ESO Faint Object Spectrograph and Camera
EFOSC2 Support Astronomer Notes

Here you may find some information about EFOSC2 which cannot be found in the manual and the normal web page.

P2PP template files (this info may be outdated now that P2PP doesn't just run  on kila):

-- See later notes on P2PP below.

Log in as instmgr@killa. The files used by P2PP are in directories:

~/instruments/EFOSC (used by engineering/testing)
             /EFOSC2 (used by P2PP)

*.tsf are the template signature files. When making changes to the.tsf files,

1) delete index.EFOSC2
2) Run p2pp from the instmgr account -- this will generate a new index
file. (See README.install.new-IP written by Oli for instructions to
archive to Garching the new.tsf files).

Making OBs with P2PP -- some notes

Select period.
Be sure to hit return or tab after making any changes so that the value is registered.

1) Spectrum

normal readout -- 1 amplifier
fast readout  -- 2 amplifiers

There is _no_gain_ from windowing the CCD readout -- don't advice it.
The binning should be 2x2 unless seeing < 0.6"

2) Image

"Return to origin" = return after the entire sequence.
offsets: each is relative to the one before. These are _telescope_offsets_ (+1" in RA goes 1" E, +1" in DEC goes 1" N.).

"Combined Offset" = move telescope and guide probe simultaneously, so that guiding is not lost, and the offset is accurate.

The observing sequence is always Point, Offset, Observe.

"Exposures" = number of offsets, cycled through the list of offsets.

"List of exposure time" = exposure time of each offset (cycled through list of offsets). For example, if number of exposures = number of offsets = 5, and list of exposure times = t1, t2 and t3, then the sequence becomes:

Exposure  1  2  3  4  5
Offset    1  2  3  4  5
Exp time  t1 t2 t3 t4 t5

3) Acquistion

For coronograph, in the afternoon, should have taken an image with the
mask in. From this image, measure the (x,y) of the mask centre. This
goes to the "move tp pixel" field in the acquistion template.

4) Polarimetry (imaging)

If "continuous rotation" = F  -- use list of fixed angles
                         = T  -- continuous rotation

"Move half wave plate in" = T

Exposure time -- it is for each angle.
              = N*T where N os the number of postions.


In the afternoon, take image of the Wallaston mask with and without
prism to find out the centre of the gaps. Measure the y positions of
the cetres of the strips where the light is coming through when the
prism is inserted. When using "Move to Pixel" the object must be in
one of these (y) positions.

Let observer decide where to put object by bringing them to the
telescope in the afternoon.

Putting Wallaston Prism into the light path shifts the image location
by about 1 pixel -- beware! (and warn observer).

Wallaston mask --> Starplate wheel
Wallaston prism --> Grism wheel

5) Polarimetry (spectroscopy)

MUST put Wallaston prism in the FILTER WHEEL -- grism has to go to
grism wheel. Observer must give enough warning for the prism to be

polarimetry-spec figure

Because the masks mask objects in one direction, can only use one of
the parallel or perpendicular prisms (I can't rememberthe definition
of parallel or perpendicular -- probably the perpendicular one) MUST

6) Standard stars OBs

OB for both imaging and long-slit spectroscopy are in the impex-stdcal directory. It can only be modified by e3p6ops, and it is under configuration control

Remedy (outdated)

login to 360cat@aylla

> setenv DISPLAY etc
> aruser
Macro - Run Macro

NRlists -- all night reports
All_problems -- every problem solved or not
Pending_problems -- problems not closed or solved.


Sciop daily meeting 1.30pm at Library
Tuesday weekly meeting at 5pm at Ritz
Thursday La Silla Action Point meeting, 2pm

Typical daily schedule:
1:30pm Sciop meeting
2:00pm La Silla Action Point Meeting (Thursdays only)
2:30pm Meet visiting astronomer. Assign p2pp number, userame.
4:00pm Bring astronomer of night to telescope for calibrations.
5:00pm Weekly team meeting (Tues only)
Sunset: twilight calibrations, start of observations
10:00pm Meet astronomer of the following night for introduction/strategy/tutorial.

3.6m email account (outdated?)

> setenv DISPLAY etc
> exmh   -- checks emails sent to 360cat account. Must forward emails to relavant people.

Other accounts
360v1,360v2,360v3 on kila
lasilla@kila   -- La Silla Account
efomgr@w3p6ins -- EFOSC magangment. EFOSC2 instrument scientist in charge of this account.
efosc@w3p6ins -- Instrument account.
guest@wp36drl  -- guest data reductions/cdburning

Telescope shutdown

1) Close mirror covers,fans.
2) Go upstairs to close hatch
3) Park telescope: in TCS press Engineering_Presets -- Zenith
4) Go upstairs, goto the TCS rack.
 - Turn telescope offline
 - Turn off the alpha, delta drive (put in the "pins"). Keep the button pressed until red
 - Turn off oil.
5) Close wind shield, close doors.

To turn on telescope

Reverse the above procedures

1) Turn on oil on TCS rack
2) Start alpha and delta drives (press button until green)
3) Press the reset button on bottom left rack
4) Online
5) Press reset again
6) Go back to control roo, to open mirror cover etc.

Image quality

NE is always bad
Near zenith is usually good.

Saoimage printer definition


> export R_DISPOSE = "lp -d lpt3p6 %s"     to print directly to printer
> export R_DISPOSE = "mv %s /tmp/saoimage.ps"   to dump to a file
Data Flow

astro@w3p6off  -- This should be where visitors reduce data.
guest@w3p6drl  -- Old linux data reduction machine. Not supported. Observer can reduce data here if he/she likes. Also for CD backups.
efosc@w3p6ins  -- Instrument machine where the logfile of the night and the raw data can be found. Do not enter unless necessary

Data flow chart

data flow chart

In order that the data is transferred from the dhs machine to the off
machine, the data subscriber needs to be running.

To run data subscriber on the off machine:


> datasubscriber
- enter ID name
- to set up preference choose "Stop", "Config"

Set up preference:
"From date" = yyyy-mm-dd 
choose "Rename to keyword 'Name on INS WS'"
No header translation.
No post command.
Press "Start Raw".

If data subscriber does not work, try to do the following:

Open xterm. Log in as efosc on ins machine.
> efoops
click "dhs + obs logs" (kill everythin on off machine beforehand")

Choose "DFSapps", "Start" (or stop), "vcsolac"

Don't do this unless really necessary.

To see if volac is running:


Data flow-- user backups

Personal backup -- recommended by ESO. User should backup data in the morning. Data are on w3p6off (astro account) in directory /data/raw/yyyy...

- Note that in the morning, the calibration files after 12:00 UT go to the directory of the following night-- remember to move them to a new directory before night begins, or files will be overwritten.

- Remember to reset the data subscriber (proposal ID, observer name) before each run.

- Remember to move files on the drl pc to new directory before each night begins, or at the end of each night, or they will be overwritten.

Observation log files:

These are kept at efosc@w3p6ins:/home/efosc/EFOSC_YYYY-MM-DD.log, and
at guest@w3p6drl:/scratch/guest/

To display log files in real time:

1) On w3p6ins

> efoops
select "dhs & obs logs" (beware of opening windows twice -- check if it is already running)

2) Alternatively, login to efosc@w3p6ins

> cd ~/logfiles
> sentenv DISPLAY w3p6off:0.0   (xhost + etc on w3p6off)
> xterm &
> tail -f -n 100 EFOSC_YYYY-MM-DD.log

Normally, 1) above should be started up during startup in the day. If
not, do 1) or 2) manually.

Rule of thumb exposure times for spectrophoto standards:

For EG274 (V=11.03, ra = 16 23 33.75 dec= -39 13 47.5, J2000),
exposure times for 5" slit and no filter are:

Grism  Exp Peak ADU
6    30    20k
3    30    20k
4    30    20k
1    5    15k
10    60    8k
2    30    45k
9    60    20k

Introduction check list

- meet astronoer 3:30
- explain car (137 -- astronomer)
- go up telescope
- explain how to enter dome
- give user p2pp password
- explain machines

To import OBs from linux machines

start p2pp on linux machines
select OBs with mouse
make dir for the exports
export the OBs
ftp OBs to dhs machine.

On tnt terminal, start an xterm from the dhs machine, start p2pp, and
import the OBs.

Telescope control panels (Adaptor F8)

To close mirror cover: "Mirror close" "Mirror"

"Set Centre field" is for another unused camera, so should be "OUT"
most of the time.

Slitviewer should be "OUT"

Calibration screen is automatically set by EFOSC to "IN" or "OUT".
When looking at sky, should be "OUT".

Rotator 270 degrees is default for EFOSC2, corresponding to 0 degrees
in BOB/P2PP:

P.A. (BOB) + 270 = Rotator Angle.

M2 focus: enter value (absolute), "Move"  (it is about 1922).

Guide Probe

GP filter is usually free.

Offsets for moving guide probe manually *Always Relative*

"Init" = initialise

"GP Centre position" to put GP in optical path.
"Guide probe focus" absolute ~ 11,100

Check that everything indicates "Online" (Except mirrir vent says "not
init" but it always is.You can hear the noise of the mirror
ventilation, and also the comments at the bottom of the display
indicates its status).

If not online, "Substate" - "Tracking" will say errors.

If telescope is not trackking, go to "TCS status" window, to look at
"Telescope State" that the "RA axis" and "DEC axis" are "ONLINE".  If
not "ONLINE", do "OFF" "INIT" "STANDBY" "ONLINE" one after the other.

One can see problems in "Log Mon + PrPrep" message window.

Telescope Interlock Panel - everything must say ok (green).

There will be messages in the l3p6ra-, l3p6dec-, l3p6ada- machine
windows.  These messages show what's happening with the LCU's "Local
Control Units". These are interfaces between software and hardware.

Under telescope control system "Dome mode" =auto or manual.

When taking observations make sure that "dome mode" is set to "auto".

Note that under "Engineering presets" option at the top of the window,
one can choose between "zenith", "flatfield" or "pole". The dome will
move to the optimum position automatically, no matter whether the
DomeMode is auto or manual.

Notes on grism alignment:

Object inclined, skylines horizontal -- grism is inclined
Object vertical, skylines inclined -- slit is inclined.

Martin Kurster's Midas Script for quickly extracting spectra:

source /scisoft/bin/setup.csh
cd <directory>

@@ efosc_extract

change cuts (remember to separate numbers by ",")
change scale

Mark 6 points for sky, object, sky.

Wavelength calibration (0, 1 or 2)?

1 -- Give the HeAr frame (eg Arc.fits)
  Order of polynomial (5 ok)
  Mark lines (left mouse button to mark, other to exit)
 need 6 lines for 5th order polynomial
 check if the fit looks ok
 save wavelength solution table (eg Gr1)

2 -- give the name of previously obtained solution table (eg Gr1)

Give root filename of resulting.fits and.ps files.

Some useful Mida commands:

> crea/graph
> set/graph pmode=1
> plot filename   (this is for a.bdf file)
> set/graph yaxis=-200,2000  (change limits)
> set/graph colo=2  (change colour)
> set/graph pmode=0 font=1 yaxis=-200,2000 (another example)
> over filename1 (overplots with file filename1)

To check the wavelength definition/pixel scale of the wavelength-calibrated file:

> read/desc file.fits

To execute unix command, preceed by a "$".

Updating CCD characteristics on web:

When CCD tests are done, an email will be sent to 360cat, and a file will be dumped in meli (now epu) under t3p6web. So update must be run to update the web.

telnet meli (now epu) as t3p6web


/home/t3p6web/bin/update is a script to update the CCD characteristic plots in the web pages. This should be done every week.

> update <no options>  -- prints out the options
> update CCD_number ID_mode

CCD_number -- 40 is the usual CCD used
ID_mode -- 42 is the most used mode (2x2 binning normal readout)

/home/t3p6web/ -- top directory of web page  /lasilla/telescopes/

/home/t3p6web/360cat -- 360cat page
/home/t3p6web/360cat/efosc -- efosc web  page

MOS Mask design and punching

Create dir eg onw3p6off eg /data/reduced/username

Copy image.fits (eg EFOSC_Image.1.fits) to dir.

> xm

Always do the following first:

"Options" -- "Punch Tool" -- "Slit"

Sizes are:
0.150 mm = 1.10" width, length 1000 microns 7.5" 48 pix. Broken.
0.249 mm = 1.78" width, length 1600 microns
0.25mm = 1.75" width
0.190mm = 1.35" width. length 1600 microns 11.5" 72 pix.
0.350mm = 2.5" width, length 1590 microns.

"EFOSC" - "LOAD and R90" load EFOSC-*.fits image.

(If doing it the second time, can do "File", "Load" *.bdf. This will load the bdf file which is the original file rotated 90 degrees.)

"Tools" -- "Cut" changes the cuts

Define slit "Options" - "Slitlength" -- "Minimum" or "Variable"

Define on screen: Interslit gap = 1 pixel. Object distance from slit edge = 1 pixel.

Slits "On cursor" or "On object"

Left mouse button to select. Twice middle button to exit.

To delete "Slits" - "Delete" First mouse botton-- delete. Twice middle button-- exit.

"Color" - "Clear" to clear screen after changes.

"Masks" - "Show all" to display slits.

To select variable length slits:

"Options" - "Varaible length"
"Slits" - "on object"

If variable length slit and on cursor, then requires 2 clicks. Bottom click defines x, ybottom of slit, top click defines ytop.

If masks overlap, the old one is automatically erased!

"Tools" - "Zoom" Zooms to a rectangular region

"Masks" - "Print maskves" Makes a PS file (note you must rename the file right away).

To save mask:

"Masks" - "Save masks"

This will create Image1_R00.msk and EFOSC-MASKS/Image1_R00.msk (for example)

To reloat the mask "Masks" - "Load mask" load the.msk file.

"EFOSC" - "Send mask files to" - "MOS@w3p6ins"

Telnet mos@w3p6ins

cd /vltdata/tmp/EFOSC-MASKS/

The file is Image1_R00.mask (for example)

Go to punch machine terminal.

Mos control panel should be working. If not, look at Emilio's web page.
Commands to try to reset the punch machine are: efoopsStartUp or efoopsMOSRun. The punching program is mosuifMOS.

Make Image

Slot the mask into position, tighten the screws witout flexing the mask.

On screen/terminal:

Move STAP to LOAD position

"Position N" - "Move Stap" to rotate the wheel.

To Punch:

Inputs: Position "Position N"
Overpunch "Overpunch"

"Load" -- path = /vltdata/tmp/EFOSC-MASKS/data/
"Start" -- check "step and "total"

If doesn't punch through:

"Start" again afterwards.

To get mask back "Position 5", "Move STAP"

** Important-- mark which side is up or down and mark the field name on the mask.

Remove tapes, keep mask flat on table.
Insert into the press, and cut plate.

Put mask into telescope. Select grism and mos, no filter, take internal exposure with quartz lamp to get the slit positions.

How to reset the punching machine

Turn off CAMARC (bottom box) using orange button at bottom.
Turn on CAMARC.
Use remote control:
Turn to position 2. Use "joystick" to illuminate "Unlock" and "LLIM" (lower limit) and "Park" (i.e. all red LEDs of these positions are on).
Reset LCU (white button on the middle rack).


Here is the status of the punch heads and tools. There has been a lot
of confusions. Do not confuse the punch tool (which is the machanism
which houses the punch head), and the punch head itself. 

(old) Punch 2: Ernesto has put a punch head of 1.75"x11" to the old
punch tool #2. So the original EMMI label on it is wrong. Now we have
renamed this punch tool + head combination as "punch 6", and removed
the old label. The EMMI label is now covered with black tape to avoid

Punch 5: This corresponds to 1.35"x11" on EFOSC. This is labelled.

(old) punch 6: both the punch head and the punch tool are bad. The
punch tool has a broken mechanism. Ernesto is going to contact
Garching about a replacement or instructions to repare it (Nov 2001).

The punches now correspond with the specifications in the old MOS manual and with those in the program xm.

New punch 6 = 1.75"x11" = punch #6 or #7 in xm    = 0.249x1.6mm in mos manual
Punch 5     = 1.35"x11" = punch #4, 5, or 8 in xm = 0.19x1.6mm in mos manual

Please note table 1.2 in the old EFOSC manual is wrong and this will be updated in the 2003 version.


EMMI 1.02" == EFOSC 1.35"
EMMI punch 1 (0.15mm x 1.0 mm) = EFOSC 1.06" x 7.1".

Instrument Templates


/vlt/insroot/SYSTEM/COMMON/SEQUENCES/*.seq  -- sequence/scripts
/vlt/insroot/SYSTEM/COMMON/SEQUENCES/*.tsf  -- definition of parameters

/vlt/introot/lib/libEFOSC.tcl/   -- directory of library files of EFOSC
                                                -- script library of subroutines
EFOSClib.tcl  -- file with all the subroutines

/vlt/insroot/SYSTEM/COMMON/MIDAS/PROCS/ -- comtains MIDAS routines (names are EXT*.prg)

/vlt/insroot/SYSTEM/COMMON/OBD/efosc/  -- these are files that can be sent to BOB if P2PP fails.

instmgr@w3p6dhs (this is account to help organise P2PP):

~/instruments/EFOSC2/*.tsf -- template signature files 
-- these are the files that P2PP on the dhs machine sees.
-- when updating.tsf files, must put copies in kila, dhs as instmgr, ins as efomgr, and also to the web link.

Under ~/instruments/ directory, keep EFOSC2.isf which contains the instrument definitions.

To update/make changes to.tsf files (note this section is old-- new procedures available via CCS-- see later).

1) Delete summary.idx (important)
2) Make the changes to templates.tsf
3) Restart p2pp on dhs as instrument manager (instmgr)

Keep old copies in ~/instruments/
eg tar cvf efosc2.tar EFOSC2* to make backup of the directory before making any changes.

Afterwards update the copies on kila and ins as well. Also must update the ESO archive.

Login to instmgr@kila

cd ~/InstrumentPackageArchive/ -- where the old templates (tsf files) are kept (eg EFOSC2_IP_20001005.tar
See README.install_new_ip under ~/instruments, can find EFOSC2/EFOSC2.isf.

How to modify templates/scripts using CMM

This is the proper way, as taught by Jorge Ibsen.

CMM Flowchart

e3p6osf -- operational stuff
e3p6msf -- maintenance
efiotm -- operational (e.g. libEFOSC.tcl).

Person in charge: jguzman.

Developement machine = efomgr@wlsops2

For developement:
Once I have been assigned keyword, add myself + password at the end of the.cmmrc file as efomgr@wlsops2.
(put myself also in.cmmrc file in w3p6ins (efomgr) as well).

When making changes, work in CURRENTLY_MOD directory.

In wlsops2

> cmmModify -l username modulename (e.g. efiotm)   This downloads from Garching and put all files with a write flag.

This will download the modules into the current directory, and locks the module.

Ask J. Guzman about simulation mode).

There will be directories called efiotm.modify/ and efiotm.

> cmmHistory efiotm  -- This shows history of module).

> cd ~/CURRENTLY_MOD/efiotm/SRC/


select file, make changes.

After modification, update the change log (alt_insert in emacs). If the header is standard, ctrl-insert will insert the log into the header directly.

> make clean (in scr directory)
> cmmCheckForArchive -l ghau efiotm  (in CURRENTLY_MOD/ directory).

Read the message to see if it is ok for archiving, if so:

> cmmArchive -l ghau efiotm "Short description of what I did"

eg "EFOSClib.tcl minor modification".

> cmmCompare efiotm 1.38 1.39    (1.38= old 1.39 = new).

> cmmLast efiotm (shows last modification to efiotm)
> cmmLast (list of last modifications of all modules -- very slow!)

To install  -- see later notes (more detailed and more correct).

login as efomgr@w3p6ins (don't do as efops -- Emilio is in charge. don't mix installations)

> cmmCopy efiotm    -- This downloads files without write flag for the source files.

> cd ~/CURRENTLY_MOD/efiotm/src

> make clean all man install copy
-- This compiles and install the module.

> cd ~/CURRENTLY_MOD/efiotm/bin efioInstall -- install files to $INSROOT.

Manuals to read about CMM


read: VLT Programming Standard, Configuratio Management Module user Manual.

Installation of new templates/scripts (CMM)

-- This is the correct method.

After a module is updated and archived, it is ready for installation:

Login is efomgr@w3p6ins

> cmmCopy efiotm        -- this downloads the module
> cd ~/CURENTLY_MOD/efiotm/src
> make clean all man install copy       -- This compiles and install the modules on the INS machine.

On the ins machine
> ~/CURRENTLY_MOD/efiotm/bin/efiotmInstall   -- to install obd.seq and tsf files to $INS_ROOT.
> rm /vlt/insroot/SYSTEM/COMMON/OBD/efosc/oldstuff.obd    -- this removes redundant obds.

Check also files in /vlt/instoot/SYSTEM/COMMON/SEQUENCES/ to see if they are up to date or need to be deleted.

Updating.tsf and.isf files on w3p6dhs

On w3p6ins as efomgr, run efiosfReplication under ~/CURRENTLY_MOD/efiotm/bin/.

efiosfReplication makes copy of.tsf &.isf files to the dhs machine. Make sure that p2pp is not running on the dhs machine when doing this.

Before running this, go to dhs machine as instmgr (can enter directly using rlogin from efomgr@w3p6ins). Remove old.tsf files under ~/instruments/EFOSC2/*.tsf, or move them somewhere else. The new.tsf files will be put under the EFOSC2 directory when the replication script is run.

Updating P2PP on w3p6dhs

on the dhs machine:

as instmgr, delete ~/instruments/EFOSC2/summary.idx

run p2pp as instmgr to create new summary file.

Making the new instrument package available

on dhs machine,

> cd ~/instruments/.
> tar -cvf EFOSC2_IP_YYYYMMDD.tar  EFOSC*  (this takes the EFOSC2.isf file as well as the EFOSC2 dir).


1) EFOSC web page. Login to t3p6web@meli and put it under efosc/docs.
2) On your computer at the same location.
3) On both your computer and meli, update all references to the IP in the html files.
4) Email amchavan, mperon, usg, dsilva@eso.org.

Bias Strip

Bias Strip diagram

New focus pyramid

Focus Pyramid

To test some of the Fiera/TCL commands

On the ins machine:

ccs command panel (ccsei Msg)

To test script
> seqWish &
source file
seq_ccsInit  (do this first to use message system of VLT)
msgSend  -- sends a command.
eg msgSend -v wefosc fcdconCI_efosc STATE " "
seqWish > source /vlt/introot/lib/libEFOSC.tcl/EFOSClib.tcl

Send comments to : ls-spectro

Last modified: Mon Sep  8 16:10:24 CEST 2003

[3.60 & CAT Homepage][La Silla][ESO][Observing][Search][Help][News]