* Filter wheel 1 failures
------------------------
We did some tests to confirm accuracy of the filter wheel 1 movements
in positive direction.
They consist in the comparison between dome flats for a certain broad
band filter after performing several filter movements (the pupil stop
wheel was fixed during the test).
The test was done three times:
Filter wheel 1 was datumed before starting the movements. The filter
wheel was always moved in positive direction.
1) At UT~17:06; flats for filter J
files: r483768-r483775
After ~20 filter-wheel1 movements, there is structure the image
obtained by subtracting the flatfield taken after 20 movements
from the one taken just after the datum, showing poor accuracy in
the position.
2) At UT~20:45; flats for filter K
files: r483777-r483784
I used the script /home/whtobs/almu/acc to move the filter wheel
and take the flats.
22 movements after the datum, the image taken was a dark (not
a K flatfield as expected).
Previous flatfields look OK. No vignetting is noticeable
in them.
The 2 expected 'K-flatfields' after the dark look like a flatfield
in Z, H or J possibly with vignetting (not possible to distinguish
which band) and like a dark respectively (images r483783.fit
and r483784.fit).
3) At UT~04:46 ; flats for filter K
files: r483855-r483862
script: /home/whtobs/almu/acc
17 movements after the datum the image taken was a dark again (not
a K flatfield as expected). See file r483859.fit
However, in this test all flats are clearly affected by vignetting; even
2 movements after the datum.
The three expected 'K-flatfields' after the dark look like a
flat in Z, H or J with clear vignetting (r483860.fit), dark with
vignetting (r483861.fit) and dark (r483862.fit) respectively.
After this test I could not datum filter wheel1 again nor make
any filter wheel movements.
The EPICS I/O Controller rejected all commands.
Error mesage:
SYS> fwheel1 ingrid 1
Moving 'fwheel1' to '1'. Please wait...
WARNING: Mechanism movement failed.
Will perform single retry...
ERROR: Command failed: EPICS I/O Controller rejected command.
SYS> Raise_alarm Mar 4 05:39:06 taurus ingrid_mchcntrl[1182]: [ID 4
71664 user.crit] ALARM (mchcntrl.c, line 471) Command failed: EPICS
I/O Controller rejected command.
* 'Rectangular' vignetting
-------------------------
There is a 'rectangular' vignetting of the INGRID array.
It can be seen in images r483714.fit and r483855.fit.
The two images have similar level of counts (~3900 ADU in the
post-pre); however, the shadowed area of the array is bigger
or better focussed in image r483855.fit than in r483714.fit.
It is located in the right edge of the array (increasing in size
from top to bottom) and in the bottom-left edge.
Image r483714.fit was taken through filter J and r483855.fit with filter k.
The INGRID internal focus in image r483855.fit was -2000, but I do not
was the value was in image r483714.fit (the value recorded in the headers is
0.; which means that ifocus was either 0. or <-1000. or higher than +1000.
The telescope was moved in elevation during the night. In fact, the
zenital distances when images r483714.fit and r483855.fit were taken were
45 and 13.5 degrees respectively.
Could it be that the element causing the vignetting is not well fixed in
the instrument?
See Figures: r483714.eps
r483855.eps
*K emission from background
--------------------------
We still have a characteristic flatfield shape in the K-band.
There are still arcs. They are located aprox. in the same
parts of the array (i.e. up, left-down and left-right) but
there are a few differences:
o The inner radius of the feature is now at ~490 pix from the center
of the array. It was at ~350 pix from the center before the stand-down.
All the structure seems to be moved towards the edges of the field, which
is good.
o The intensity in all the three arcs was ~8-10% higher than the intensity
in the surroundings before the stand-down.
The arcs are now ~4% brighter than the surroundings in the arcs located
in the upper part and right-down corner.
However, the arc located in the left-down corner is ~16% brighter
than the surroundings.
The feature is therefore more asymmetric now.
The work done during the stand down has improved the situation we had before.
The new internal baffle seems to be blocking part of the unwanted reflected
thermal radiation. Could the asymmetry seen now be due to a misalignment of
this new baffle?
The flatfields in K also show now more 'vignetting' in the upper-right part.
o See files: flat.k.dark.nov01.norm.ps and flat.dark.k.norm.ps
for a comparison between the old (flat.k.dark.nov01.norm.ps)
and the new K flatfield shapes (flat.dark.k.norm.ps).
*Pupil imager (PI)
------------
files: r483805-r483818 (focussing)
r483799-r483804 combinations lights on/off and no_pstop/cass-k
r483819-r483828 combinations lights on/off and no_pstop/cass-j/cass-k/cass-s/cass-h
-images in focus (good news!) for INGRID internal focus -2000 microns.
(lower end of the INGRID internal focus at the moment)
For higher values of the images look out of focus.
It is worth it to have a look at the images!
-cass-k pupil stop is better focussed than the others (cass-j,
cass-h or cass-s). For cass-j/cass-s/cass-h we would probably need
an INGRID internal focus < -2000 microns.
-The best way to check aligment of INGRID or pupil stops to the telescope
was to obtain pairs of images (lights on and off) with the all the pupil
stops and the clear position.
All the images were taken through the K filter and with the PI in.
Filter wheel 1 and the pupil stop wheel were datumed before taking
the images. The wheels were also moved in the positive direction.
Images are:
r483819 pi in on nops ==> PI in; lights on; no pupil stop
r483820 pi in off nops ==> PI in; lights off; no pupil stop
r483821 pi in off cassj ==> PI in; lights off; cass-j pupil stop
r483822 pi in on cassj .
r483823 pi in on cassk .
r483824 pi in off cassk .
r483825 pi in off cass-s and so on
r483826 pi in on cass-s .
r483827 pi in on cass-h .
r483828 pi in off cass-h .
See:
nops-cassk2.eps
nops-cass-s.eps
nops-cassj.eps
nops-cassh.eps
These are differences no-stop minus stop image of the lamp_on minus lamp_off
pair differences (i.e. first compute the differences lights on minus lights
off for each stop and then the differences no_stop minus stop).
It is clear from them that there is a misalignment, but there are two possibilities:
INGRID could be misaligned with respect to the telescope or the pupil stops are
misaligned.
The misaligment is seen for all the 4 pupil stops used and it is seen always
approx. in the same direction: pupil stops are down and slightly towards the left
of the position the should have to match the secondary mirror.
Some differences are observed between cass-k, cass-h and cass-s when compared with
cass-j. The misalignment in the latter is bigger towards the left of the image.
This could be due to poor pupil stop wheel movements accuracy.
*Debris on the detector
----------------------
-The array is much cleaner now. The debris have dissapear from it.
A quick comparison of flat.k.dark.nov01.norm.ps with flat.dark.k.norm.ps
shows that they are not there now.