Alignment status & plans
Alignment status & plans
2 2
ndndmodulation frequency modulation frequency
Virgo mod. (6.26 MHz) 2nd mod. (8.35 MHz)
2nd modulation frequency:
absolute (demodulated) reference for common end alignment
(not available with one fmod) Previously aligned with ITF reflection Q21_DC
IMC mod. (22.4 MHz)
2nd mod. on the same EOM
=> line forest => difficult Solution: 2nd EOM
=> Much less problems
but modulation must be weak (m=0.01 instead of 0.17) for avoiding perturbations => to be investigated
2 2
ndndf f
modmodlayout layout
6.26 MHz 8.35 MHz
WSR10 alignment layout
2 2
ndndf f
modmod: results : results
8 MHz modulation is working
Low modulation index => Q22 signal noisy => use at LF
Common end loop adapted
Error signal remains Q21_DC
Absolute Q21 position set by shifting until Q22_AC = 0 (Alp loop; stopped in science mode)
Alternative possibility
“Mix” Q22_AC (LF) <=> Q21_DC (HF)
One error signal; filters in sensing matrix Advantage: absolute DC reference is kept
during science mode
Q21_DC Q22_8 MHz
B7_q1 B7_q2
B7
B8_q2
B8
B8_q1
WE
WI
NI NE
BS PR
End mirror beam centering - before End mirror beam centering - before
Before
DSP Drift control
QD error signal
B8_q1_DC B8_q2_DC B7_q1_DC B7_q2_DC
B7_q1 B7_q2
B7
B8_q2
B8
B8_q1
WE
WI
NI NE
BS PR
End mirror beam centering - now End mirror beam centering - now
Now
Alp Drift control
Alp error signal
7...8 Hz line on NE/WE tx/ty Locking correction in z
=> control input mirrors Noise reduction
but: upconversion 10-15 Hz ?!
10 Hz
before after
(high ISYS noise!)
Alignment noise in dark fringe Alignment noise in dark fringe
Differential end mirror tx mode almost limiting
=> more power on Q1p diode + whitening filter
=> alternative: improve corrector cut-off (but: need all margin for high LF gain)
G. Vajente Automated alignment noise projections (10 min.)
Coherence
Recent Gc-Ali upgrades Recent Gc-Ali upgrades
Sensing – Filtering – Driving available
Driving switch needed during thermal transient (switch NE-WE => Common-Differential)
Noise injection
For measurement of open loop transfer functions
Filtering gain change on-fly
For migrating all gain changes from DSP to Gc
Sensing Filtering DSP
Global control
Driving
Noise
Alignment OL transfer function measurements Alignment OL transfer function measurements
Differential end tx
Excellent fit
with Matlab model
Gain adjustment
=> matrix calibration
Differential end ty
?
Less excellent fitBut still good in the important region (a few Hz)
Other improvements Other improvements
Thermal transient robustness
Improved by switching on 1. differential 2. common end alignment Driving matrix NE/WE => NE+WE/NE-WE after thermal transient PR alignment acts on reference mass
Local control on marionette remains on
High gain filters (end mirror control)
Further tuning needed
Sometimes more gain than needed
=> Rather invest in HF cut-off
Histogram
Diff end tx error signal
Plans up to scientific run I Plans up to scientific run I
Improve alignment stability (filter work)
Sometimes oscillations during thermal transient Check stability in bad weather conditions
Reduce alignment noise
10-50 Hz: alignment noise starts becoming important
Error signal improvement (B1p: more light + preshaping)
Filter tuning (LF gain / HF cutoff trade-off)
Continue transfer function measurements
Line injections for better coherence
Understand model deviations
Measure resonance frequencies: rad. pressure effects?
Update sensing matrix
Continue alignment globalization
More logical system: all gain adjustments in Gc, …
Plans up to scientific run II Plans up to scientific run II
Common mode error signal improvement
Create composite error signal (mix Q22_ACLF + Q21_DCHF)
Observe beam centering loops
Improve if needed (up-conversion, better centering, …)
Centering of input mirrors?
TBC
To be done if needed To be done if needed
Switch BS/ISYS <-> NI/WI control
LA on NI/WI mirrors
BS/ISYS steer beams in arms
Globalize other LA degrees of freedom
PR, BS
PR alignment
Keep LC on marionette on?
Optimize IMC alignment filters
Reduce LF beam jitter (higher gain)
Increase 8 MHz modulation index
First understand where locking signal perturbation comes from
End
Alignment configurations Alignment configurations
C7
14/09 – 19/09/2005tx IB PR BS NI NE WI WE ty IB PR BS NI NE WI WE
10 d.o.f. fast alignment
WSR1
08/09 – 11/09/2006tx BM PR BS NI NE WI WE ty BM PR BS NI NE WI WE
6 d.o.f. fast alignment; input beam and BS control
WSR10
09/03 – 12/03/2007tx BM PR BS NI NE WI WE ty BM PR BS NI NE WI WE
7 d.o.f. fast alignment; input beam and BS control
XX Linear alignment XX LA (ref. mass) XX Drift control XX Local control XX DC error signal XX DC + AC err.sig.
WSR10 sensing matrix WSR10 sensing matrix
BMS PR BS NI NE WI WE ThetaX
0.5 B1p_q1_ACq 4 B2_q1_ACp 2 B5_q1_ACq -0.5 B8_q1_ACp -1.5 B2_q1_DC 1 B7_q1_DC -1.5 B7_q2_DC 1 B8_q1_DC -0.8 B8_q2_DC BMS PR BS NI NE WI WE ThetaY
-0.6 B1p_q1_ACq 4 B2_q1_ACq 2 B5_q1_ACq 0.4 B8_q1_ACq 4 B2_q1_DC 1 B7_q1_DC -1 B7_q2_DC 1 B8_q1_DC -0.7 B8_q2_DC
17 Commissioning meeting, Cascina 02.04.2007 H. Heitmann
WSR10 alignment control overview WSR10 alignment control overview
Bea m Dio
de De mo d.
Loop
Arm diff. B1p Q Fast
Arm commo n
B2 1
2 for centering
- Fast
PR B5 1 Q Fast
BS B8 1 P
(tx) (ty) Q
Fast (tx) Drift (ty)
NI B7 1+2 - Drift
WI B8 1+2 - Drift
BMS B2 1 P
(tx) Q
Drift
WSR10 alignment filtering WSR10 alignment filtering
Sensing Filtering DSP
Sensing Filtering 1
DC Filtering 2
Initial Filtering 3
Boost DSP
PR tx
--- --- --- --- Ref.mass
+ mario LC
PR ty
--- --- --- --- Ref.mass
+ mario LC
End tx
--- --- INITIAL BOOST ---
End ty
--- --- INITIAL BOOST ---
Global control
Driving
WSR10 alignment driving WSR10 alignment driving
Sensing Filtering DSP
Driving 1
Initial_Driving Driving 2 Driving
Common
End => NE => NE+WE
Differ.
End => WE => NE-WE
Global control
Driving