/schedule/dual2000.cat
Create Star List, Browse Star List, filename: /home/pti/schedule/dual2000.cat, Load.

NOTE: When observing in single star mode, always make sure that you have only one star in the entry (i.e. no secondary), or you won't find fringes. This is because the LDL will go to the expected fringe position of the secondary FT, which differs from the primary.

If you do them after dual-star, be sure to fidSet the small delay line and put the primary fringe tracker search step back to 4.4 um.

Zero point calibration

• Enter the same star as both primary and secondary. Click Create Star list, browse catalog, and load current schedule or FK5 as appropriate. Type in the same star for primary and secondary (eg. for Sept. 2000 at 7:30 pm try HDC174881 or HDC182488), set the integration time to 10000 sec, add star, send list to sequencer.
• Make sure boresight laser, and WL are off, and ST corner cubes are covered.
• Block CT, leave in manual mode (this is so it doesn't interfere with the star boresight step below).
• Make sure you've done: fidSet on all three delay lines; strokeSend, strokeSet on long active; chop calibrate on all four star trackers (CT must be off); calLow and calHigh on both fringe trackers; primary and secondary laser boresights.
• Click dual observe.
• Wait for all four star trackers to enter track mode.  On the N and S acquisition systems, click star boresight; when done, click use star boresight.  During this time, the long delay lines should have slewed to their final position.  (Hint: if the secondary tracker is taking forever to find the star, turn it off, and change "SID Sprl Srch X(and Y) Step" from 2 to 8; restart search; restore to 2 when done.  This changes the step size it used when doing the spiral search; it's normally kept small on secondary to avoid cross-locking). Click dual observe off.
• Wait for primary-primary fringes on both primary and secondary.  This will establish their lock offsets.
• On the primary and secondary fringe trackers, update the search pos step size from 4.4 um/step to 0.55 um/step - this will minimize interactions between the fringe trackers, and with a good lock offset, the slower search rate should not be a problem.
• Dual observe off.  CT should be in computer mode.

Starting dual observe

• For dual-star observing, create a list with just one dual-star entry in it.  When only one entry is in the list, the sequencer will repeatedly cycle through the list, going p/p, p/s, p/p, ...
• Load a catalog (ie. /home/pti/schedule/dual2000.cat).
• Select a primary star (ie. HDC201091) enter it into the Primary box.
• Select a secondary star (ie. HDC20192) and enter it into the Secondary Star box.

61 Cyg: HDC201091 (primary) and HDC201092 (secondary).
16 Cyg: HDC186408 (primary) and HDC186427 (secondary).
• Enter an integration time.  Typically we use 130 sec (default).
• Add Star. Send list to sequencer.
• Be sure the short delay line is in off mode.
• The following steps are probably unnecessary with the autoalignement system:
  • Click Dual Observe.
  • Wait for to the delay lines to arrive at about the right position for the target star. Wait for the passive and active delay lines to stop slewing. Then turn dual observe off. It's now time to tweak the CT metrology manually using the picomotors.
  • When the CT is aligned, put the CT filter wheel to blocked and to computer mode.
  • Resume dual observe - click Dual Observe
  • At the end of every fourth or fifth triplet or so of data, the system used to prompt you to tweak the metrology. You would have needed to turn dual observer off, tweak the CT metrology per the instructions above on Fine alignment of metrology beams at huts, and then resume dual star.

Running dual observe

• CT filter wheel to blocked and to computer mode.
• Run dual observe.

• The signal is displayed on the scope in the lab.  Fluctuating signals mean it's past time to re-tweak the metrology.
• Look at the line SDL-CT in the observing window. This is the difference  between the small delay line metrology and the CT.  It's in units of mm.  It should remain constant to ~10 um during a pp or ps.
• Look at the constant-term metrology window.  Record the value of the metrology at the first primary-primary, and compare it to the second primary-primary.  If these aren't the same, the metrology has glitched, and the data will not be useful for that triplet.

Sequencer parameters

• bSeqDoMult - do the pp/ps/pp/ps sequence (default = 1)
• INumMultiplesToDo - number of ps in sequence above (default = 3-5).
• iSeqDualTrackDivisor,  which is how much shorter pp is compared to p-s; defaults to 2.  It can be changed as desired if you want a shorter pp time).
• bSeqSkipAcq.  defaults to 0; if 1, the sequencer doesn't try to acquire but just turns on the Star Trackers. This usually works fine, although occasionally one of the star trackers will lock on the wrong component.
• (n/a) iBackgroundsToSkip.  Default is 3.  This is a counter of how many iterations to go before doing a background. It gets decremented, and is initially set so that the first time around  it will do a background, then it will do one every third time.
• (n/a) bSeqSkipBack will skip backgrounds entirely.
• bSeqAlwaysDoSecondaryFidSet. Default is 1.  If set to 1 the system will do an SDL fidset every iteration. If not it will only clear the laser counters.

RTVIS and end-of-night processing.

• rtvis works in dual-star mode - it's a useful diagnostic to have running
• /home/pti/vis/dodual 123 will process the data for 99123. If the data is not in /home/pti/bin/data, setenv DATA_DIR actual_location. You will need to first copy over /home/pti/vis/dualstar.ini.  If you have multiple directories, you can use setenv DATA_DIR /home/dataX/99ddd/:/home/dataY/99dddb" before running dualstar.
• /home/bode/pti/dualNight/dualNight will do a nightly pipeline reduction, similar to endNight for single star.  Load both cdWriters with blank CDs.  If you have more than one directory, you will be prompted in the morning to add more CDs.  See cdBurning.txt for notes on labels and checking for errors in the disks.
• Please!!! Depower the secondary APDs when done, and block sunlight from getting into the detectors: we don't have any more spares, and are currently using noisy ones.

Troubleshooting

• Verify the metrology is working correctly.  The CT metrology signals are displayed as the + and - entries in the CT metrology delay line window.  When the big delay line is moving at sidereal rates, the delay rates for it and the + and - entries of the CT should be the same.  This means you have a valid CT signal for both tables.  The small delay line uses its own (local) metrology signal.  The way things are now configured, the CT doesn't even have to work to get dual-star fringes. However, without CT, the data is not very useful.
• Checking for proper sequencing.
• The Motor Only delay line should go into Slew and then Off.
• The Long Active delay line should go into Slew and then Track.
• The system should acquire (if enabled), and the 4 star trackers should lock up.
• The small delay line should fidset, then Slew and then Track.
• The search for fringes then begins, and the Long Active Delay Line will be commanded to an offset to look for fringes.
• If things are going normally at this stage, you should expect to see the following:
• The Long Active delay line is in Track mode, with a HP jitter <100 nm
• The Short Active delay line is in Track mode, with a HP jitter <100 nm
• The fringe trackers go into Search
• The fringe trackers eventually go into Lock.
• Note: the secondary fringe tracker depends only on the main delay line metrology.  The primary fringe tracker uses the small delay-line metrology.  As mentioned above, the CT metrology is no longer used in real-time.  You need to check separately that it's working properly.
• Check periodically that data is being recorded.  In principle, you should be warned automatically if there is a problem, in which case you must stop observing, quit the gui, and then restart it.  But it wouldn't hurt to check a few things...
• In /home/pti/bin/data type du -k . and repeat, so you see the files changing size.   You can also:
• Check that the Log is being updated correctly. If the Log file is not being updated then the data is probably not being recorded. You need to scroll to the bottom of the window to check this.
• Check the time stamp of the data files in /home/pti/bin/data. When were they last updated? Do an "ls -l". The time stamp should be no more than a few minutes old  on .FA* and .DL*  if you are fringe tracking.
• [You could also run dualstar to check that all is well.  On horatio, go to /home/pti/vis and type dualstar -s 98:DDD where DDD is the current day number]
• Check periodically for errors in the scroll window of the observing window.   Errors like BUF_ERROR are OK (really!).  Errors like BUS_ERROR or errors associated with DATA_HANDLER are not (for latter, restart gui); for former, see common_problems.txt.

If you find data recording stops without being notified by the system, please note it in the log.
 
• GUI Freezes du to sequencer hang

This can happen after you've been doing observations for a long period. To recover from this error, do the following:
• Exit the gui; exit normally works. If not, kill -9.
• Type DoSystemHibernate in the sequencer rlogin window, or on any subsystem - it's a global commmand. Please record the bus error message with the access addresses so we can fix it.
• In the sequencer rlogin window, or on the terminal (position A), reboot sequencer with ^X.
• Once the Sequencer is up, set the switch box to Star Tracker and turn off all four star trackers by typing stop. You need to do this, otherwise the sequencer will be too busy to do an RPC connect. Also, switch over to siderostat and type stop.
• Restart the gui and reload the star list. Note that the SSSMs will likely be in wrong position, and you won't get a secondary metrology signal until you restart dual observe.

 
• You get photon count too low messages

For star trackers 0, 2 (North Primary, South Primary), you can make Acq2StCountsScale smaller.  However, this will only take effect after any subsequent acquisitions.  For more immediate gratification, for any star tracker (including 1, North Secondary, and 3, South Secondary), you can just reduce the value of sumLowLimit (upper right hand corner of constants window).  This is the light value below which the star tracker starts complaining.
 
• The CT metrology drops out in the middle of a run of data

This has already been discussed as part of normal observing.
 
• The  signs of the CT and long delay line metrologies are  different

Assuming it was right initially,  the  N and S primary polarizers have probably been reversed.  This should only be an issue at the start of a run.
 
• Can't find fringes on both primary or secondary

First check that the long delay line metrology wasn't blocked by doing fidCheck on the long delay lines(fidCheck doesn't work on the small delay line).  If the fidCheck value is more than a few 10's of microns, do a fidSet.

Verify that you have sent and set the stroke.
 

• Can't find fringes on primary

Possibly the fidSet occured when the CT was blocked, or the CT got blocked subsequently. Redo fidSet on the small dl.
 
• Can't find fringes on either primary or secondary

It's possible that the star trackers are confused and locking on the wrong components. In this case, you need to reload a star list with a single star in it to get the offsets right.  You can check this by looking at the star tracker photon counts displayed on the observing window.
 
• (This should be fixed too)Centroid for the boresight was (580,380)

Note that if the centroid for the boresight was (580,380) then the acquisition simply latched on a pixel in the bottom right corner. So recover by any one or several of the following.
• Reset.
• Off, then Hardware Reset in the constants window.
• Camera power reset using the remote control buttons
• Go to the Sequencer window and check Hibernate. rlogin acq, Ctrl-X.
• Call Mark.
Note that if your name is Mark the last item is of no help at all.
 
• Star Trackers refuse to lock on the star

This can be caused by badly misaligned pupils.  Make sure that the CT signal you're seeing on the scope is real, and not self interference, by triggering on the references (channels 3 (primary) or 4 (secondary)).  If the primary or secondary traces are fixed in phase, you've only got self interference and the pupil is probably way off.  You'll probably need to turn on the boresight laser and head to the outside to check the pupil location.
 
• SSSM home gives error

Infrequently, this will fail. First try "Stow" a few times and then try homing again. If this fails you need to go outside and move the SSSM actuators manually so that the mounts are positioned near the trip point of their limit switches.  It's sometimes easier to move the motors if you turn off their power; it's the illuminated switch on the panel on the back of the outside rack; you need to open the door to get access.  You basically back off the actuators (ccw), then rotate forward until you hear the limit switches click.  Turn the power back on and close the door.
 
• Disabling the star trackers for CT diagnostics.
• Power off FSM at piers - not needed if install new ST code.
• Set ST X,Y Gains to 100 (removes integral term)