NRAO Home  >  Green Bank  |  Wiki Topic:    GB > Pennarray > MustangOof

Changes | Index | Contents | Search | Go

Requirements and Considerations for MUSTANG OOF

MustangOoofData -- data from Oct 2006 posted here

NewMustangOofData -- plots from 2007/2008 posted here

First thoughts during the GB OOF Workshop

• The out-of-focus PSF is different for each detector on the array.
• We need a reasonably gentle algorithm that doesn't destroy all the large scale structure in the image.
• Getting 1.4 Hz out with a gentle algorithm will require some thought.
• Scan pattern ideally would concentrate integration time away from the source (rectangular; or spirograph)
• Use of secondary tilts to increase scan speed? Probably a bad idea.
• We will give Bojan timestreams. Before that interface we will:
  • do a gain calibration (from cal). The 1-2% stability on 20 min timescales is ok.
  • perhaps a gentle 1.4 Hz and/or atmosphere cleaning

• TO DO: give Bojan a processed datastream from last season to play with.

Brief notes of optical simulations of MUSTANG on the GBT

These notes are follow up work by Simon Dicker from the GBT workshop amd the e-mail that was sent round. Please direct questions and comments to Simon (simon.dicker at gmail.com). It should be noted that further simulations may have to wait as I am moving between code v and zemax.

2. Qualitatively compare the OOF images that are obtained with and without truncation present in the CODE-V model.

While setting up the code V models I noticed several things.

1. Defocus of the secondary was made along an axis 36.7 degrees from the axis of the main mirror. A slight movement of the central ray on MUSTANG's focal plane (~1mm for 10mm defocus) could be seen. A positive defocus I took to be away from the primary.
2. A look between Bojan's phases and mine shoed they looked much the same.
3. MUSTANG's lyot stop is not perfect but fairly good. Until vignetting occurs the illumination pattern of the primary is quite constant both between fields and between focus positions. If we can assume a constant illumination it may make life easier but we will need to study this.

3. Quantitatively examine the level of truncation in the system

• Examine the cases of -15mm, 0, +15mm (because 15mm = 35mm * (3mm/7mm)). Identify what fraction of power is truncated by the various apertures (vacuum window, lens1, lens2, Lyot stop).
• Identify the maximum change in focus (from optimal focus) which causes no truncation, as a function of the pixel radius from the array center.

1. Eight positions in the focal plane were studied. The are (az,cross el measured in arcseconds):
   • field 1 : 0,0 (central pixel)
   • fields 2&3: 0,+-14.88" (top & bottom array edges)
   • field 4: 14.88",0 (left array edge)
   • fields 5&6: 14.88",+-14.88" (array corners)
   • fields 7&8: 6.4",+-6.4 (well away from corners of central 4 pixels)

1. Vignetting of the beam occured at the cryostat window for negative (towards primary) defocus and the 1st lens for positive (away from primary) defocus.
2. Vignetting first occured at these positions (to within 0.5mm): (see summary plot)
   • Field 1 (0,0); +-14.2mm
   • Field 2 (0,+14.88); 14.9mm and -8.1mm
   • Field 3 (0,-14.88); 7.5mm and -16.2mm
   • Field 4 (14.88,0);+-9.5mm
   • Field 5 (14.88,+14.88);7.4mm and -6mm
   • Field 6 (14.88,-14.88);5.4mm and -8.8mm
   • Field 7 (6.4,6.4); 14.1mm and -10.5mm
   • Field 8 (6.4,-6.5); 10.5mm and -14.7mm

4. Compare the amplitude and phase differences across the aperture of on-axis and off-axis pixels in a defocused position. Output the results as Zernikes on a 256x256 grid. First do it at Gregorian focus. Then repeat it at the Mustang focus. Finally, remove the truncating apertures in the optics chain and (but keep the Lyot stop).

1. These simulations were carried out on a 512 square grid as the outer half is empty (but there for FFT if wanted).
2. The maps are of the phase in the entrance aperture - where there is illumination of the primary (this can be assumed to be uniform) there is a phase value other than -9999 whic means no data/no field.
3. I did a 55 term standard zernike fit to the data (why not) and saved it. These are code V zernike's.
4. Defocus positions of +15mm, +9.56mm, 5.4mm, 0mm, -5.4mm, and -9.45mm were used (with -15mm on the GBT gregorian data). These are in files attached at the end of here. The three files are for the GBT only, the GBT with MUSTANG and the GBT with MUSTANG but with all apertures increased to 200mm (radius). For information on the format of the files ask Simon (or look at code V documentation). There are sets of files in each tarball - a large text file for each defocus containing the phase data and 8 other files of similar names containing the fits (note the numbers of the fit files are in reverse order to the field numbers). Positive defocus have files that begin with p and negative with 0 and the number is the defocus (eg m5_4 is -5.4mm defocus).

-- BrianMason - 27 Sep 2007; Simon Dicker - 14th Oct 2007

• GBT_gregorian.tar.gz: The gregorian defocus files.

• GBT_mustang_ap.tar.gz: OOF files from MUSTANG with real apertures

• GBT_mustang_no_ap.tar.gz: OOF files for MUSTANG with no cold apertures.

Attachment:	Action:	Size:	Date:	Who:	Comment:
GBT_gregorian.tar.gz	action	11791386	15 Oct 2007 - 16:19	SimonDicker	The gregorian defocus files.
GBT_mustang_ap.tar.gz	action	7350440	15 Oct 2007 - 16:20	SimonDicker	OOF files from MUSTANG with real apertures
GBT_mustang_no_ap.tar.gz	action	8773241	15 Oct 2007 - 16:21	SimonDicker	OOF files for MUSTANG with no cold apertures.
vignetting.pdf	action	11444	07 Dec 2007 - 00:37	ToddHunter