|
Flat Fielding
for Mimir in the JHK bands is normally performed in the dome against the
flat field screen. The following procedure may be followed and the numbers
listed used as verification that Mimir is operating normally. The integration
times have been set to achieve less than 50% of saturation, while being
short enough to avoid excessive dark+thermal current effects and long
enough (seconds) so that the first read of the CDS does not have a significant
slope across it (from a bright illumination and the 0.5 s readout time).
The following list assumes the user has already brought Mimir up normally.
This page
is organized into five sections:
- Basic
Set-Up
- Bias
Images
- F/5
Imaging Mode Flats
- F/5
Polarimetric Mode Flats
Note that
taking data for the linearity correction is located on another page (click
this link).
For advanced
users, the following table summarizes key settings:
|
Mode
|
Band/Filters
|
F/5
|
F/17
|
|
Illuminator
Setting
|
Integration
Time
|
Illuminator
Setting
|
Integration
Time |
|
Imaging
|
J
|
18
|
5.0
|
30
|
5
|
|
H
|
18
|
0.7
|
30
|
1
|
|
K
|
18
|
0.7
|
30
|
1.5
|
|
H2-on
|
25
|
4
|
60
|
10.0
|
|
H2-off
|
25
|
4
|
60
|
10.0
|
|
Polarimetry
|
H
|
18
|
1.6
|
TBD
|
|
Basic Set-Up
- Turn
off the sun
- that
is, dome flats should be done at night since sunlight scattered
into the dome will change brightness with time. While this may not
harm imaging flats, it render polarimetric flats useless.
- Close
the dome slit
- same
reason
- Point
the telescope at the flat field screen
- this
is done by issuing the FS command in MOVE
- after
motion has stopped, use the hand paddle to fine position the telescope
to accurately point at the flat screen.
- Open
the mirror cover
- Open
the instrument cover
- Turn
on the flat field illuminator light
- Select
the "Narrow Band" option of the switch marked "Broadband
/ Narrow Band"
- Turn
to "On" the other switch marked "On / Off
/ Test"
- Set
the black rotatable knob (located just above the flat field illuminator
switch plate) as indicated in one of the sections below
Bias Images
It
is vital that users obtain bias frames (zero exposure darks) before their
flat field images and after the flat field images have been acquired (for
all modes).
- Take
a set of Bias Images
- Put
a value of "32" in the argument of the "go_bias"
button and push that button
F/5
Imaging Mode Flat Fields (for F/17, substitute
values given in table above)
- For
the J filter (J+PK50)
- Select
the J filter, by clicking the "set_J" button
- Set
the integration time to 5.0 seconds
- Put
the black knob setting to "18"
- Take
a test image
- the
mean count level should be about 4,000 counts
- Populate
the FITS header
- "J-Band
Dome Flats; lights On" , "J+PK50"
- Select
N_exp = 32
- Press
"Go" to collect 32 images
- After
this "lights on" sequence has finished, turn the illumination
lights off using the "On / Off / Test" switch
- Change
the FITS header to indicate the lights are off
- Press
"Go" to collect 32 images
- the
mean count levels at the image center for the lights off condition
should be about 115 counts
- For
H-Band (H+PK50)
- Select
the H filter, by clicking the "set_H" button
- Put
the black knob setting to "18"
- Set
the integration time to 0.7 sec
- Repeat
the steps 2-4 through 2-10 above
- lights
on should give about 4,000 counts/pixel
- lights
off should give about 35 counts/pixel
- For
K-Band (K+PK50)
- Select
the K filter, by clicking the "set_K" button
- Put
the black knob setting to "18"
- Set
the integration time to 0.7 sec
- Repeat
the steps 2-4 through 2-10 above
- lights
on should give about 4,000 counts/pixel
- lights
off should give about 450 counts/pixel
- For
H2ON (H2ON+PK50)
- Select
the H2-ON filter, by clicking the "set_H2On" button
- Put
the black knob setting to "25"
- Set
the integration time to 6 sec
- Repeat
the steps 2-4 through 2-10 above
- lights
on should give about 4,000 counts/pixel
- lights
off should give about 240 counts/pixel
- For
H2OFF (H2OFF+PK50)
- Select
the H2-OFF filter, by clicking the "set_H2Off" button
- Put
the black knob setting to "25"
- Set
the integration time to 5 sec
- Repeat
the steps 2-4 through 2-10 above
- lights
on should give about 4,000 counts/pixel
- lights
off should give about 240 counts/pixel
F/5 Polarimetry
Mode Flat Fields
The following
will provide a set of polarimetric flats sufficient to achieve 0.8%
photometric uncertainty in each pixel for each HWP, or under 0.01% polarization
uncertainty for stars spanning 16 pixels (about 1.6 arcsec FWHM). For
cruder uncertainties, reduce the number of exposures or number of cycles
(step 9).
As a benchmark,
the full set of observations takes about 2 hours of clock time.
- Set
up Polarimetry Mode
- In
LOIS window, "cd ../tcl.mimir"
then "source gpips.tcl"
- Click
"Show User Buttons"
- Select
H-band filter, PK50 blocker, H-band Half-wave Plate, and Wire Grid
Analyzer
- Click
the "set_POL"
button
- Test
the HWP rotation
- Click
the "pol_reset"
button
- Put
the black knob setting to "18" with the illuminator lights
turned on to the narrow band setting
- Set
the integration time to 1.6 sec
- Take
a test image - the mean counts should be around 4,000 ADUs. If greatly
different, adjust either the light setting or the integration time
to achieve a mean of about 4,000 ADUs.
- Run
the pol_FF_cal polarimetry data collection script
- Do
a "go_bias 32" first
- Edit
the object title to say "Pol Flat Lights On"
- Find
the "pol_FF_cal"
user button and set the exposure time ("expt") to 1.6
and the number of exposures ("nexp") to 4
- Click
this button to collect 32 biases, then the polarimetry data and
to move the HWP to different positions taking
4 exposures at each position
- After
the script ends, click the pol_FF_cal button,
again with the lights still on
- Turn
the illuminator lights off
- Run
pol_FF_cal twice with the lights off (edit the object title to "Pol
Flat Lights Off")
- Repeat
steps 5-8 three times
- totaling
32 lights on and 32 lights off images at each of 16 HWP angles.
DPC 20070927
|