Title: Compact Fiber Laser for 589 nm Laser Guide Star Generation
1Compact Fiber Laser for 589 nm Laser Guide Star
Generation
- Jay W. Dawson, Deanna M. Pennington, A. Brown
-
- Lawrence Livermore National Laboratory
- 2007 CFAO Spring Retreat
- March 26, 2007
Work done under the auspices of the U.S.
Department of Energy by the University of
California Lawrence Livermore National Laboratory
under Contract W-7405-ENG-48. This work has
been supported by the National Science Foundation
Science and Technology Center for Adaptive
Optics, managed by the University of California
at Santa Cruz under cooperative agreement No.
AST-9876783. This work has been supported by
the National Science Foundation Adaptive Optics
Development Program, managed by the Association
for Research in Astronomy.
2We are developing CW and pulsed fiber laser
technologies for next generation LGSs
NIF fiber amplifier chassis
- Fiber lasers provide an elegant solution
- Compact, rack mounted, fiber delivery
- Efficient operation (limited electrical power
and cooling) - Turnkey operation
- Reliable (high MTBF)
- Robust
- Safe (all solid-state, no chemicals)
3ELT designs include several laser guide stars
- Baseline TMT architecture includes three 50 W CW
lasers designed to produce 9 LGS - Preferred upgrade architecture includes six 50 W
pulsed lasers with dynamic refocusing - Power requirements for upgrades can be reduced
by - Implementing AO on the LGS uplink to produce a
smaller focus - This will be tested on the Nickel Telescope at
Lick Observatory in 2007 - Mitigating spot elongation effects
- AODP funded development of pulsed fiber laser and
custom CCD capable of tracking laser pulse
through Na layer - Pyramid wavefront sensing
4Spot elongation can be mitigated by tracking
laser pulses in the Na layer
- Key laser times
- Time to Na layer 300µs (h/90km)/cosz
- Round trip 600µs (h/90km)/cosz
- Time through Na 33µs (t/10km)/cosz
- Pulse separation for single pulse in Na
- layer 66µs (t/10km)/cosz
- Max pulse frequency
- 15Khz (10km/t)cosz
- Pulse duration integration time
- lt 8.7µs (blur/0.5 arcsec)/(s/15m)/cos3z
5Arbitrary pulse format can be achieved by adding
modulators to the seed lasers
Ppeak Pavg trep/to , Duty cycle ton/trep ,
Repetition rate 1/ trep
- Rep rate lt 2 kHz is not optimal for CW pumping
NDFA - - Nd3 upper state lifetime 470 ms
- For efficiency, repetition rate should be gt 2
kHz with gt1 duty cycle - Consistent with proposed ELT pulse format (6 ms,
16.7 KHz)
6Gain competition from the 1088 nm 4-level line
make the 938 nm Nd3 laser challenging
- 938 nm operation requires an Al-free glass
composition - Al or P pull the emission wavelength shorter to
915 nm - Significant limitation on the Nd ion
concentration (lt10 dB/m _at_ 808 nm) because of
concentration quenching, forcing a long laser
amplifier - 938 nm operation is hampered by ground-state
absorption at 938 nm and parasitic emission at
1088 nm
7Reducing core/clad ratio enables room temperature
938nm operation with manageable 1088nm gain
1088 nm peak has 40 dB lower gain than 938 nm
peak
8We generated gt15 W, CW at 938 nm with a 100 W,
808 nm pump, with narrow linewidth
938 nm output power at various spots in the
system
M2 lt 1.01 Polarization 101
9Initial 938 nm pulsed experiments yielded gt10 W
avg. power with 500 MHz bandwidth
- Pulsed at 100 kHz with 20 duty cycle
- gt95 of optical power was in 938 nm signal line
- No sign of SBS with 500 MHz signal line width
- Square pulse distortion will be implemented to
scale to 10 W in the 10 kHz repetition rate regime
101583 nm fiber laser is constructed from
commercially available components
PM Lithium Niobate phase modulator Polarization
sensitive, 20 dB extinction ratio
Koheras SM 1 mW/1583 nm
Single mode fiber pre-amplifier
Lithium niobate phase modulator
EAR-15k-1583-LP-SF IPG Fiber amplifier 14 W
14 W CW 10 W pulsed
Isolator
111583 nm system produces 14 W in CW mode with gt98
of the power in the signal
Power vs. Pump Current
Output spectrum at full power
121583 nm laser produced gt 15 W at 20 duty cycle
at 100 kHz
However, the lithium niobate amplitude modulator
is leaking significant CW light. So the peak
power is only 1/3 of expected value
The cause of the CW leakage was poor polarization
control from the oscillator
13The front end of the 1583nm laser has now been
improved to be all PM (no leakage issues)
- Square pulse distortion increases the peak power,
driving SBS - - Add more bandwidth to suppress (gt 400 MHz)
- By programming the modulator drive signal, we can
pre-compensate for square pulse distortion
- IPG 15W amplifier unit failed twice. It has been
repaired and the root cause of the prior failures
is believed to be a flakey key switch
14Preliminary SFG in PPKTP yielded 2.7 W of 589 nm
light for CW format
- 2.7 W _at_ 589 nm with 6 W _at_ 1583 nm and 11 W _at_ 938
nm - Power scaling in PPKTP limited by damage and
available 1583 nm
- Switch to PPSLT which is less susceptible to
damage effects - Pulse laser to achieve higher conversion
efficiency
Na cell, 589nm
15We generated 3.8 W at 589 nm in 3 cm of PPSLT at
100 kHz and 10 duty cycle
- 1583 nm laser had significant CW leakage, so a
large percentage of its power was not
contributing to frequency conversion - PPSLT showed no signs of damage at these power
levels
16Where are we? Where are we going next?
- 3.8W is less than the 10W original target
- However, it does demonstrate the basic
feasibility of the concept - It was hoped we could generate the full 10W in
the breadboard phase, but enormous efforts were
being undertaken to do this with little practical
payoff in the long run - To this end, breadboard level experiments have
been discontinued - Our recent internal experience with our short
pulse laser systems indicate that packaging leads
to much better performance and simplified
ease-of-use - Our focus is now on engineering the system for
packaging and turn-key operation and installation
on the Nickel telescope at Lick in late 2007 or
early 2008 - The 1583nm laser sub-system is essentially in
this state now - The only thing the 1583nm subsystem needs to be
ready to go to Lick is some software control and
a better driver for the phase modulator - The 938nm laser system a bigger, but solvable
challenge (see next slides) - It appears PPSLT will work for this laser
- We now have AR coated PPSLT crystals and need
only to design the sum-frequency mixing and
diagnostics breadboard which should be
straightforward - We will be working with the CFAO and Lick teams
over the summer to ensure the final packaging
will integrate well at the telescope
17938nm laser engineering and packaging
- A new fiber coupled 938nm master oscillator has
been ordered and received, along with a fiber
pigtailed AOM and LiNbO3 phase modulator - We have also designed and ordered a PM 938nm
fiber, an all fiber pump signal combiner and 7
LIMO pump diodes appropriate for driving the
938nm laser to full operating power of 15W - This task was tricky and involved simultaneous
negotiation with several vendors in order to
ensure the custom parts will be all created in a
way that they operate together - We have also had some purchasing and
institutional bureaucracy issues that have
created some schedule delays - The project is currently dormant in order to
conserve funds while we wait for these components
to arrive - We have identified two people internally with
appropriate skills in software control, optics
and mechanical engineering who can work on this
problem starting in late April when we expect the
above components to arrive and we anticipate
generating 589nm light with the new system by the
early fall
18938nm system block diagram
19Internal schematic of 938nm amplifier (up to 4
25W LIMO diodes can be employed)
20Our laser will be installed on the Nickel
Telescope at Lick Observatory in late 2007 or
early 2008 for a visible light AO demonstration
- Need 10x more laser fluence per spot to do
visible light AO - A factor of 100x 200x gain is available from
- Uplink AO correction
- Pyramid sensing
- 4x more beacons needed in the tomography
constellation - Other gains?
- CW vs micropulse format
- Tracking beam in Na layer
21Summary
- We are developing CW and pulsed 589 nm laser
systems for ELTs - We have achieved gt 15 W at 938 nm in a Nd3 based
amplifier system - We demonstrated 11 W in a pulsed format with 20
duty cycle at 938 nm - 10 is achievable with an extra amplifier stage
for additional gain - We constructed a 14 W, 1583 nm laser system from
commercial components - Pre-compensation for square pulse distortion is
being implemented to achieve 10 duty cycle - We have achieved 3.8 W at 589 nm with a 10 duty
cycle via sum frequency mixing in PPSLT with no
signs of optical damage in the crystal - Final pulse format of 3 ms at 15 kHz will enable
tracking of pulses through the Na layer to
mitigate spot elongation - System scheduled for installation on the Nickel
telescope at Lick Observatory at end of 2007 for
a visible AO demonstration with AO corrected
laser uplink