Quality Control and
Data Processing

The minimum Detector Integration Time

The minimum Detector Integration Time [DIT] is determined by the configuration of the electronics of the detector system.
The HAWK-I system was initially configured such that the minimum DIT was 1.2572 seconds.

When	Value [seconds]
2007-04 -- 2009-05-06	1.2572
2009-05-07 -- present	1.6762

Eliminating Crosstalk: 2009-05-07 After approximately one year of operations and feedback from the user community, in order to address concerns regarding Crosstalk (see below) the confguration of the electronics was adjusted with three consequences:

• Elimination of crosstalk
• Reduction of Readout Noise
• Increase in minimum DIT (from 1.2572 to 1.6762 seconds)

• Crosstalk between amplifiers: [All data prior to 2009-05-07, i.e. all periods prior to, plus part of P83] Crosstalk between the amplifiers of the chips produces a series of 'crater-like' artifacts arrayed horizontally with respect to each 'bright' star in the field.
  By adjusting certain voltages of the detector electronics this effect has been virtually eliminated. In addition this change reduced the Readout Noise of the detectors, but increased the minimum-DIT from 1.2572sec to 1.6762sec. This change was applied at approximately 2009-05-07T21:30.
• Change of the detector read speed 2009-05-07 : The HAWKI detector read speed has been changed for reasons to reduce the electronic ghosts.

  Before 2009-05-07T21:30 it was:

  HIERARCH ESO DET VOLT1 DC6   = 2.4000 / Set value DC-Voltage
  HIERARCH ESO DET VOLT1 DC7   = 3.4000 / Set value DC-Voltage
  HIERARCH ESO DET RSPEED      = 6 / Read-Speed Factor
  HIERARCH ESO DET MINDIT      = 1.2572000 / Minimum DIT                
  

  From 2009-05-07T021:30 on it is:

  HIERARCH ESO DET VOLT1 DC6   = 2.1000 / Set value DC-Voltage
  HIERARCH ESO DET VOLT1 DC7   = 3.6000 / Set value DC-Voltage
  HIERARCH ESO DET RSPEED      = 8 / Read-Speed Factor
  HIERARCH ESO DET MINDIT      = 1.6762000 / Minimum DIT     
  

  Consider 2009-05-07T21:30 as a breakpoint to associate calibrations. Do not use twlight flats or darks acquired before 2009-05-08 to calibrate science data acquired after 2009-05-08 and vice-versa.

Pupil Ghosts

When

Between the last Commissioning/Science Verification run (which ended January 30th) and the start of operations in P81 (Apr-Sep 2008), the entrance window to the instrument cryostat suffered a significant degradation and operations began in less than optimum conditions. The problem was resolved during an intervention conducted 2008-08-20 -- 2008-08-24.

All data acquired between 2008-04-01 and 2008-08-20 are affected by this problem, though to what degree generally depends on the circumstances of the observations, though often in an apparently random way...

What

During the Aug intervention it was discovered that the entrance window degradation was caused by a greasy film. The greasy film is believed to have been deposited during the 3 month period between the end of Commissioning/Science Verification and the start of observations when, for technical reasons, HAWK-I was kept under vacuum but at room temperature, instaed of it's usual cryogenically cooled state.

The degradation caused TWO effects to be evident in reduced SCIENCE data:

• Pupil-Ghosts: This was the first noticed, and most noticable, effect of the degradation, hence the name of this web-page. This problem tends in general to be worse when the pupil angle changes quickly between individual DITs. However there are plenty of exceptions to this rule, i.e. data sets with small changes in pupil angle between each DIT where the effect is none the less strong.
• Sudden Lateral shifts: Randomly, there appear suddenly between two DITs 'sudden latteral shifts' in the alignment between entrance window and detectors, thus changing the pattern of the degradation on the detectors and thus affecting the sky subtratction for the science. This effect could affects both data which show strong pupil ghost effects as well as data with weak or even non-existant pupil ghost effect.

Example data from April 2008 where the Pupil Ghost pattern and the effects of a 'sudden lateral shift' are clearly present [AB = HAWKI.2008-04-29T02:47:25.267_tpl.ab]:

RAW frames

RAW frame N minus RAW frame 1, N=[2,TPL.NEXP]

Product frame

Radiation-Induced Charge Collection

When

As reported on 04/06/2007 in the HAWKI Test report on the Detector and Acquisition System (VLT-TRE-ESO-14800-4062) detector with DET.CHIP.NO = 2 ( DET.CHIP.ID = ESO-Hawaii2RG-chip78 ) is subject of radioactive events. Here we report on the number and the character of the radiation-induced charge collections.

The following data have been analysed:

• 10 consecutive raw dark frames with DIT=2.0 sec, NDIT=1, DET.NCORRS.NAME=NonDest, acquired on 2008-06-17, 2008-10-20 and 2009-02-20
• 5 consecutive raw dark frames with DIT=60.0 sec, NDIT=1, DET.NCORRS.NAME=NonDest, acquired on 2008-06-17, 2008-10-20 and 2009-02-20
• 3 consecutive raw dark frames with DIT=300.0 sec, NDIT=1, DET.NCORRS.NAME=NonDest, acquired on 2008-06-17, 2008-10-20 and 2009-02-20

What

Fig.1 Difference of two DIT=300sec raw dark frames taken on 2009-02-20.

Fig.2: Shape of the strongest event with 730 ADU peak counts.

+--------+--------+
|   4    |   3    |
| chip88 | chip79 |
+--------+--------+
|   1    |   2    |
| chip66 | chip78 |
+--------+--------+

Fig.3: HAWKI detector head layout and naming scheme (chip number: 1 to 4, and chip id: 66, 79, 79 and 88)

Fig. 4: Details of a difference between two consecutive dark frames (DIT=300, NDIT=1) showing several weak events.
The right head of the event amounts to more than 8000 ADU.

Summary: From the events total flux, three frequently occuring types of events can be distinguished:

• weak events with peak = 10 ADU. This is by far the most frequent type of event.
• intermediate events with peak 100 ADU.
• strong events with peak 500-700 ADU (2 found in total in the sample).

Events can show up in any kind of shape from circle-like, ellipsoide, comet-like up to 400 pixel long narrow trails. There is no prefered orientation of the trails (derived by visual inspection).

The number of events is linearly increasing with DIT (exposure time) , The event rate is is about 13 events per minute. The event rate is constant within the last 8 months.

The large number of diffuse events with highly elongated shape which are so far homogeneously distributed over chip #2 with uncorrelated trajectory orientations gives rise to the suggestion that the radioactive source itself is close to the detector plan, and that the source is extended.

The few particle trajectories derived from the two neighboring chips #1 and #3 do not point to chip #2. This is interpreted in the way that the radiation registered on chip #1 is not emitted from chip #2, but emitted in the same way from a source layer in chip #1 with a much smaller event rate of < 1 event per minute. The same holds for chip #3. Chip #4 does not show any stochastic events of this kind.

For a DIT=300 NDIT=1 dark or science exposure, chip #2 registers about 65 events. No detailed distribution of the event characteristics like flux, shape, position is availabale. The ESO HAWKI data reduction pipeline does not correct for these events.

Central Flare-like Feature

When

The Central Flare-like Feature appears and disappears intermittently, usually for a period of some months at a time.

The following periods clearly have the feature:

1. 2008-04-28 [P81] -- 2009-03-05 [P82]
2. 2010-12-08 [P86] -- 2011-04-05 [P87]
3. 2011-04-22 [P87] --
4. and others; since P87 the list has not been updated

At the centre of the array, i.e. where the four chips meet, there is a "flare-like" structure. It becomes evident in frames with DIT greater than about 30sec.

The presence of the feature is monitored in the 60 and 300 sec Daily Health Check [DHC] DARK frames.

The following images are montages of the central regions of the the 300 sec DHC DARK frames:

2008-04-27 -- 2008-09-27	2008-09-28 -- 2008-12-02	2008-12-03 -- 2009-02-16
2009-02-17 -- 2009-05-03	2009-05-04 -- 2009-07-05	2009-07-06 -- 2009-09-18
2009-09-19 -- 2009-12-08	2009-12-08 -- 2010-02-13	2010-02-14 -- 2010-04-25
2010-04-26 -- 2010-07-08	2010-07-09 -- 2010-09-16	2010-09-17 -- 2010-12-11
2010-12-12 -- 2011-02-22	2011-02-23 -- 2011-04-24