1. Design of the NFIRAOS PWFS module
Jean-Pierre Véran
National Research Council Canada – Herzberg, Victoria, BC
Presentation at the "Wave Front Sensing in the ELT era" workshop
Padova - Oct 2, 2017
TMT.AOS.PRE.17.XXX.REL01

2. NFIRAOS Simplified Control Diagram in LGS Mode
WFS/DM telemetry (PSF recon)
Telescope offloads
TWFS is an NGS WFS that measures radial modes at low speed to track changes in the profile of the sodium layer. It is nominally a ~12x12 WFS. Also track drifts in NCPA compensation?
Reconstruction parameters
RTC Param. Gen.
RTC
Star selection mechanism
LGS Offsets
Truth WFS
LGS Pixels
DM/WFS Statistics
Actuator commands
OAP
LGS Trombone
LGS BS
6 LGS WFS
6 LGS WFS
OAP
OIWFS Pixels
On Inst. WFS(s)
6 LGS
Inst. Fold
Inst. Fold
NGS(s)
Sci. Obj.
OAP
DM11
OAP
OAP
DM0 + TTS
SCI BS
OAP
To Sci. Instrument

3. NFIRAOS Simplified Control Diagram in NGS Mode
WFS/DM telemetry (PSF recon)
Telescope offloads
NGS WFS is a high-order WFS. This mode is typically used for high-contrast imaging, or when the lasers are not available.
Reconstruction parameters
RTC Param. Gen.
RTC
Star selection mechanism
NGS WFS pixels
NGS WFS
DM/WFS Statistics
Actuator commands
OAP
LGS BS
OAP
OIWFS Pixels
On Inst. WFS(s)
Inst. Fold
Inst. Fold
NGS(s)
Sci. Obj.
OAP
DM11
OAP
OAP
DM0 + TTS
SCI BS
OAP
To Sci. Instrument

4. Context within NFIRAOS
Input window
SS- Source simulator
Rotating pupil mask
OAP3
DM11
OAP1
OAP4-S
OAP4-L
OAP2
VNW- Visible natural wavefront sensor
Instrument selection mirror
Beamsplitters
Tip-tilt stage & DM0
LGS- Laser wavefront sensors & trombone
OAP4-V

5. PWFS top-level parameters
CCID-96
256x256 pixels frame transfer CCD with 64 output amplifiers
<3e- noise at 800 Hz frame rate
On-chip binning
Each pupil image has 96 pixels in diameter
Oversamples 60x60 Fried geometry
Binning to 48, 24, 12 and 6 possible
PWFS module design sub-contracted to ABB and INO (Quebec City, Canada)
PDR: October 2017
FDR: March 2018
NFIRAOS FDR: June 2018
TMT.AOS.PRE DRF02

6. Current CAD model

7. Double Pyramid
Canadian and Italian double pyramids have same optical effects
Italian pyramid already prototyped for GMT
Canadian pyramid has better thermal properties
NFIRAOS operates at -30C so ΔT=50C !

8. Can we use a single pyramid?
PWFS operates between 610nm and 785nm
Chromatic blurring with single pyramid ~0.9 pixel
Assumes BK7
Current budget for quality of pupil image is ~0.9 pixel FWHM
Dominated by charge diffusion (0.7 pixel FWHM)
Double roof-prism set-up at NRC (Van Kooten 2016)

9. Relaxation of quality of pyramid edges and roof
Original specification:
Edge and roof < 5um
We have simulated the effect of having a 22um edge / roof
f/45 beam on pyramid
Incremental wavefront error is only 16nm RMS
Specifying edge / roof < 20um makes pyramid much easier to manufacture

10. Relaxation angle tolorances on pyramid (Lardiere AO4ELT4)
Wang, L, AO4ELT4 2017

11. -30C operated mechanisms
Rotary stage (ADC)
customized version of PI stage DT-80
Slow tip, tilt and focus stage (pupil pointing)
customized version of PI stage PLS-85 (linear stage)
PI S330.8 (tip-tilt)
Required to achieve repeatability spec (<5 urad)
Fast Steering Mirror
X-Y stages
customized version of PI stage LS-180
Main challenge are:
Travel: 300mm
Bi-directional repeatability: 2um
Differential accuracy: 4um over 60mm (2mas over 30” on sky for AO guider offsetting) and 20um over whole travel range

12. FSM trade-off Traditional choice for PWFS modulation: PI S330
However, this module has several downsides:
Heat dissipation (~4W when modulating at 800Hz)
Would require active cooling
Lifetime (109 cycles or about 4 months at 1000 hours/year)
PI has proposed a custom mono-crystal stage
No heat dissipation
Improved lifetime (to be quantified)
Very linear (no capacitive sensor, no feedback electronics)
But limited stroke: ≤±0.3mrad or 10λ/D (at all frequencies)

13. Testing of mechanisms
PI has been the most responsive / interested vendor
PI mechanisms are rated to -20C
PI will test the mechanisms at room temperature and at -30C
No optical test
We are in the process of procuring all mechanisms now to conduct our own optical tests at -30C

14. TMT Pupil Mask
NFIRAOS on Nasmith platform and does not have a derotator
Telescope pupil will rotate on DMs and WFSs
For calibration and testing, a rotating pupil mask could be implemented at an image of the telescope pupil in the high-altitude DM (DM11) relay
Could simulate ragged pupil edge and spiders
Would have to retract during operations
If no pupil mask, pupil is defined by oversized circular mask on DM0

15. Interaction matrix calibration
For LGS WFS, we are planning to use a hybrid approach
Experimental interaction matrix to derive low-order mis-registration parameters.
Then the actual interaction matrix is computed analytically, including details of TMT pupils.
Interaction matrix and reconstructor are updated as the TMT pupil rotates (every 0.5 degree)
Can this approach work for PWFS, or do we need to use an experimental interaction matrix
In order to properly capture diffraction effects
In the coming months, we are planning to compare experimental and synthetic interaction matrices on our PWFS bench