La Silla - Science Operation Department

Title of the page 

Version: 1.0 2004-11-05 Paul Le Saux, Gaspare Lo Curto

TIMMI2 polarimeter study

This study reports on tests performed with the TIMMI2 template TIMMI2_pol_tec_minAngle which allows, by using the polarized filter N-filter+pol and the rotating polarimeter INS.ROT1.POS , under the light of the SIMULATED_SKY of INS.OPTI1 function to determine a "zero" angle of the polarimeter,
i.e. when it is orthogonal to polarized filter.

Test procedure

For this test an instrument mode had to be modified. It is defined "SIMULATED_SKY" and apart from inserting the "simulated sky" plate at the calibration wheel leaves the polarizer available for use.
In parallel a new instrument mode was created, defined "TEST", which leaves total freedom to the user about the setup.

Moreover the MIDAS-PCO interface had to be generated in the TIMMI2 instrument machine (wtimmi) in order to allow analysis of the data directly by the template.

A new entry had to be put in the modelist table to define the working paramters for the "N+pol" unit. After some tests the paramters were good enough to keep maxima and minima well inside  <sensitivity - saturation> range .

Instrument setup:

  • Filter wheel: N + pol
  • Calibration wheel: Simulated sky
  • Aperture wheel: 0.2" mask
  • Lens wheel: 0.2" N band imaging
  • Polarimeter: IN
  • Polarizer angle: varying
  • DIT: 75ms
  • NCYCLE: 1
  • NREAD: 66
  • NCDUMMY: 0
  • Read out mode: Single-IWR-Chopped
  • Data analyzed: HCYCLE1


     Some of the first results presented below refer to the maximum of the polarization curve, all the other ones to the minimum. This is because the different tests were performed in different phases of the optimization of the paramters (DIT). This has however no influence on the conclusions reported.

    The template fields are:  

    INS.POL.LOWANGLE.LABEL          "Polarimeter test lower angle"; # Label used in P2PP
    INS.POL.UPPANGLE.LABEL          "Polarimeter test upper angle"; # Label used in P2PP
    INS.POL.STEPNUMBER.LABEL        "Polarimeter test step number"; # Label used in P2PP
    INS.POL.BACKLASH.LABEL          "Polarimeter angle hysteresis correction"; # Label used in P2PP
    INS.POL.START.LABEL             "Polarimeter start angle between to tests"; # Label used in P2PP
    SEQ.PASS.NUMBER.LABEL           "number of pass along the lowerangle upperangle interval"; # Label used in P2PP
    SEQ.LOOP.NUMBER.LABEL           "number of times the test is executed"; # Label used in P2PP
    SEQ.INIT.FLAG.LABEL             "reinitialize ROT1 before next iteration"; # Label used in P2PP
    SEQ.INIT.PERIOD.LABEL           "number of iterations bewteen two reinitializations"; # Label used in P2PP
    SEQ.INTERLOOP.DELAY.LABEL       "number of seconds bewteen two test iterations"; # Label used in P2PP

    Therefore the template allows to measure the behaviour of ROT1 rotation or initialization according to the value FALSE or TRUE of

    SEQ.INIT.FLAG respectively.

    Test results

    The following "features" were noticed:
  • There seems to be some "hysteresys" on the positioning of the polarizer wheel; the SEQ.PASS.NUMBER parameter was created to allow testing it. When scanning the test interval backwards, it moves the polarizer rotator POL.BACKLASH degrees away from  the target angle before presetting to it. This  corrects only partially the problem :

  • -rw-r-----   1 plesaux    users        52505 Nov  7 09:56 not corrected
    -rw-r-----   1 plesaux    users        52698 Nov  7 09:56 corrected by -10 degrees

     Two things can be noticed in the two plots above:
    1) the curves are different in the two cases;
    2) the drift of the position of the minimum is minor when the backlash is applied.
        It is not clear however whether the difference between the two cases (with and without backlash) is significative.

  • INIT wheel : after reinitializing either all TIMMI2 ICS or the polarizer wheel ( angle control )

  • we found several times that maxima in a quite different position : jumps up to 60 degrees.
    -rw-r--r--   1 plesaux    users        24901 Nov  3 18:49    120 deg.
    -rw-r--r--   1 plesaux    users        22470 Nov  4 12:06        60 deg.
    -rw-r--r--   1 plesaux    users        52746 Nov  4 17:08        120 deg.

    The three previous examples plus another one not registered may lead to think that maxima can fall randomly
    in two areas : around 60 or around 120 degrees.

    Later on it was not possible to find again the 60 deg. maximum thru 6 consecutive initializations of the polarimeter
    rotation wheel or carriage, but it is clear that between two consecutive wheel restarts
    the minimun position moved up to more than 10 degrees.

  • The carriage initialization did not modify the previous minimum position :

  • -rw-r--r--   1 plesaux    users        34984 Nov  5 17:41   34.20 deg.
    -rw-r--r--   1 plesaux    users        33367 Nov  5 17:48  23.70 deg.
    -rw-r--r--   1 plesaux    users        35140 Nov  5 17:55  29.40 deg.
    -rw-r--r--   1 plesaux    users        35569 Nov  5 18:06  29.20 deg.
    -rw-r--r--   1 plesaux    users        34449 Nov  5 18:13  33.50 deg.
    -rw-r--r--   1 plesaux    users        34735 Nov  5 18:38  36.60 deg.
  • Shift may  also happen without any reinitialization between two measurements:

  • -rw-r--r--   1 plesaux    users        40614 Nov  5 17:16 34.40 deg.
    -rw-r--r--   1 plesaux    users        35214 Nov  5 17:26 29.90 deg.
    or with TIMMI untouched during one night:
    -rw-r--r--   1 plesaux    users        49491 Nov  6 19:03 38.30 deg.
    -rw-r--r--   1 plesaux    users        47115 Nov  7 08:53 35.10 deg.

    Out of suspicion we decided to iterate the test with a delay between each other iteration :
    here are the curves representing the drift of minimum position angle

    ( between two iterations the rotator returns to  encoder position 0 :
      reinitializing it resets minimun around 30 degrees : therefore while not reinitializing,
    rotator drifts a couple of degrees per test cycle )


    Three severe problems make the behaviour of TIMMI2 polarizer quite unreliable and require
    a mechanical diagnostic:
  • Hysteresys of the polarizer rotator.
  • Large drift of the minima (up to 60 degrees) when the polarizer rotator is initialized.
  • Small drift of the minima (few degrees) from one measurement to the next.



    On the good side it has to be said that initializing the polarizer carriage, or moving in and out does not seem to add any more troubles to the ones already mentioned.
    The scientific usefulness of the polarizer in this conditions is at best doubtful. Time and manpower must be found for it before the next polarimetric run. There seem not to be any polarimetric run until the end of period 74.