Transcript Micromirror Arrays for Adaptive Optics

Micro-Deformable Mirrors for Adaptive Optics
Emily Carr
Department of Electrical and Computer Engineering
University of California, Davis 95616
Lawrence Livermore National Laboratory
Livermore, CA 94550
Lawrence Livermore
Scot Olivier, Peter Krulevitch
Lawrence Livermore National Laboratory
Livermore, CA 94550
Olav Solgaard
E. L. Ginzton Laboratory
Stanford University, Stanford, CA 94305-4085
CfAO Video Conference - March 29, 2001
Outline
• Why use micromirrors in an AO system?
• Current Research
– Boston University’s Deformable Mirror
• Characterization of Mirror Surface
• Voltage vs. Displacement Results
– AO Testbed Results with BU Mirror
Lawrence Livermore
• Far-field spot
– Airforce/Cowan/Sandia Mirror
• Characterization of Mirror Surface
• Voltage vs. Displacement Results
• Future Research
– New mirror design
CfAO Video Conference - March 29, 2001
Conventional
Deformable
Mirror
Conventional
Deformable
Mirror
Lawrence Livermore
http://www2.keck.hawaii.edu:3636/realpublic/inst/ao/about/slides/dmirror.html
CfAO Video Conference - March 29, 2001
Boston Micromachines Corporation Mirror
Lawrence Livermore
http://www.bostonmicromachines.com
CfAO Video Conference - March 29, 2001
Optical Characterization of BU Micromirror
3.3mm
Lawrence Livermore
300mm
3.3mm
• 2 mm deflection achieved
with 240 V applied to each
individual actuator.
Electrostatically
actuated
diaphragm
Attachment
post
Continuous mirror
CfAO Video Conference - March 29, 2001
Membrane
mirror
Static Mirror Surface - No Voltage Applied
Lawrence Livermore
•Mosaic of 4 images taken with a Zygo
white light interferometer.
Line Scan ~ 8 actuators
rms: 35.84 nm
CfAO Video Conference - March 29, 2001
Influence Function of BU Mirror
Lawrence Livermore
• 100 V applied to one central actuator on
the BU mirror
CfAO Video Conference - March 29, 2001
Lawrence Livermore
•Pink data points - measured data
•Blue data points - curve fit: y=0.02028 x2.2105
CfAO Video Conference - March 29, 2001
Flattening the Good Mirror
Lawrence Livermore
• Zero Voltage applied to all the actuators
• Best Flat with an 8x10 array of actuators
Good Mirror - W25C27
CfAO Video Conference - March 29, 2001
Flattening different areas of the mirror
Lawrence Livermore
Flattest 8x8 elements - central actuators
Flattest 10x10 elements
This is the flat being used in the AO
testbed.
CfAO Video Conference - March 29, 2001
Zero Volts on MEMS – Far-field Spot in AO Testbed
Lawrence Livermore
Zero Volts on all the actuators
• This is the far-field spot we obtain when we put the mirror on the left in the AO
testbed.
CfAO Video Conference - March 29, 2001
Far-Field Spot with Flat Voltages Applied
Lawrence Livermore
Flattest 10x10 array
• Far-field spot with the mirror at left in the AO testbed.
CfAO Video Conference - March 29, 2001
MEMS Mirror Far Field
Lawrence Livermore
Row and column slices: red is measured, black
is perfect beam. Strehl~0.43
CfAO Video Conference - March 29, 2001
Reconstructed Waveforms from DM
Bad Actuator
Lawrence Livermore
Big Checkerboard – 125V applied
To sets of 4 actuators.
Small Checkerboard – 125V
Applied to every other actuator
• Reconstructed waveforms from the Shack-Hartman wavefront
sensor.
CfAO Video Conference - March 29, 2001
Coupled
Actuators
Airforce/Cowan/Sandia 256 Actuator Segmented Mirror
Lawrence Livermore
•256 Active Actuators
•Segmented Mirror
•Maximum voltage needed 20V
•0.65 microns of stroke
CfAO Video Conference - March 29, 2001
Airforce/Cowan/Sandia Mirror Voltage vs. Displacement Curve
Actuator Structures
Lawrence Livermore
200 mm
•Light blue data points - measured data
•Red data points - curve fit: y=0.39789 x2.3676
CfAO Video Conference - March 29, 2001
“New” Mirror Design
Lawrence Livermore
CfAO Video Conference - March 29, 2001
Conclusion
• The Deformable Mirror (DM) is a crucial part of an AO system.
• MEMS based DM’s are cheaper, faster, and have the potential
to have more stroke then conventional DM’s.
• Boston Micromachines Corporation has a commercially
available continuous face-sheet MEMS DM with 2mm of stroke,
7kHz resonance frequency, and an rms surface error of 30nm.
240 V needed to get maximum deflection.
• The Airforce mirror designed by William Cowan has a better
surface quality, but only 0.65mm of stroke, and it is segmented.
20 V needed to get maximum deflection.
• Will continue to work on new high-stroke design based on the
Justin Mansell’s deformable mirror architecture.
Lawrence Livermore
CfAO Video Conference - March 29, 2001