Transcript L.Barsotti

Status of VIRGO
Lisa Barsotti
- University and INFN Pisa –
on behalf of the Virgo Collaboration
 Locking of Full Virgo
ILIAS
CASCINA - January 24th, 2005
VIRGO Optical Scheme
PR
3-km Fabry Perot
cavities in the arms
BS WI
NI
Commissioning Plan
Steps of increasing complexity:
Sept 2003 – Feb 2004
 A SINGLE FABRY-PEROT CAVITY
PR
misaligned
North Cavity
Commissioning Plan
Steps of increasing complexity:
Sept 2003 – Feb 2004
 A SINGLE FABRY-PEROT CAVITY
West Cavity
PR
misaligned
• Check of the performances of the sub-systems
• Check of the control systems in a simple
configuration
Commissioning Plan
Steps of increasing complexity:
Feb 2004 – Dec 2004
 A FABRY-PEROT MICHELSON ITF
West Cavity
“RECOMBINED” MODE
• Intermediate step towards full Virgo
• Start of noise analysis
PR
misaligned
North Cavity
Commissioning Plan
Steps of increasing complexity:
Since Sept 2004
 A POWER RECYCLED MICHELSON ITF
West Cavity
“RECYCLED” MODE
 Final configuration
PR aligned
North Cavity
Commissioning of a Single Fabry-Perot Cavity - I
Lock at the first trial
28th Oct 2003
Transmitted Power
Power
Fluctuations
WE
laser freq noise
&
mirror angular
motion
WI
PR
BS
NI
NE
T=8%
Demodulated
osymmetric beam
Control Scheme
Commissioning of a Single Fabry-Perot Cavity - II
Three Commissioning runs in a single cavity configuration:
 C1 (14-17/11/2003)
- North cavity and OMC locked
IMC control noise
reduced
 C2 (20-23/02/2004)
- C1 + Automatic alignment
- West arm locked
 C3 (23-27/04/2004)
- C2 + Laser freq stabilization
Transmitted Power
Commissioning of a Single Fabry-Perot Cavity – III
 Sensitivity Progress
Three Commissioning runs in a single cavity configuration:
 C1 (14-17/11/2003)
- North cavity and OMC locked
IMC control noise
reduced
 C2 (20-23/02/2004)
- C1 + Automatic alignment
- West arm locked
 C3 (23-27/04/2004)
- C2 + Laser freq stabilization
C1
C2
C3
Commissioning of the Recombined ITF
WE
WI
PR
BS
NI
NE
Commissioning of the Recombined ITF
PBS expected in
Start of some noise
characterization
Sensitivity ~
WI
10 W
P0
PR
recycled mode ~ 500 W
WE
( 500 W)
~1W
PBS
BS
PBS
Sens_Recyc led
500

~ 20
Sens_Recom bined
1
NI
NE
Recombined ITF Optical Scheme
8
West Transmitted
beam
WE
WI
PR
2
Reflected beam
NI
BS
T=8%
NE
7
5 Pick-off beam
1
Asymmetric beam
North Transmitted
beam
Recombined ITF Optical Scheme
8
WE
L2
Simple Michelson
2
 Lengths of the kilometric arms: L1 and L2
 Michelson asymmetric length: l1 – l2
West Cavity
PR
 3 d.o.f. ‘ s to be controlled:
 fields not mixed
WI
l2
BS
NI
l1
T=8%
5
1
North Cavity
NE
L1
7
Recombined ITF – Lock Acquisition
8_demod
 Lock of the two arms indipendently
with the end photodiodes
Corrections sent to NE and WE
West arm
 Lock of the michelson with the
asymmetric port signal
Corrections sent to BS
North arm
7_demod
2_quad
1_demod
Michelson length
Recombined ITF - Linear Locking
 End photodiodes very usuful for lock
acquisition but too noisy
West arm
 Cavities controlled with the reflected
and the asymmetric beams
Michelson
North arm
2_quad
2_phase
Common mode
of the cavities
1_demod
Differential mode
of the cavities
Commissioning Run C4
- June 2004
 Recombined Data Taking Mode
 ITF controlled with the reflected
and the asymmetric beams
 Automatic alignment of the cavities
 Laser frequency stabilized to
cavities common mode
 Cavities common mode locked to
reference cavity
 Output Mode Cleaner locked on the
dark fringe
 Tidal control on both arms
Commissioning Run C4
- June 2004
 5 days of run
 Lock losses understood
 Longest lock ~ 28 h
 h reconstruction on line
Commissioning Run C4:
Noise Characterization
Coupling of IB resonances into the
michelson controller signal due to a
see between
Flaminio’
s talk
mismatch
modulation
frequency
and input mode-cleaner length
Michelson controller signal
C4
After frequency
modulation tuning
 July – August
After C4
 Upgrade of the terminal benches -> Re-tuning and
improvement of the linear automatic alignment
 Suspension full hierarchical control started
 Commissioning of the Recycled ITF started
 Effect of the backscattered light in the IMC
-> attenuator installed between the IMC and the ITF
 Mid September: Re-Start
 October – November:
-> Recombined ITF locked with the full hierarchical
control of the end suspensions
-> ITF locked in recycled mode
Suspension Hierarchical Control
 Locking acquired and maintained acting
at the level of the mirror
marionette
Actuators noise: current status
-5
10
Reference Mass - Mirror Actuators Noise
Filter #7 - Marionetta Actuators Noise
VIRGO Sentivity
-10
reference mass
m/Hz 1/2
10
-15
10
103
-20
10
y
-1
x
z
mirror
z
10
0
10
1
10
Frequency (Hz)
2
10
 Reduce the strength of the mirror
actuators by a few 103 to reach
Virgo design sensitivity
3
10
Suspension Hierarchical Control
TIDAL CONTROL
Corrections sent to the
marionette
DC-0.01 Hz
0.01-8 Hz
8-50 Hz
RE-ALLOCATION OF THE FORCE
Force on the mirror reduced of a factor 20

Switch to low noise coil drivers
Corrections sent to the
mirror
Suspension Hierarchical Control
 SUMMARY
 Single arm locked with the hierarchical control
for the first time in July
-> controllability of the superattenuator demonstrated
 Last main result: hierarchical control of the
recombined ITF in the C4 configuration, with
automatic alignment and frequency servo engaged
 Stable lock -> tested in the last commissioning run
(C5, 2-6 December 2004)
Lock Acquisition of full VIRGO
 Chronology
 Simulations on a lock acquisition technique developed
following the LIGO experience
 Locking trials with this baseline technique
(first half of July)
 Attenuator installed (summer)
 Restart of the locking trials with the baseline technique
(21st September)
 Debugging of the sub-systems
 Establishement of theVariable Finesse lock acquisition
technique (October)
Recycled ITF: Base and Photodiodes
West Transmetted 8
beam
4 lengths to be controlled:
• MICH = ln-lw
WE
• PRCL= lrec+(lN+ lw)/2
LW
• CARM= LN+LW
• DARM= LN-LW
WI
lW
PR
lrec
2
Reflected beam
NI
BS
lN
5
1
Asymmetric beam
NE
LN
7
North Transmetted
beam
Baseline Technique
 Based on the LIGO technique
 Multi–states approach
 Dynamical inversion of the sensing matrix
Experimental Activity:
Lock of Stable States - I
• Sidebands locked in the recycling cavity
Reflected f-demod
signals to control
MICH and PRCL
2_phase
2_quad
STABLE STATE 2
Experimental Activity:
Lock of Stable States – II
• Sidebands locked in the recycling cavity, carrier locked in the FP
Reflected f-demod
signals to control
MICH and PRCL
2_phase
2_quad
STABLE STATE 3
From f-demod to 3f–demod signal
 CARM contamination in the PRCL reconstruction
State 4 Simulated Sensing Matrix
1P

  1e  2

 
1Q

  4e  2

  1e  4
2P

 
2Q

  3e  3
 2 _ 3 f _ P    5e  3

 
 2 _ 3 f _ Q   1e  2

 
5
P

  7e  4

  3e  3
5Q

 
2e  5
6e  6
3e  3
3e  4
1e  1
1e  1
7e  3
3e  4
3e  1 5e  7 

1
1e  7 
1e  5
1  MICH 


1e  6 5e  2  PRCL 
1e  4
1  DARM 



6e  4 1e  1  CARM 

5e  2
1 
8e  3
0.1 
 Frequency Response of the f-demod signal very
sensitive to the ITF losses
PRCL Frequency Response - I
 SIMULATION
Input FP Mirrors Losses 1%o
B2_f_phase
Non - Minimum Phase
PRCL Frequency Response - II
 SIMULATION
Input FP Mirrors Losses 1%o
B2_3f_phase
Minimum Phase
VIRGO Lock Acquisition Scheme
 Good decoupling
MICH / PRCL
 Less CARM contamination
in the PRCL signal
 Almost Diagonal
Sensing Matrix
3f - Demod
Signals
REF BEAM
phase
REF BEAM
quad
5_phase
1_phase
First Locking Trials
BS – PR Corrections
NORTH and WEST Power
Recycling Cavity Power
NE – WE Corrections
Drawbacks of the Baseline Technique:
PR Transfer Function
 PR transfer function
The lock acquisition
technique is “statistical”.
 transients, ringing
MARCH
Compensation of the PR
Resonances: critical, high Q
OCTOBER
Drawbacks of the Baseline Technique:
the CARM contamination

The optical design of the ITF makes the response of the
reflected 2_f signal very depending by the losses
Use of the 2_3f signal in the lock acquisition phase

The CARM contamination is anyway critical :
 use of SSFS is possible only in a steady state regime
A new strategy:
theVariable Finesse
Lock Acquisition
The Variable Finesse
Locking Strategy
“A recycled ITF with a low recycling factor is
similar to a recombined ITF “

End photodiodes

Lock immediately the 4 degrees of freedom of the ITF
on the half/white fringe (low recycling factor)
 lock of PR prevents ringing and transient effects
 lock of the cavities prevents CARM contamination

Bring the interferometer adiabatically from the half
to the dark fringe increasing the recycling factor
The Variable Finesse
Locking Strategy
 Lock immediately the 4 degrees of
freedom of the ITF on the half fringe:
 end photodiodes to acquire the
lock of the long cavities
WE
Low Recycling
Factor
 simple michelson locked
on the half fringe with
the asymmetric DC signal
WI
PR
NI
 3f demodulated reflected
signal to control the
recycling cavity length
NE
BS
1
0.8
0.6
0.4
Half Fringe
0.2
0
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
The Variable Finesse
Locking Strategy
End photodiodes start to see both the cavities:
north_phd   1 0.5 North_cavi ty 

  


 west_phd   0.5 1  West_cavit y 
We can not continue to control the arms
indipendently
The Variable Finesse
Locking Strategy
 Laser frequency stabilization engaged
 One of the end photodiodes used to
control the differential mode of
the cavities
WE
 Laser stabilized on the common
mode of the cavities
Low Recycling
Factor
 PR realigned
WI
PR
NI
 Offset in the mich DC error
signal reduced approaching
the dark fringe
NE
BS
1
0.8
0.6
0.4
Half Fringe
0.2
0
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
The Variable Finesse
Locking Strategy
WEST TRANSM
BEAM
 From the DC to a demod signal
to control the michelson length
LASER
2_3f_ph
1
0.8
0.6
5_ph
0.4
5_q
0.2
0
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
The Variable Finesse
Locking Strategy
 Final Step : To the Dark Fringe
1
0.8
0.6
0.4
0.2
0
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
1
0.5
0
ITF on the operating point
The Variable Finesse
Locking Strategy
RUNNING MODE: Switch to
the main GW signal to control
the DARM mode:
end photodiode very noisy
LASER
2_3f ph
5_ph
5_q
ASY BEAM 1_demod
The Variable Finesse
Locking Strategy
POWER IN THE RECYCLING CAVITY
ITF locked on
the dark fringe
ITF not
locked
Lock Acquisition
“Variable Finesse”
of the recycling cavity
The Variable Finesse
Locking Strategy
POWER IN THE RECYCLING CAVITY
Recycled interferometer (~ 17 W)
TPR=8%
-> Recycling factor ~ 25
Recombined interferometer (~ 60 mW)
The Variable Finesse
Locking Strategy
Recycling Cavity Power
Lock duration limited by the natural
misalignment of the mirrors
Longest Lock: 2h30
( Usually about 30-40 minutes )
Need of the linear
automatic alignment
The Variable Finesse
Locking Strategy
 SUMMARY
 First lock of the recycled ITF on the end of last October
 Stable lock of the recycled interferometer ~ 40-50 minutes
 no linear automatic alignment yet  next step
 Locking procedure tested several times
 lock acquired in few minutes
 New original lock acquisition procedure established, combining
end photodiodes, frequency servo, 3f-demod signal, slightly
misalignement of PR mirror, and lock on the half fringe
 1 day and half of test in the last commissioning run C5
Commissioning Run C5
- December 2004
C5 configurations:
- RECOMBINED ITF as in C4
(automatic alignment, laser frequency stabilization servo, OMC locked)
+ suspension hierarchical control
-> end of the commissioning of the recombined ITF
- RECYCLED ITF (1 day and half)
Commissioning Run C5
- December 2004
RECOMBINED
RECYCLED
Best
VIRGO Sensitivity
Noise hunting: C5 sensitivity
 Sensitivity limited by control noise
Longitudinal locking control signal
Short michelson control Recycling cavity control
Laser freq control
COHERENCES with the GW signal
Local control signal
BS tx local control
Noise hunting: C5 sensitivity
• What about the noise at high frequency ?
?
• Observation: noise level change with time i.e. with alignment
Noise hunting: C5 sensitivity
• 2 minutes of C5 data
Power on dark fringe
Other quadrature
Main ITF output
Averaged noise spectrum
Noise hunting: C5 sensitivity
• Noise variation at high frequency vs alignment
Noise hunting: C5 sensitivity
Next Steps:
• At low frequency (< 100-300 Hz)
- Switch OFF local controls (possible when automatic alignment will be used)
- Use of less noisy error signals to control the ITF (2_3f -> 2_f)
- Use of more complex controller filters
- Reduce sensitivity to IMC length noise ( tune IMC length and Fmod)
• At high frequency (> 100-300 Hz)
- Implement ITF automatic alignment
- Have a better look into noise when alignment is/will be better
Something not undertood yet:
lock losses in C5 data
RECYCLING STORED POWER
Lock acquired,
but not stable
Stable Lock
 Any evident difference in the two periods (analysis in progress)
Something not understood yet:
the “JUMPS”
 “Jumps” in the powers observed with the recycled locked
Recycling Cavity Power
Maximum power
 Jumps very big -> less than half power
 They can unlock the ITF
 More frequent in these last weeks
 Some days it was impossible to work
 Not always present: any evident
difference observed in the ITF status
when jumps appeared with respect to
the quite situation
First idea: jumps connected with the alignment of the ITF
 Some experimental tests: NI misaligned of few urad
Jumps start to appear
when the mirror is
misaligned of 2-3 urad
Aligned
position
Misaligned
positions
Same results obtained
misaligning the PR
mirror
…but jumps are seen also with the “ well aligned” ITF
(maximum stored power observed)
 Some experimental tests: change of the PRCL error signal (2_3f)
demodulation phase with respect to the alignment of the PR
Recycled
Stored Power
Aligned
position
PR - ty
2_3f
demod phase
Aligned position: no jumps for a scan of several tens of degrees
of the demodulation phase
 More PR is misaligned and more the demod phase is critical
ITF locked in a “bad” way?
 Sometimes the ITF works better - higher power, more stable when it is still present an offset in the michelson error signal
(5% out from the dark fringe)
 A constant offset is present in
the out loop reflected signal when
the ITF is locked. The 2_f signal
is planned to be used to control
PRCL (switch 2_3f -> 2_f needed
for noise reduction)
IN LOOP
Offset equivalent to 5 nm
PR displacement
OUT LOOP
ITF locked in a “bad” way?
Stored Power
Switch to 2_f
Refl 2_3f_phase signal
IN LOOP
OUT LOOP
Refl 2_f_phase signal
OUT LOOP
When the switch
2_3f -> 2_f is done the
stored power decreases
of the 50 %
The offset in the 2_f
signal is independent
from the alignment
conditions
IN LOOP
An offset in the
2_3f signal ?
Something not understood yet:
offset in the end signal
Stored Power
Dark Fringe Power MICH error signal
ITF on the
dark fringe
ITF
LOCKED
DARM error signal
Offset
GW signal
IN LOOP
 As soon as the switch from the end to
the GW signal to control DARM is done,
an offset appears on the end signal
 The dark fringe is “ darker” if the ITF
is locked with the GW signal
An offset in the laser frequency servo?
The error signal used to control DARM is
one of the end signals
It sees not only DARM, but also CARM
An offset in the laser frequency servo
error signal could keep the ITF bad locked
in the CARM d.o.f
 the end signal sees the CARM offset, which is transferred to
the DARM d.o.f and which is visible on the dark fringe power
 the GW signal sees only DARM, so it does not see the offset
 could it explain also the offset in the reflected signal?
An offset in the laser frequency servo?
 ITF locked with the GW signal, offset added to the frequency servo error signal
Offset
Stored Power
DARM error signal
Ref 2_f_phase
OUT LOOP
same offset
Conclusions
 1 year of commissioning
28th Oct 2003
First lock of the
north cavity
26th Oct 2004
First lock of the
recycled ITF
Next Steps
 Improvement of the recycling locking robustness
and understanding of jumps and offsets:
- real time simulation under development + dedicated shifts
 Automation
- pre-alignment (in progress) and locking procedures (done)
 Linear automatic alignment of the full ITF
- work started, other 3-4 weeks planned
 Laser frequency stabilization optimization
- preliminary measurements done