Transcript Document

B-pol optical configurations
Telescopes and Focal plane options
B. Maffei for the B-pol collaboration
G. Pisano
A. Murphy
T. Peacocke
Configurations
Requirements
Resolution goal ~ 0.5 deg at 100GHz
Large focal plane  no or very low FP curvature
Low systematic effect
• distortion, ellipticity, cross-polarisation
• beam homogeneity across FP
• Similar beam for both polar. orientation
Let’s assume an imager
Optics
Needs telescope: refractive (lens) or reflective (mirrors)
Focal plane options – Beam formation
Feed horn arrays
Bare detectors
• Antenna coupled detectors
• Lenslet
Refractive telescope configuration
Goal for any telescope: fit as many detectors possible + low sys effects
SAMPAN and EPIC/Spider studies
BICEP and Re-imaging optics of QUAD (working instruments)
 Needs 2 lenses for low FP curvature
At 217GHz
Aperture
Stop
Lenses
Telecentric
Focus
SAMPAN study:
Lenses material: HDPE, Silica or Germanium
Mass: from 3 to 8Kg – up to 30Kg with mount
Reflective telescope configurations
QUAD: Cassegrain with
secondary supported by Zotefoam
Pros
• on-axis
• Edge pixels are similar
Against
• Secondary mount
• Needs re-imaging optics for low
FP curvature
Planck + many others
Gregorian off-axis with D-M condition
Curved focal plane
Reduced FP size
Possible for large arrays
Compact test range configuration
Used in many CATR + Clover and Quiet
Large Focal plane possible
Example: Clover
• FP diameter = 250mm diameter
• Size limited by filter diameter not
by aberrations
• Flat FP
• 256 horns @ 150GHz
• 165 horns @ 97GHz
• Edge pixel eccentricity ~ 0.02
• Optical configuration allows good
baffling
Design for B-pol
Focal
plane
Sec
Primary
~ 40cm
First order comparison
(based on simulations – Ideal telescope)
Will need the same aperture size in either case ~ 30-40cm
Lens based system advantages:
Possibly Mass ? (maybe not an issue if telescope + mount in CF)
Compactness
On axis:
• Symmetry of pixel beam characteristics
• Advantage for scanning strategy ?
Mirrors advantages
Achromaticity
Lower Losses
Comparison
Lenses need to be cooled (larger dewar) to reduce emissivity
Mirrors do not have to (already low emissivity), BUT if we want to cool
the telescope to lower background will need much bigger dewar
Real systems
Telescopes are not perfect: will produce systematic
effects
Even if mirrors affect the beams, the effects are very
well know (predictions and measurements)
Lenses imperfections knowledge is not that well
advanced in the microwave range
Reflective telescope systems
Planck RFQM test with a high dynamic range: telescope + feed horns
Alcatel-Alenia Space
X-polarisation beam at 100GHz
Co-polarisation beam at 100GHz
Main beam
Planck-RFQM – mid and far sidelobes
Alcatel-Alenia Space
Far sidelobes
Moon rejection
Earth rejection
Intermediate beam cuts
Raw measurements
Measurements
dynamic limit
Predictions
Sun
rejection
Lenses knowledge
Several big lenses have been developed and used so far (QUAD,
BICEP)
Both are using Polyethylene
Max transmission (with A/R coating) ~ 99% in ideal case (no loss)
Problems seen on both experiments
Non systematic variation in pointing differences for detector pair (PSB) 
strong suspicion on extra lens effect (bi-refringence ?)
Models to predict performances lack accuracy
 not well simulated at the moment
Performances and models will improve but
current lenses are causing problems still
A/R coating
Coated UHMW polyethylene lens
operated at 4K for QUAD
Single layer works fairly well at 4K but not large
band enough
If multi-layer A/R could achieve a much larger
BW, we do not know how it will behave when
cold
Comparison summary
Mirrors
Lenses
Size due to off-axis config.
Inhomogeneous for large diameter ?
Different pixels might see different parts of the
lenses  beam variation
Needs mount adds weight
Bi-refringence (increased when cooled?)
Not telecentric  non symmetric beam
variation across FP
Chromatic aberration + A/R coating Band-Width
 Might need different telescopes (see Spider)
If needs to be cooled large dewar
Standing waves - even with 99% transmission
Material ? : either problems or lack of data
Well understood and modelled
Lack of knowledge and many systematics  will
heavily rely on extremely good calibration
Experience on manufacture and use
Properties variation with T  ground calibration ?
Thermal gradient across diameter
Clover design (CATR) shows very low
xpol and aberrations across large FP
More compact  but needs to be cooled
If all problems solved and/or understood
(feasible?)  potentially better system
Focal Plane based on feed horns
Well known and understood technology
Very good performances and low systematics
Low cross-pol - below 40dB typically
High efficiency, better than 25dB return loss
Low sidelobes and good control of straylight
The Good
Works with various detectors (Bolos and HEMT)
The bad
Mass / volume
Reliability/yield for mass production
Several design/manufacture possibilities
are being investigated
Large aperture diameter (4-8 l)
Reduced number of pixel – hundreds
The ugly
Antenna coupled bolometer
Adopted for several new instruments but poor
amount of data so far
The latest data right now (that I know) are from JPL
Talks from J. Bock
Publications by Kuo et al.
Main advantages
Reduced size
• SAMPAN study: 20000 pixels for 165mm diameter FP
Fairly easy filtering (intrinsic filters)
Antenna coupled bolometer beams
Measured Beam Patterns
Single polarisation
antenna
X-polarization
Single polarisation
antenna + filter
Dual polarisation
Dual polarisation
Y-polarization
antenna + filter
antenna
Radiation pattern
(meas.) of antenna
coupled detectors
FT of radiation
pattern with cold stop
Kuo et al, 2006
Ant. coupled bol. spectral response
Long slot single
pol. antenna
Dual pol. antenna
Single pol. Antenna
+ BP microstrip filter
Dual pol. Antenna +
BP microstrip filter
Measured Spectral Response
Low efficiency due to reflections in test system.
After correction
Lenset
The idea is to put a small lens on top of each detector.
No data available
We will have the problems related to lenses (see
previously)
Possibly more serious problem
the lens is small any small imperfection (fraction of the
wavelength) will potentially impact the beam
Bare detectors potential Pbs
Cross-talk ?
Staylight ?
Use of cold stop: how each detector across focal
plane will be affected ?