Optical Qualification and Plans for PFM

1
Optical Qualification and Plans for PFM

• N. Geis
• MPE

2
FPFPU Alignment

• Alignment precision achieved for CQM
  (PACS-KT-PR-014)
• Alignment made at KT, at ambient, with VIS light.
  
• All offsets refer to TEL focus/LOS Meas REQ
• Alignment Cube ? OB lt 0.3 mm position 0.5 mm
• (for satellite alignment) lt 0.4 angle 1
• OB ? detector arrays (VIS) lt 0.3 mm position 0.5
  mm
• (internal alignment) lt 5 angle 1
• ( linear resolution limit at Lyot stop)
• M2 ? ext. Alignment Mirr lt 0.3 mm position 0.5
  mm
• (to transfer alignment to MPE OGSE) lt 0.3
  angle 0.5
• except 4, all requirements are met gt Repeat
  for PFM.
• 5 angle could be improved to 3, but IIDA/B
  requirement of 1 unlikely to be
  verifiable/achievable

3
FPFPU Alignment

• Alignment precision achieved for CQM (ILT tests)
• Descoped program allowed neither characterization
  of OGSE, nor of CQM.
• Measurement precision achievable with test
  equipment is lt 0.2 mm position lt 0.3 angle
• ? OGSE suitable for verifying at the PFM
  requirements level.
• 1) Test Optics ? Cryostat OB not done
• 2) Test Optics ? FPFPU done (see next
  slide)
• by evaluation from 1) and 2) ?
• 3) FPFPU ? Cryostat OB not done (missing 1...)

Full PFM verification program should suffice to
demonstrate PFM alignment at requirements
level. However
4
FPFPU Alignment

• Caveats
• Known dimensional problems in OGSE, affecting the
  internal calibration path.For CQM, ad-hoc
  (large) alignment corrections had to be
  applied.For PFM tests, tracking and fixing of
  problem may be necessary. gt Manufacturing
  changes in OGSE and re-alignment ( 2 month
  effort).
• Stability of Test Cryostat 1. Unexpectedly large
  shift (4mm) in LOS to external ( point) sources
  on cool-down of test cryostat.2. After warm-up
  a residual LOS shift of 0.43 mm remained.
  (This effect may be responsible in part for 1 mm
  boresight shift observed between VIS and IR
  cooldowns -- other contribution mechanical
  distortion of chopper assy, defocus see 3.)3.
  Tilting the cryostat shifts the external LOS by 2
  mm (caused by hardware failure?).4. Possible
  defocusing of external sources by 10 mm (reason
  unclear)
• Either - fix these shifts by hardware mod.
  (unlikely) (1-2 month delay of OGSE test program)
  - characterise effect (repeatability) -- may not
  be repeatable (1 month / cool-down)- modify test
  cryostat to allow monitoring/compensation of
  internal alignment during ILT measurements.
  (1-2 month delay of OGSE test program)

5
FPFPU Optical Performance

• FPFPU Itself
• The Good
• Fairly sharp images, in photometry and
  spectroscopy modes, close to expectations (see
  presentation by E. Sturm). Peak Strehl not
  reached within nominal focussing range of OGSE,
  PSF FWHM 30 too wide.
• ? optical system basic performance qualitatively
  verified, however focus mismatch between OGSE
  focus and PACS focus, reason unknown.
• Reduction of PSF peak height should be easily
  noticeable in the 105 mm photometer band when
  defocussing by 3 mm. (informal test, but
  confirms the focussing requirements for best
  performance claimed by PACS in the IID-B)
• Further details are still under analysis, but no
  show stoppers have been flagged in the imaging
  area as yet.

6
FPFPU Optical Performance

• FPFPU Itself
• The Ugly
• Grating Ghosts
• Full Scans of the spectrometer section during CQM
  ILTs showed unexpected deviations from expected
  SED (see next slide)
• Bumps of 10-50 of total power over an
  extended wavelength range in 1st order.
• Extremely strong spikes at the long-wavelength
  end of each order.
• This added signal which moreover seems to vary
  for different pixels may prevent proper
  calibration of astronomical spectra
  (line-continuum ratio, e.g.).
• gt Non sequential ray trace model for
  spectrometer section in progress to identify
  causes.

7
Representative Spectrum from ILT
H20 features
2nd order leak 0th order ghost
3rd order leak 0th order ghost
2 pass ghost
200
8
l-l Plot of Ghost Responses in 1st Order
0 t h o r d e r
Ghost 1st order 3pass
Ghost 1st order 2pass
Normal Signal
Ghost 2nd order 2pass
Leak 2nd order
9
l-l Plot of Ghost Responses in 2nd Order
107
Ghost 2nd order 2pass
0 t h o r d e r
Lambda Seen um
Normal Signal
Ghost 3rd order 2pass
Leak 3rd order
65
106
70
Lambda Set um
10
l-l Plot of Ghost Responses in 3rd Order
75
0 t h o r d e r
Lambda Seen um
Ghost 3rd order 2pass
Normal Signal
Ghost 4th order 2pass
Leak 4th order
50
74
53
Lambda Set um
11
FPFPU Optical Performance

• Grating Ghosts Ongoing Work
• Preliminary results of a simplified optical model
  of the Collimator-Grating assembly using
  non-sequential raytracing
• Leakage from next higher order affects the long
  wavelength end of each operating order this is
  normal and can be explained/expected from the
  tail of the filter blocking curves.
• The high temperature of the telescope background
  enhances the effect such that it becomes
  important over a 510mm wavelength range.
• Interesting ray paths to detector array
  identified (see following slides).
• The observed sharp spectral spikes at longest
  wavelength, and some surrounding extended signal
  are caused by 0th order ( direct mirror)
  reflection from the grating, ghosting into the
  beam path via Collimator 1 mirror (see following
  slides).
• A broad, 10-50 excess near 170-180mm is caused
  by a double pass of the grating in 2nd order
  (Ghost-l 105 mm) supported by Collimator 2
  acting as a retro-reflector. The signal is strong
  because of the short wavelength involved. Bounded
  on one side by blocking filter curve, on other
  side by falling off the mirrors.

12
Beam Path of 0th Order Ghost, Simple Baffle
Coll 2
Coll 1
Input Slit
Det.
Baffle
Grating
13
Beam Path of 2nd Order 2pass Leak, Mechanism
14
Beam Path of 2nd Order 2pass Leak, simple baffle
Coll 2
Coll 1
Input Slit
Det.
Black baffle
Grating
15
2 (Preliminary) Baffles in place
Coll 2
Coll 1
Input Slit
Det.
Black baffle
Black baffle
Grating
16
l-l Plot of Ghost Responses in 1st Order
0 t h o r d e r
Ghost 1st order 3pass
Ghost 1st order 2pass
Normal Signal
Ghost 2nd order 2pass
Leak 2nd order
17
l-l Plot of Ghosts, 1st Order, with Baffles
Ghost 1st order 2pass
Normal Signal
Ghost 2nd order 2pass
Leak 2nd order
18
Ghost Responses Magnitude
19
FPFPU Optical Performance

• Grating Ghosts Several identified Possible
  fixes
• Mounting of additional baffles in
  Collimator-Grating area. This means Modification
  of FPU !
• Simple blocking baffles
• Try to find positions for baffle(s) which
  maximise suppression of unwanted beam paths,
  while minimising vignetting for normal beam.
• Trade-off necessary.
• Some losses expected (5-10), mostly at long
  wavelength.
• Can block most, but not all rogue paths.
• Minor modification of FPU, qualification of
  design change could be handled by
  analysis/similarity.
• Needs to happen very soon, since all PFM parts
  have already been manufactured, black-painting
  run imminent.

20
Summary

• Consequences for PFM Programme
• Spectrometer Ghosts need to be fixed, fix ready
  to be implemented soonest Feb. 2005.Manufacturing
  /assembly of PFM collimator section will have to
  be delayed.However, overall PFM M/I schedule is
  probably not affected yet.
• Proper re-alignment and characterisation/calibrati
  on of OGSE and test cryostat, including hardware
  modifications, will take 5 months (no margin),
  worst case 8 months, starting mid Jan.
  2005.Still compatible with PFM schedule.
• Schedule extremely tight - no margin. For the PFM
  there will be a full performance characterisation.