SDSS Spectrograph Guide Sensor

Patrick Waddell

---

Table of contents: 

• Guider functions
• System requirements
• Design overview
• Camera parameters
• Fiber bundles parameters
• 2-Axis Calibrated Accuracy
• Guide Sensor Performance
• Transverse Position Error Budget for science fibers
• Guide Sensor Sky Coverage/ Star Availability
• Guide Star Selection Criteria
• Sky Selection Criteria
• Photographs
• Figures
  • Guider block diagram
  • Guide harness in slit head (front view)
  • Guide harness in slit head (side view)
  • Guide camera on spectrograph
  • Guider harness
  • Guide fiber organization and reach at plug plate
  • Guider S/N performance as a function of star temperature (Mv=15.5)
  • Guide Fiber Bundle Transmission

Guider functions:

• 4-axis info. for field acquisition, pointing accuracy and stability:

  X, Y, rotation, scale

• Pointing information for ~1 hour integrations
• Image quality monitor
• Sky brightness monitor for spectrograph exposure control
• Wide FOV for one guide star for initial pointing

System requirements:

• Use objects from photometry survey bright limit (m_v=14.25 - 15.25)
• 15 second exposures
• Plug plate holes for guide stars are identical to those for science FO

Design overview:

• Acquire images at plug plate surface via coherent FO bundles
• Coherent bundles terminate at slit head wall
• Sensor camera, mounted to one spectrograph chassis, images FO bundles
• Low power TE cooled Photometrics Ltd. Sensys camera
• Camera optical bandwidth is selected to match atmospheric refraction of spectrograph band (thick Kodak CCD + 2 mm BG-38)
• Accuracy and 4-axis capability is achieved via 10 stars/field:

  - 2 stars at plate center --> one is selected for image quality monitor

  - 8 stars arrayed at ~1.2 deg. radius

• FOV for each of 9 guide stars is 7.4 arcsec dia. (0.45 mm FO, Sumitomo IGN-05/10)
• One guide bundle is 12.4 arcsec dia. (0.75 mm FO, Sumitomo IGN-08/30)
• A second large FO bundle, 12.4 arcsec dia., monitors sky brightness
• Use guide star selection to avoid proper motion problems and close neighbor confusion
• Camera/fiber registration may be calibrated during spectrograph

  flat field calibration or by lookup from previous calibration

   

   

   

   

   

Camera parameters:

Image scale:	60.6 µm/arcsec
Assumed sky background:	4.1 e- s-1(arcsec2)-1 [ 21.7 Mv/arcsec2, Solar]
Sensor optical bandwidth:	450-650 nm (thick CCD + 2 mm BG-38)
Sensor detector parameters (binned 2x2):
Detector type:	Kodak CCD, Thick (frontside illum.)
Readnoise:	13 e- rms pixel-1
Dark current:	2.8 e- pixel-1 s-1
Array geometry:	384 x 256 (768 x 512 - engineering mode)
Pixel size:	18 µm sq (9 µm sq - engineering mode)
Pixel sampling:	0.30 arcsec/pixel

 

Fiber bundles parameters:	IGN-05/10	IGN-08/30
Diameter of fiber bundle (mm):	0.45	0.75
Distance between neighboring cores ( µm):	4 - 5	3 - 4
Number of cores:	~ 10,000	~ 30,000
Net throughput efficiency (%)	~25	~25
Numerical aperture:	0.35	0.35
Core material:	GeO2 +silica	GeO2 +silica
Cladding material:	F + silica	silica

2-Axis Calibrated Accuracy:

Accuracy with which each measured guide star centroid ties the focal plane reference (plug plate) to the astrometric reference frame

Error Source	mm RMS	Milliarcsec RMS
Guide star astrometry	TBD (6 mm)	(100)
Accuracy of FOV distortion calibration	TBD (10)	(165)
Plate deformation	5	83
Hole drilling	9	149
Ferrule/hole concentricity	8	132
Fiber bundle/ferrule concentricity	7	116
Calibrated fiber position (output end)	TBD (2 mm)	(33)
Camera/fiber flexure effects (over 60 min)	TBD (2 mm)	(33)
Centroid calculation and detector flaws	TBD (6 mm)	(100)
Total	TBD (20)	(331)

Guide Sensor Performance: (2-Axis, RMS) (using any fiber cartridge)

(per sensor star)

•Calibrated Accuracy:	TBD (0.331 arcsec)
•Sensor Noise (per 15 seconds):	TBD (0.06 arcsec)

 

(sensor ensemble)

•Calibrated Accuracy:	TBD (0.110 arcsec) (7 µm)
•Sensor Noise (per 15 seconds):	TBD (0.02 arcsec)

Transverse Position Error Budget for science fibers:

Error Source	µm RMS	Milliarcsec RMS
Science object astrometry	TBD (6 µm)	(100)
Accuracy of FOV distortion calibration	TBD (10)	(165)
Plate deformation	5	83
Hole drilling	9	149
Ferrule/hole concentricity	8	132
Fiber bundle/ferrule concentricity	7	116
Temperature gradients	5	83
Guide sensor	TBD (7)	(110)
Total	TBD (21)	(340)
Max desired (10% of core dia.)	18	297

 

Guide Sensor Sky Coverage/ Star Availability:

• At r' = 14 to 15, model (Bahcall & Soniera) and data (Stoughton) indicate

  ~ 60 stars/sq. degree/mag. at high galactic latitude

   

However, to meet g'-r' criteria (blue objects), throw out ~ 2/3

Assume other criteria, close neighbors, etc., eliminate another factor of 0.5

We have a density of 7 stars, for each of 10 evenly distributed samples -->

yielding 99.5% probability of finding a star in each sample area -->

We should have a sufficient number of available guide stars.

 

Guide Star Selection Criteria:

• Guide star brightness must range between m_v = 14.25 to 15.25
• To ensure that proper motions over 10 years will be < 200 marcsec, guide star color index: 0.4 < g'-r' < 0.9 (G0 through K5 stars, 5900 to ~4400 k)
• Selected guide stars must have no neighbor stars within TBD (5 arcsec) radius that have a brightness within TBD (5 mag.) of guide stars

Sky Selection Criteria:

• Select an area of sky which is within TBD (0.7 degree) of plate center, which has no detectable objects to the photometric survey bright limit in r' + i' bands. 
• Selected area must have no near neighbor, within TBD (7) arcseconds, brighter than TBD (r'=23). 

Photographs:

Figures:

Guider Block Diagram

Guide harness in slit head (front view of slit)

Guide harness in slit head (side view)

Guide camera on spectrograph

(Top and Side views / reimaging optics not shown)

Guider harness

Guide fiber organization and reach at plug plate

Guider S/N performance as a function of star temperature (Mv=15.5)

Guide Fiber Bundle Transmission