AAO Instrument Scientists

UHRF Instrumental Tests

This section describes test observations taken with the UHRF. These notes are likely to be of use only for staff scientists.

1. First Observations with MITLL2 + UHRF on 18/04/98
2. Further MITLL tests on 09/06/98

1. First Observations with MITLL2 + UHRF on 18/04/98

CCD Orientation and File Sizes

The MITLL2 was first used with UHRF on 18/4/98. The detector was tested during the daytime and used in the evening for scheduled observations.

The CCD was mounted with the 4K axis in the dispersion direction.

The MITLL2 pixel size = 15 x 15 microns. The detector size = 60 x 30 mm = 4096 x 2048 pixels, normally operated at 160K

On the XMEM display: x-axis = spatial direction, y-axis = dispersion direction, top = red, bottom = blue. To move the line rightwards increase ug. On the XMEM 1 cm ~ 200 ADU.

I then set up a CCD window which was binned by 4 in the spatial direction. Window = disk$raw:[observer]mitll_jw. The full, unbinned MITLL2 window is mitll_full.

Using an uncompressed data format (data float) the file sizes are 34 MBytes for full chip and 8 MBytes for chip binned x 4.

Using the CCD command data ushort writes the data in compressed format. This gives a files size of 10.4 MBytes for the full, unbinned chip.

Set up on laser line

The configuration file laser_6328 was used for initial tests @ 3E5 detector position. The laser line was well centred on chip. To focus the collimator I used a modified version of foc_unix.com as normal. This worked fine.

Focus scale with MITLL @3E5: 1 pixel = 106.3 ADU (i.e. 170 x 15/24).

Resolution test on laser line: FWHM = 3.10 bins gives R = 319,286 when used with slicer. Close to the expected value.

Slit mode Observations

The observations taken on 18/4/98 (by Webbet al.) used the slit in place of the image slicer.

This required a semi-hybrid mode : Observing software was set up as UHRF but the TV was placed in the UCLES position. The UCLES beam rotator could then also be used. One complications is that the UCLES slit and UHRF slicer are NOT on same optical axis. This means that the UHRF arc lamps do not illuminate the slit correctly and need to be positioned manually. The laser lamp is visible throught slit in UHRF mode but the ThAr lamp is too far off the slit axis.

For this mode the collimator focus should be done using the slit as the slit focus position is slightly different to the slicer position. (The offset between the slit and slicer focus positions was 6 pixels).

Resolution tests with slit:

SW = 0.620 arcsec gives R = 104,732
SW = 0.074 arcsec gives R = 290,000

Set up for new lambda

Went through the set up procedures to get a new solution for 5422.7 A. This process was very slow but worked OK.

To get the sdf frame in correct orientation for Figaro tasks - just use figaro rotate.

Dark Current

During the Webb observations there appeared to be some excess dark current contribution in frames. This was however not well established as no dark frame was taken during the night.

Dark current test taken the following morning -

5 x 3600 dark frames
5 x bias frames

mean dark + bias = 488.92
mean bias = 477.73

So the dark contribution over 1 x 4 pixels
= 11.5 ADU in 3600 s
= 2.9 ADU/pix/hr
= 1.05 e-/pix/hr

The expected dark rate from Barton notes = 0.3 e-/pix/4000s. So there was some evidence for slightly higher dark level. Level at start of night was probably higher. Frames level (object + sky + dark + bias) = 520 ADU approx which would give an upper limit of 4 e-/pix/hr.

2. Further MITLL2 tests on 18/04/98

Splicer MASK

To test the full number of image slices, three frames were taken with the quartz lamp and the UHRF mask IN, OUT, IN

frame 1:    nslices = 27  slice  1  @y = 1800 x =  558
mask IN                   slice 27  @y = 1800 x = 1614

frame 2
mask OUT:  nslices = 35   slice  1  @y = 1800 x =  464 
                          slice 34  @y = 1800 x = 1836 



frame 3    
mask IN:   nslices = 24  slice  1  @y = 1800 x =  632
                         slice 24  @y = 1800 x = 1582

At 1E6 and without the mask, a single order covers the full CCD quite well. At 1E6 could observe without a mask. At 3E5 and 6E6 would probably have order overlap without the mask. Note that the mask was not put back optimally.

Vignetting Problem

In June 1998, frames taken with quartz lamp clearly showed sharp vignetting at each end. This was due to the camera shutter being too small for the long axis of the CCD. For 1E6 data, the measured usable region was from pixels ystart = 400 to yend = 3712 out of 4096 pixels, so approximately 10 per cent of the frame was vignetted at each end.

(Dec 1998: A new, larger camera shutter was ordered - and is available at site but not yet installed.)

Resolution Test

The MITLL2 was binned by 4 in the spatial direction and focussed on the laser line at the frame centre. I then took frames with the laser line at 4 offset locations on frame, with a shift in dispersion direction only between each position.

The figaro emlt routine was then used to measure the FWHM in bins fo3 3 positions across the line.

Results:Wavelength increases towards LOWER y-values

Position Xval   Yval   BINS
____________________________
1        160    592    3.77
1        256    592    3.50
1        350    592    3.42
____________________________
2        160   1165    3.45
2        256   1165    3.43
2        350   1165    3.55
____________________________
3        160   2000    3.32
3        256   2000    3.62     chip centre
3        350   2000    3.55
____________________________
4        160   3084    3.55
4        256   3084    3.47
4        350   3084    3.75
____________________________
average of all values  3.53 bins

The dispersion = 0.1473 A/mm so 3.53 bins = 7.799E-3 A giving R = 811,330.

The resolution was significantly lower than expected but appeared to be consistent across the chip.

Throughput test on standard star

Star = HR7596, lambda = 5576, T = 20 mins, AB = 5.62 mag, seeing ~ 3 arcsec Res=1E6, HA = 2:27

Rough estimate gave S/N = 30 in 20 minutes per dispersion pixel.

Lower than previous values due to poor seeing.

Page created: 29th December 1998 by Jessica Chapman

Last Updated by Jessica Chapman on 29th December 1998 (jmc@aaoepp.aao.gov.au)