PAVO User Guide (Updated 22 July 2010)
--- The Detailed Node Listing ---
Aligning PAVO in the Lab
Adjusting PAVO Coarse Alignment
Alignment Out to the Telescope - Skip this Section (2010 April)
On-sky Alignment
Troubleshooting
1 Introduction
PAVO - Precision Astronomical Visible Observations
These notes are designed to be instructions for basic PAVO operation,
assuming familiarity with the rest of the CHARA system. The basic
observing information, moving telescopes, moving carts, etc is not covered
here.
2 Cooling Down the PAVO Camera
The PAVO camera needs to be cooled down at the beginning of the run. Turn
on [CCD Cool] on the power gui. Type the following command into the pavo
server:
> coolon -70
This cools the camera to -70C. "CCD Cool" has to be on
for this to work. The fan is also on, so nothing should break if CCD
cool isn't on; it will just stop at about -55C and not go any lower.
The gain numbers in the current mode only apply to this temperature.
You can read the current temperature from the pavo server. Keep [CCD Cool]
turned on throughout the PAVO run.
If you leave the pavo computer on, the camera plugged-in and "CCD
Cool" on, you don't have to do anything until a PAVO run is over.
At the end of the run, type "coolon 0" into the server and it will raise the
temperature to 0 Celsius. Once it gets to about 0, it is "safe" to
switch anything and everything off without damaging the CCD.
3 Overview of Alignment Steps
There are three different phases to the PAVO alignment. An overview of these
steps is given below:
- Course alignment is done in the lab during the day at the beginning of a
PAVO run.
During this step, we put in the LDCs, put in the PAVO beam
splitters, and align beams 1-3 through the mask and lenslet
("glass" in front of pavo camera). Details of the lab
alignment procedure are given in Aligning PAVO in the Lab.
- Alignment at the beginning of a given baseline setup. In this step we align
the VIS/IR dichroics out to the telescope (flower laser align). This only has
to be done when the telescopes are changed to different beams or after
changing POPs. Details of the flower laser alignment are given in
Alignment Out to the Telescope. NOTE: This step is no longer needed
as of Spring 2010 because we can do M10 alignments using the green lab
alignment laser.
- On-sky alignment of image and pupil planes. This should be done when
switching to a new target/calibrator set in a different part of the sky.
Details of the on-sky alignment are given in
On-sky Alignment.
Note: The alignment on Beams 1 and 2 are done using VisBeams 1&2. The
alignment of Beam 3 is done with VisBeams 3&4.
Note to telescope operators:
Now that we have the green alignment laser, it's possible to check the visbeam,
dichroic, and M10 alignment on a nightly basis.
4 Aligning PAVO in the Lab
These instructions are for aligning the PAVO beams at the start of a run.
After the PAVO beam splitters are put into the path and the optics are aligned,
they shouldn't need to readjusted for the remainder of the run.
4.1 Basic Alignment of PAVO in the Lab
- Open PAVO server and PAVO gui (needed to adjust alignment)
- Home and zero the PAVO motors by typing "zhome 0" and "zzero" into the pavo
server. There is some potential that this will screw with the alignment,
but experience has shown that not doing this can be worse, especially in the
event of power failures.
- Align beams before putting in the PAVO beam splitters
- Align Visbeams using pico 1 (B2&1-1, B2&1-2, B3&4-3)
- Put in the LDC units - align with markings and stops on table, use index card
to make sure the beam is going through
- Align VIS/IR dichroic
- Align out to M10
- Put pavo beam splitters into beam path (at edge of classic table)
- Re-align laser on visbeams target for Beams 1, 2 ,3
- The green laser is now bright enough that we can re-align visbeams directly
using the visbeams target located on the E-table. Double check the dichroic
and M10 alignment afterwards.
- Put corner cubes into the beams
- Open pavo shutters ([B1] [B2] [B3] buttons on PAVO gui)
- Slip target cover over M1 and align PAVO beam splitter using hand-adjustable
knobs
- Check to see if light is going through L1
- Slip target onto L2 and use PAVO gui to do M1 alignment on B1, B2, B3
- Align beam through prism and then through the hole in the mask
- Note: reflections from the front and back of the pavo beam splitters cause
the green laser to have two spots (one above the other). Align the top spot
on each beam through appropriate hole in the mask.
- Use mask in position [3] - should be in place if [mask out] is not pressed.
Use ENG gui to select mask if it is not in the correct position.
(Note - the good position used to be [2]). Looking toward the PAVO camera, the correct mask is the one on the far left.
- Do L2 alignment on B1, B2, B3 (~500 step)
- If beam is aligned properly, can see laser on other side of mask
(use index card to check)
- Mask holes from top to bottom (non-redundant spacing):
- Take off PAVO black box cover
- Align pupil so that it goes through the LENSLET (glass in front of camera)
- Click [LENSLET] on the PAVO gui to put the lenslet into position.
- Change filter to FB500-40 for the green laser (LP02-568 for red laser).
- Put small square target on the lenslet stand for reference
- Adjust pupil so that it's centered in the target box
- If the flux in the laser spot starts to fade, adjust the image plane
to bring flux back to a good quality.
- If it's a long way off, will have to alternate between moving pupil
and image planes
- e.g. if laser spot needs to move up, move pupil "up" and image "down"
- may have to occasionally adjust left/right to keep laser going through mask
- Type "zwrite" in pavo server to save current positions
- Leave PAVO shutters open when exiting server
- This will save M1 mirror positions in the up/open position
- Do a tip/tilt alignment after completing the PAVO alignment
- This will compensate for adjusting VisBeams after putting in the PAVO beam splitters
- Put the black box cover back in place over PAVO camera. Slide the front panel on the front of the PAVO table (this is the panel with the inset cut out for the beams). Leave off the side panel, as this will block the light from beam 1.
4.2 Checking PAVO alignment
- After beams are aligned through the mask and lenslet, go back to check
the alignment on the L1 and L2 lenses
- Slip target cover on L1 - check offset from center: +/- 8 mm tolerance
- Slip target cover on L2 - check offset from center: +/- 9 mm tolerance
- Also check for vignetting by the L2 lens/M2 mirror
4.3 Adjusting PAVO Coarse Alignment
4.3.1 Step 1 - Correct for offset on L1
- See where beam lands on L1
- Shift L1 (move on table) so that the error is halved
(move halfway between the original position and where the image is)
- Re-align pavo again through to the end
- The error on L1 should be just about cancelled out
4.3.2 Step 2 - Correct for offset on L2
- Move L2 (using pavo gui) so that the beam lands in the center of the lens
- Then move M2 (hand-adjustable knobs), so that the beam goes through the mask
- This should correct for the offset on L2
4.4 Changing focus
If the L1 or M2 mounts have been physically moved on the table during the
course alignment, the focus of the beam might need to be adjusted.
- Turn on PAVO camera [GO/Stop]
(make sure camera is cooled)
- Turn laser on ND 3.0
- Open PAVO shutters ([B1] [B2] [B3] buttons)
- Look at images of each beam, most of the light should fall within 1-2 pixels
- If the light is too spread out, need to adjust the position of L1
along the direction of the beam
- Easiest way to do this is to take out the mask [MASK OUT] and slowly move L1
while watching the image on the PAVO camera
- After this is done, will probably have to align beam through mask again
4.5 If you need to home only one of the PAVO motors
- "zhome N" where N is the motor number listed in the ENG gui
- Look at /ctrscrut/chara/etc/pavo_zaber_positions
- line 1 - motor 1
- line 2 - motor 2
- ....
- first number in each row is position 0 (i.e. the on-sky shutter open position)
- "zmovabs 5 12345" - moves motor 5 to position 12345
- Alternate: "zread" - will read last saved positions and set motors.
- This could be dangerous because it can overwrite any unsaved offsets
- Motors have allowed ranges out to 132000.
4.6 General Notes
- With the green laser, POP changes can be done directly on the PAVO beams using
the normal procedure.
- The PAVO alignment out to the telescopes (VIS/IR dichroics) must be re-done
if any changes are done in the lab (see Alignment Out to the Telescope).
- Make sure the PAVO Shutters are open before aligning. The laser position will
shift very far down on L1 if the PAVO shutters are closed.
- Leave the shutters open when shutting down the PAVO server.
5 Starting PAVO Software
This section describes the PAVO-specific servers and guis that need to be
open at the beginning of the night to observe with PAVO. This list does
not include the general CHARA software that needs to be opened.
- Open PAVO server from pull-down menu (Servers -> PAVO). If this doesn't work,
then try "ssh -X pavo" then typing “run_pavo”. Note that for now (Oct 2008)
it is OK to do this from ROCS/AROC, although there have been times where there
seems to be fewer communication issues if the server is run from the mountain.
- If the server is being run from the mountain during a remote observing run,
then the remote observer should open a terminal window and type “speak pavo”
to communicate directly with the PAVO server. All commands that are usually
entered into the PAVO server can be entered into the speak pavo session. Use
CTRL-C to exit out of the speak session, do NOT type exit (that will close
down the server).
- Open PAVO GUI from pull-down menu (GTK -> PAVO)
- For remote observers, clicking the [STATUS] button will bring up the status
window that is usually displayed on the top part of the PAVO server.
- On Grand Wazoo, select PAVO from the beam combiner menu and turn off
[BL SOL]. Click [REOPEN] on GrandWazoo to set baselines after starting
the PAVO server.
- Click [ENG] on PAVO gui to open engineering gui [ONLY NEEDED FOR WAVELENGTH ALIGN]
- If the server is re-started:
- Click [REOPEN] on PAVO GUI and PAVO Engineering
- Might need to close and then reopen displays
- Open LDC controllers from pull-down menu
- ESP Controllers -> LDC S1/S2/W1
- ESP Controllers -> LDC W2/E2/E1
- Wait to turn on LDC power (green) until after the first acquisition alignment
and reference cart has been set (this will avoid sending LDCs to a bad position).
- At start of run, you may want to home each of the LDC units
- Make sure vel and acc are nonzero (or else the units won't home)
- If you need to home LDC glass during the night, turn off ASTROMOD using the
ople gui, home the glass, then turn back on ASTROMOD.
- Sync the time for pavo after the operator completes the sync gps sequence
(do this in a pavo login window)
- Note: As of July 2009, the PAVO connection with OPLE stays open.
Check PAVO server to see if the connection with OPLE is open or closed.
If needed, typing “opleclose” into the PAVO server will close the connection.
This used to be necessary to prevent PAVO from changing goffsets automatically.
6 Alignment Out to the Telescope - Skip this Section (2010 April)
NOTE: Now that we have the green laser in the lab and can do M10 alignments
with the PAVO beam splitters in, there is no longer a need to do the
laser alignment out to the telescope. Also because we are no longer using the
red laser, we cannot do the laser wavelength alignment anymore.
In this step we align the VIS/IR dichroics out to the telescope
(flower laser align). This only has
to be done when the telescopes are changed to different beams or after
changing POPs. When this alignment is needed, it is done at the beginning
of the night. Check with the telescope operator to make sure it is safe
to open the M3 and M5 covers on the telescopes.
You don't need to do the alignment out to the telescope if you're using the
same telescope setup as last time and the dichroics haven't been adjusted
on the PAVO beams. It is always needed each time PAVO is used with a new
telescope or a different beam setup.
6.1 Align VIS/IR Dichroics
Align Vis/IR dichroic using telescope corner cubes:
- Telescopes must be stowed. Carts should be at the back switch (?).
- Make sure you set gain to 0 in the GUI (or "gain 0" in the server)
- Click on [DISPLAY] to turn on PAVO camera display
- Click [MASK OUT]
- Click [GO/STOP] to start camera
- Select laser ND 3.0 and turn the laser on
- Open M3 and M5 covers using Telesope Controls
- Open VisBeams 1 & 2 (or 3 & 4)
- Click [IMAGE]
- Set filter on PAVO gui to FB500-40 for green laser (LP02-568 for red laser)
- Check that the laser is visible in the PAVO display. Beam
1 should be at the top, beam 3 at the bottom and beam 2 about 2/3 of
the way up. If the flower shape is cut off by a sharp edge, use the (e.g.
for beam 1) "B1 Image" selection of the "Move" drop down box in the
PAVO gui to move the image so it isn't vignetted.
- Repeat for other beams.
- Close IRIS, using the IRIS gui (from the drop down menu), then [INC] x 2
- Select [PUPIL] in the PAVO gui.
- Set filter on PAVO gui to FB500-40 for green laser (LP02-568 for red laser)
- Select laser ND 1.0
- You should see a "flower-shaped" image, although it likely isn't evenly
illuminated. Tweak PICO1 -> S2DICH for S2 (other applicable dichroics for
other telescopes) so this image is evenly illuminated and bright. You
will have to move the dichroic by a hundred steps to see it move, so a
step size of 100-200 is good. PLEASE do not move dichroics too far from
nominal positions - the system alignment cannot tolerate too much adjustment.
- Tweak "B1 Pupil" in the "Move" drop down box (e.g. for beam 1) so that this
evenly-illuminated pupil image is in the center.
- Repeat for other beams.
- Reopen IRIS to BEAM.
- Close laser shutter.
6.2 Align Wavelength Scale
Align wavelength scale of spectrum. Experts only for now...
- ND 3.0
- Open the shutters (laser and B1-B3)
- Click [LENSLET]
- Select LP02-568
- Click TV [FINDER] then [OFF] to shut off dome lights
- Type "dispmode zoom" at pavo prompt
- Look at XY numbers on pavo server (two numbers; centroid of 4 pixels)
- Want the XY numbers to fluctuate around 0,0
- Adjust on engineering gui:
- 1 lenset [<] 401522(1) [>] [-] 2 [+]
- Click [-] [+] to adjust step size
- Click [<] [>] to adjust X,Y numbers
- Example: X was too big, so click left [<]
6.3 After Finishing the Alignment
- Type “zwrite” in PAVO server to save alignment
- Set gain back to the correct value
7 On-sky Alignment
On-sky alignment of the image and pupil planes. This should be done when
switching to a new target/calibrator set in a different part of the sky.
You need to re-align after a large telescope slew or more than about
1 hour on target.
7.1 Before Aligning
Before aligning (i.e. while slewing) do the follow setup steps:
- Set the reference cart and glass position by following the procedures in Observing with PAVO.
- If LDC glass is not in the right position, you will not see the laser in ACQ
(make sure ldc velocity is nonzero or glass will not home/set properly)
- May need to use laser on ND 2.0 to see it on Beam 2 in ACQ
- Set the gain using the PAVO gui
- gain 173 - for mag 5.5 and fainter.
- gain 150 - for 3rd magnitude or fainter.
- gain 140 - for brighter target (up to about 1.5).
- gain 100 - for very bright target.
- If changing the gain, you must take new backgrounds. Using a gain mode that is too high
makes the camera age (it costs $30,000) and the PAVO server will give a “flux dangerously high”
if it calculates that you're in that regime. One good reason to go to a higher gain mode is that
the Flux number is more reliable and doesn't go negative.
7.2 Take Backgrounds
- Make sure the [LENSLET] is in
- Click on [DISPLAY] to bring up the PAVO camera display
- Click [GO/STOP] on pavo gui to start camera and activate numbers on pavo server
- Click [ACQUIRE] on pavo gui
- Take backgrounds
- Close pavo shutters using [B1], [B2], [B3]
- Click [back] - turns on background
- Wait for flux on PAVO Server to go to zero (i.e. less than 10$^5$ and
preferably fluctuating above and below zero).
- Click [back] - turns off background
- S/N should be low
- Take foregrounds if you want V^2 to be correct on PAVO display
(do PAVO shutters have to be open during foregrounds???)
(this step is not typically done)
- Click on [VIDEO] to continue with the alignment
- Note: Only have to re-do backgrounds if changing the gain or re-starting the
PAVO server. Just pressing “Go/Stop” is fine and won't effect the background.
7.3 Align Image
- When doing the alignments, you may want to set display to average 10 frames
- In PAVO gui, set Average to 10
- Set this to 1 later if you want to see fringes in big window display
- Click [IMAGE] on PAVO gui
- Make sure [B1], [B2], [B3] shutters are open on pavo gui
- You should see star light from all three beams (B1=top, B2=middle, B3=bottom)
- With mask in, click on [TARGET] button
- Position color-coded boxes by using mouse to click on center of light
in B1, B2, B3 images. The title bar of the PAVO Image/Pupil Display will
change, telling yo uwhich beam you're up to.
- Take mask out [MASK OUT] - images of star will get bigger
- Align centroids of B1, B2, B3 images to lie within the color-coded boxes
- Do this by moving "B1-image", "B2-image", and "B3-image"
- Click [MASK OUT] again to put the mask back in and check that the boxes are
evenly illuminated
7.4 Align Pupil
- Click [LENSLET] on PAVO gui (mask should be in)
- Want to center the black hole in center of the pupil and get
beams to overlap in pupil plane. If you can't see well-defined hole,
just center the light in the display.
- Open each beam individually using shutter buttons [B1, B2, B3]
- Align B1, B2, B3 pupils by moving "B1-pupil", "B2-pupil", "B3-pupil"
- When finished, open all three PAVO shutters and make sure that the mask and
lenslet are both in to begin searching for fringes.
- If bandwidth issues are a problem, click [No Display] to turn off large display. However,
it is a kind-of-useful thing to have up if there are no problems.
7.5 Save Alignment Positions
- Type “zwrite” into the PAVO server when an alignment is complete and you're
happy with it. This records the positions of the PAVO optics. This is
especially important to do after aligning for the first time at the beginning
of a run.
8 Observing with PAVO
- Sync pavo, by typing "psync" into pavo login window
- Slew to target
- Check reference cart position on OPLE server
- On the OPLE CONTROL gui, make sure that MAN is depressed grey for the active
and non-fixed carts. PAVO does this for you in principle, but good to check.
(OL must also be clicked for all three active carts.)
- Check reference glass position on LDC. If you are *not* using AUTO for the reference
cart, i.e. you are tracking on it, then the glass should be set automatically (but good to check).
- Set the LDC position for the reference cart
- The LDC glass on all beams should be between -10 mm to 50 mm.
- Look out for "<-10mm" or ">40mm" on OPLE server and use the ESP LDC gui to
move the glass for the reference cart.
- Hint: this will roughly work if you put
the reference glass position in mm the same as the reference cart
position in meters, minus 5mm.
- Tell OPLE to send offsets to LDCs by typing the following commands into
the OPLE server:
- Align image/pupil planes on-sky if needed. See On-sky Alignment.
- Make sure mask and lenslet are in before searching for fringes.
- Click [SEARCH] button to start the search, in a zig-zag pattern. For the first star especially,
you'll next want to
- If the alignment doesn't need to be checked, PAVO will automatically start scanning for fringes after the star is acquired.
- PAVO will lock onto fringes when found
- Check pavo server to make sure the servo is on for beams that are in use
e.g. if beam 1 is the reference cart, turn on beams 2 & 3 (type into server):
- >
servo 2 aoff
- >
servo 3 aoff
- If a cart is not being used, set it to "OFF"
- Reference cart should be listed as "FIX" in pavo server
- AOFF - automatic search but not locked on fringes
- AON - locked on fringe
- AMOV - moving cart fringes
- FIX - cart active but fixed as reference (type FIXED when using servo command)
- OFF - cart off, not searching
- Set integration times for each baseline:
- The three numbers are itime in seconds for beams 1, 2, and 3
- In the above example, pavo will wait until it acquires 120 seconds of
locked fringes on all three baselines before automatically doing a
shutter sequence
- If you are not fringing on a particular baseline or if you don't
want to wait too long to acquire enough signal on a particular
baseline, set those baselines to 0.
e.g. "itime 0 120 0" will acquire 120 secs on baseline 1, regardless
of how long fringes were locked on baselines 1 and 3
- If you want to end an integration, click [SHUTTER] to go immediately
to shutter sequence or [GO/STOP] to end with no shutters
- Adjust S/N level that turns auto fringe on or off (AON/AOFF)
- Adjust these values by looking at the S/N values on the PAVO server. Look
at the range of S/N values for the background (no fringes).
- Set “snon” above the highest S/N value in the background. This determines when the fringe tracker turns on. Setting SNON too high won't stop the fringe search when fringes are found (the fringes will pass right by). Setting SNON too low will cause the search to stop frequently and track on noise.
- Set “snoff” roughly to the average S/N in the background. If SNOFF is set too low, then the fringe tracker won't turn off after fringes are lost. This will cause PAVO to track on noise. If this happens, the fringes will appear to be yanked away after they are found.
- Set autosearch parameters that define how to get fringes back after being lost
- Set autoseach: beam 2, move -5 um away, search at 5 um/s
- Autosearch will do a back-and-forth search, increasing the width of the
search on each pass
- These seem to be pretty good settings - set for both "moving" beams
- To restart autosearch, type “sas” into pavo server
- Set linear search parameters
- Click [LINEAR] button to enter cart offset and velocity for linear fringe search
- The units displayed in the LINEAR search box are um and um/s
- For a slow search, set velocity to 10 um/s
- For a fast search, set velocity to 20-30 um/s
- Manual commands that do the same thing (entered into PAVO server):
- Adjust search parameters: “fs 2 4000 30”
(fringe search: beam2 4000 um offset at 30 um/s)
- Send cart 2 to specified offset: “foffset 2 3900”
- Set cart 2 to specified velocity: “fvel 2 30”
- [PLOTS] button will bring up three fringe display plots (one for each baseline)
- They plot group delay along x-axis (+/- 19 um wide) and time along the y-axis
- Fringes show up as bright verticle "streaks" in fringe display plots
- Fringes will drift in/out from the sides as fringes are found/lost
- [DELAY] gives a plot of the delay velocities of each cart
- [SAVE] button can turn on or off file saving
- Turns on automatically when fringes are found continuously for 1 sec
- Turns off when slewing to new target
- [SHUTTER] button will begin shutter sequence (if want to end integration early)
9 Detailed Observing Notes
- Two-Telescope Mode
- For 2-telescope mode, using beams 2 and 3 (figure it out for yourself
if you're using other beams), type "servo 1 off" to turn off the servo for beam
1, and type "itime 0 120 0" to make sure PAVO is not waiting for a long
integration on baselines 1 and 3.
- Moving Cart Fringes
- If you get moving cart fringes frist, you'll see "AMOV" on the pavo server.
GW will print out the position. If you are doing a linear search then it
should continue while locking the two moving carts against each other, finding
the other fringes automaticallly. If it's in back-and-forth search
mode then AMOV
might get kind of screwy. If the AMOV fringes are lost, it's always a good
idea to check if the carts are still heading in the right direction.
- Changing the Integration Time - Faint Targets and Good Seeing
- On faint targets, S/N will improve by increasing tint to 20 msec (default is 9.7 msec).
- To change integration time, type “tint X” in PAVO server (where X is in msec)
- Re-do backgrounds after changing tint.
- It is ok to change to longer tint if the seeing is good and the Rat
column on the PAVO server is above 0.5 consistently on a star with
good fringes (e.g. a bright calibrator). It is better to leave at the default of 8 msec in bad seeing.
- Calibrator and science targets must be observed using the same tint.
- Aligning on Bright Targets
- For bright targets (V > 2 mag), align on a fainter calibrator or nearby star
so that you don't saturate the detector during alignment.
- High Dispersion
- If the dispersion of the fringes is high or if you can't find fringes,
you may want to adjust the glass offsets used by the ople server (this is an
offset sent to the “moving” LDC glass, that adjusts it from the nominal
position on the LDC display)
- High dispersion will produce smeared, asymmetric fringes
- When tracking on fringes, the DISP column on the PAVO server should be < 2.0.
- If the dispersion is high, type “tg” into the PAVO server while on a
fringe finder or calibrator to correct for high dispersion.
See Troubleshooting for more details on how to correct this.
- If the dispersion number is near zero (i.e. you've found fringes on a bright
star and done "tg"), you get lower noise by typing "dispersion" into the PAVO
server to zero the dispersion. Otherwise the software searches over 1
dimension of group delay and 1 dimension of dispersion to search for the
biggest peak. (Helps by about 0.3 mags only.)
- Tips to Help Track Fringes Better (adjust these parameters depending on data quality and seeing conditions). Note: to see default values before changing, type the command with no value into PAVO server.
- Change tint in pavo server (see above; default 9.7 msec)
- If you're having problems with R=6.6 and V^2C is more than 0.2, then you
should go to tint 20. Calibration will be a little worse but better chances
of trackable fringes on faint stars.
- Calibration sequence must be taken all with the same tint
- Re-do backgrounds after changing tint
- Adjust dtime. For good seeing (stable fringes) use dtime up to about
2. For bad seeing (fringes jumping left and right) decrease dtime to
about 0.4.
- Adjust snon and snoff
- Change autosearch parameters to smaller offset and velocity values:
- Adusting the Incoherent Integration Time
- Adjust “dtime”. This is shorthand for "gddtime" i.e. "Damping Time" and
combines both "gddtime" and "sgain". If you want to have a longer incoherent
(software) integration time, increase dtime. e.g. dtime 0.5 is appropriate
for fast seeing, and dtime 2 is appropriate for fainter targets in good
seeing. It doesn't change the recorded data, only the analyzed data used for
fringe tracking.
- Photon Counting Mode (pc)
- For photon counting mode to work, stars have to be at least 6th mag or
fainter, and it only makes a
noticeable difference at 7th mag. To use pc, first change the gain to 173
(e.g. the gain box on the GUI), then take a new background, then type "pc",
then take _another_ background, then you're good to go.
10 Description of PAVO Server/PAVO Status Window
This Section describes the parameters listed on the top of the PAVO server display and on the PAVO status window.
- Camera temperature
- Saving mode on/off
- Filenumber - current saved file number
- GD - group delay
- S/N
- Tint - total itime locked on fringes
- Flux
- Disp - glass dispersion
- CP - closure phase in five channels
- POS - gives cart offset for each beam (1 cart will be fixed)
- State - the beam will show up as OFF, AOFF, AON, AMOV, FIX
- OFF - cart off, not searching
- AOFF - automatic search but not locked on fringes
- AON - locked on fringe
- AMOV - moving cart fringes
- FIX - cart active but fixed as reference
- Vel - velocity of cart
- TL - time locked (for a given scan)
- S - indicates whether shutter is open/closed (X means closed)
Here is the order that some of these parameters are listed on the Pavo server:
Pos Vel TL S
V2 V2C V2S V2CS (different visibility estimates)
11 Shutting down PAVO
- Leave PAVO shutters open when exiting the PAVO gui and server. This will save
the M1 mirror positions in the up/open position
- If you will be using PAVO the next day, you can leave the camera cool
and just “exit” out of the server.
- If you are finished with the PAVO run, then warm up the camera before exiting
by typing the following command into the PAVO server:
- Wait for camera to warm up to 0 degrees and then exit out of server.
- Turn off power for LDC units.
Compressing the data at the end of the night
- ssh to pavo and “cd /raid/090430” or whatever your UT date is.
- Type “compress *fits”. Or if that fails (file list too long)...
“compress pavo0*fits &” then “compress pavo1*fits &” etc.
- Do a “df -h” and check there is at least 200GB left on /raid.
12 Troubleshooting
12.1 Can't Find Fringes
This list outlines the simple things that can prevent you from finding fringes:
- Check CHARA time with official US Time. Also look at “Ext-Sys” line on the GPS server - this gives a comparison with an external clock. Also check ople and pavo time.
- Check carts: within delay line range (0-44 meters) and small errors (< 1)? Do the carts need to be homed?
- Make sure [MAN] is pressed on OPLE Control for the moving carts (fixed cart will be green).
- Check glass position: within allowable range (-10 to 50 mm) on all beams?
- Check flux: is light getting through on all beams? Are PAVO shutters open? Check alignment?
- Check flux levels: Is there any light in the lab entering PAVO (this will show up as vertical stripes in the PAVO pupil display when taking backgrounds)? Are the acquisition mirrors out (PAVO pupil will appear cut-off if they are not)?
- Check “goffset” command on ople server - see Dispersion is High.
If OPLE is not tracking - go to OPLE Server:
- If mean errors are high (should be 0.00) -> anything bigger than 1 is high
- Click [OFF] - [BACK] - [OFF] - [TRACK]
or... [OFF] - [FRONT] - [OFF] - [TRACK]
- This often happens with W1 - keep you're eyes open in case it gets stuck.
Check LDC glass position:
- The LDC glass on all beams should be between -10 mm to 50 mm.
- Change the LDC position for the reference cart via LDC controls.
- Possible reason for not finding fringes
- Look out for "<-10mm" or ">40mm" on OPLE server and use the ESP LDC gui to
move the glass for the reference cart.
- IMPORTANT HINT: this will roughly work if you put
the reference glass position in mm the same as the reference cart
position in m.
12.2 Can't See Laser in Acquistion
Can't see laser in [ACQ] when aligning TV Tracking on telescope:
- Beam 2 will be too faint to see - change laser to ND 2.0.
(Set back to ND 3.0 when finished.)
- Check glass offsets - these must be within allowed ranges or you won't
be able to see the laser (change ref cart glass if necessary) See Can't Find Fringes.
- If you need to change glass position on individual beam, type in pavo server:
- Move glass by hand using LDC gui
12.3 Can't See Laser in PAVO Camera
You need to switch the filter to LP02-568 in order to see the laser in the
image plane or the lenslet plane.
12.4 Dispersion is High
If the dispersion is high or if you can't find fringes, you may want
to check the glass offsets being sent to the LDC controls by the ople server.
High dispersion will produce smeared, asymmetric fringes. This is mostly an issue in terms of tracking on the broadband fringes (post-processing is done using small sub-bandpasses). There are three ways to adjust the glass offsets:
- As of July 2009, you can type “tg” into the PAVO server on a fringe finder
or calibrator to correct for high dispersion.
- Opening the connection to ople to allow PAVO to communicate with ople and
adjust the glass offsets automatically
- PAVO server says whether the connection to ople is open or closed (default is closed)
- If the dispersion is high (DISP column on PAVO server > 2), type “opleopen” into pavo server while tracking on fringes (AON)
- This will allow PAVO to send commands to ople to adjust the goffsets
- Wait until the dispersion gets low while tracking on fringes
- Type “opleclose” in PAVO server to close connection and continue taking data
- DISP < 2 is good
- Change goffsets directly using the ople server
- Type “goffset” in ople server
- Select the telescope for which you want to adjust the glass offset
- Enter the goffset value you want to set
- Ask Mike for recommendations on setting the goffset values
(e.g. during Ellyn's run a goffset of 1.8 on W1W2 worked well)
Mike - since the glass offsets still occasionally pop up as a problem, could you provide some insight on how to determine what the goffset values should be set to?
12.5 Large Jump in Image Position
If there is suddenly a large jump in the position of the image, it might be
caused by L2 being at the end of its range. To check:
- Open PAVO [ENG] gui
- Look at numbers for (e.g.) B1 L2 X
- That number can't go above 132000 when aligning the image and pupil.
12.6 LDCs Won't Home
- Make sure VEL and ACC are nonzero (or else the units won't home)
- If you need to home LDC glass during the night, turn off ASTROMOD using the
ople gui, home the glass (wait for position to go to zero), then turn back on
ASTROMOD. If you don't turn off astromod first, OPLE may send a new offset
to the LDC glass before it has a chance to get to the home position.
13 Recommendations for Setting Beam Order
These were Mike's recommendations for setting the W2-S2-W1 beam order (Nov 2008):
- The best baselines are between beams 2 and 3, then between 1 and 2, then
between 1 and 3. So - for the best possible fringe tracking for faint,
resolved stars, try W1 B1, W2 B2, S2 B3.
- HOWEVER, the very shortest wavelength channels of PAVO don't work for the
baseline between beams 1 and 3 (not quite Nyquist sampled), and have poorer
signal-to-noise.
- So if you really want maximum resolution, and maximum signal-to-noise at
high resolution, then you want W1S2 on e.g. beams 2 and 3. In that case I'd
recommend W2 beam 1, S2 beam 2, W1 beam 3. Then you have best
signal-to-noise on your longest two baselines.
14 Fringe Finding Sources
- It is often helpful to use a bright, unresolved star to find fringes for the
first time. If your targets and calibrators are on the fainter side, it's
a good idea to look for additional fringe finding targets.
- There is an idl routine available to aid in selecting bright fringe finding stars
- Log on to pavo using “ssh -X pavo”
- Type “idl” at the prompt
- Run the star finding routine
- >
pavo_ptsrc, [10,30,00,+45,00,00], 10
- Where the array in the call is the RA and DEC in [hh,mm,ss,dd,mm,ss] of the
location where
you want to find a bright nearby star and the last number is
the search radius in degrees.
15 Random and Important Notes
- There is currently a bug that if the PAVO gui is reopened, the gain does
not update, so it will go to 0 and still read a values such as 140 or 150.
- Clicking [image] will automatically change filter selection to narrow-band
filter. If prism is left in, we would see long spectrum rather than star
image.
- You need to switch filter to see the laser in the image plane or the lenslet plane.
- Restarting pavo:
- Click [REOPEN] on GW.
- Click [object] on GW to re-acquire target in order to set target name
(otherwise type "HD 12345" on pavo; type "star" to check)
- If you need to reset the glass in primary ople server, e.g. if you've changed baselines and the glass for an old telescope is at -20mm, type into the ople
server or speak session:
- >
autoldc off ; tells ople to stop moving the LDC's
- Using the LDC Gui, set the velocity of the glass in question to 50,
home the glass, then turn it off. Back in the ople server:
- If the display freezes: click [NO DISPLAY], [REOPEN] and [DISPLAY].
If that still doesn't work, close PAVO GUI and reopen it.
- If you totally screw up the alignment... you should do a "zzero"
- If you totally screw up the zaber save file... look for old files
with dates as part of their name in /ctrscrut/chara/etc/zaber. These files
are automatically created whenever the PAVO server is shut down. Copy one of
these files to /ctrscrut/chara/etc/zaber_default_positions, do a "zread" then
a "zzero".
- What happens if alignment was good, PAVO crashes, then alignment screws up?
This generally only happens if you adjusted the PAVO alignment considerably,
didn't do a "zwrite" and left the "shutters" closed. Then the M1 mirrors will
be left in their "shutter closed" postition, and the software will not know
where you adjusted it last time. However, it generally isn't too hard to
find the beam again because only M1 vertical (y) will be out.
- How does PAVO remember the saved positions (e.g. in the PAVO engineering GUI)?
Whenever you go to another saved position (e.g. using the engineering GUI, or
clicking [MASK OUT] or one of the shutters [B1] [B2] [B3] or [IMAGE] or
[PUPIL], if you are selecting a position you are already at, you go to the
last saved raw position for this setting. If you select a different setting
(e.g. by clicking [B1] after adjusting it) the previous raw position is saved
in memory.
- Tricky power-user shortcut: to get ctrscrut window just type "ctrscrut" on rocs
- How can I see the Wazoo log from ROCS? From a ctrscrut terminal "cd /ctrscrut/chara/data"
then look at the appropriate status_xxx.txt file. A new one is written with each reopen.
16 PAVO Engineering GUI
The PAVO Engineering gui is no longer needed as much as it used to be.
You may need it to set the mask to the right position at the start of the lab
alignment (Basic Alignment of PAVO in the Lab), to adjust the wavelength
scale (Align Wavelength Scale), or to correct for one of
the lenses reaching the end of its limit (Large Jump in Image Position).
A largely incomplete listing of the PAVO Engineering lines:
3 Mask [2] [3]; 2 - mask in, 3 - mask out
9 B1 [0] [1];
11 B2 [0] [1]; shutters: 0 - open, 1 closed
17 Index