Some Advanced Virgo Developments l l l Laser

Some Advanced Virgo Developments l l l Laser Benches suspensions DAQ/Controls B. Mours (with slides from Alain Brillet & Jo Van de Brand) LCGT-Virgo meeting November 5, 2010

Pre Stabilized Laser for Ad. V l Reminder : u Main requests: Virgo quality (noise, reliability), 200 W power u present reference solution : LIGO-like DPSS l Preparation of a change-request for a fiber laser solution u Cheaper, smaller, better l Remaining choices u Thin fiber (NUFERN) u Rod fiber (EOLITE) » Amplifier or Injected laser 2010/10/07

LZH-LIGO Laser: Nd-SSL 2010/10/07

Pros and Cons l l Pro: u. Extrapolation of present lasers u. Developped in 2009 for LIGO, known performances Con: u. Large and expensive u. Inefficient Thermal management problems » 32 x 45 W diodes > 1 k. W heat loss u. Maintenance 2010/10/07

Y doped fibre: thin fibre design 6 x 50 W Pump diodes Master laser Nd-YAG EOM > 30 m. W Green box Noise performances and beam geometry OK Power limited to 100 W (Stimulated Brillouin Scattering) coherent addition Towards higher power : SBS threshold x 6 Reliability still TBD 2010/10/07 5

Examples of deployed systems Custom rack mounted 120 W fiber lasers (32 units deployed with one customer) Example of 2 x 1 k. W fiber lasers deployed with one customer Eight x 100 W fiber lasers 16 x 150 W PM Amplifiers deployed with one customer 6 17/11/09 Nufern Proprietary 6 www. nufern. com

RIN/Frequency/Beam shape measurements Master laser EOM Beam Dump NUFERN Amplifier ~ 100 W ~300 m. W Power control Camera CCD Frequency control 17/11/09 7

Relative Intensity Noise control Blue curve: -Noise floor at -165 d. B/√Hz between 20 Hz and 1 k. Hz 10 -6 - Noise floor close to shot noise 10 -8 Unity Gain Frequency (UGF): 30 k. Hz 10 -9 The servo loop UGF can be improved 17/11/09 8

Frequency noise UGF : ~200 k. Hz Limitation: a 500 ns delay coming from the fibre amplifier. Solution: a phase corrector 10 -2 Noise floor at 6. 10 -3 Hz /√Hz, not limited by the power amplifier 17/11/09 9

Beam shape and matching measurement Evaluation of beam geometry : contrast in reflection from a high quality symmetric Fabry-Perot cavity Absence of Amplified Spontaneous Emission Contrast ~ 95% Absence of residual pump power Result independent of the output power, between 3 W and 100 W. Blue curve: power reflected by the cavity as a function of the laser frequency 17/11/09 10

Reliability 30% power loss in 500 hours : too much, probably due to the transportation accident, being checked by NUFERN …. 2010/10/07 11

Double pass rod amplifier Dichroic MO 200 W Rod fiber Mirror Core diamater up to 80µm 1 x 300 W diode Industrial design: lithography… Proved reliability 150 lasers/year (50 -300 W) 2010/10/07 12

Tentative planning for ROD laser l Eolite produces a (140 W) test laser for Virgo, u To be tested as single and double pass amp, and as injected laser u Delivery 12/2010 l l l Spring 2011: choice of the final 200 W solution Fall 2011: delivery of the first 200 W laser End 2011: retrofit of the prototype as a 200 W laser. 30 + 100 k€ 350 k€ (ARTEMIS-EOLITE program)

Benches seismic isolation Developments for Virgo+ and Ad. Virgo 5 Nov 2010

External Injection Bench suspension § Needed to reduce § Beam jitter noise § Diffused light noise § Seismic attenuation filters § Inverted pendula § GAS filters F. Mul, Nikhef § Installation in 2011?

Geometric anti-spring development § GAS filter development § Features § Simplicity of design, reduced number of components (< 50 kg) § Bakeable at 200 o. C § No body or cross-bar resonances § Improved thermal stability § Fast construction, assembly and tuning, wide range of loads § Active control possible (e. g. EMAS control) due to integrated LVDT and voice coil, vertical ID § Excellent attenuation performance (80 d. B above 10 Hz) § Applications § SAS systems § Suspension chains: with Riccardo De. Salvo and Alessandro Bertolini § Width profile New blade profile (half model) top surface, smax = 1. 59 GPa

GAS as chain element M. Doets Nikhef Could be tested in collaboration with LCGT Bi-metal plate GAS spring blade LVDT – voice coil Magic wand

Possible GAS filters use: MSAT for recycling cavities § Could solve space issue for recycling cavities § Folded cavities with reduced requirements § Still a proposal under evaluation among other solutions

IB and DB suspension § Attenuation § Horizontal: IP and wires § 127 d. B (1 fm/rt. Hz) at 10 Hz § Vertical: GAS springs § 3 sets of 2 blades § Vertical – horizontal coupling § 1% yields 1 pm/rt. Hz residual vertical § Easily handled by GAS § No wands needed § Criteria and procedure needed § GAS systems § Short IP legs § Input § Optical paths § Control issues A. Bertolini, AEI F. Mul, Nikhef

New DAQ electronic for Virgo+ l New timing system u Distribute GPS time to all electronic modules » Optical fiber distribution (between racks) u Local clock generation » Low jitter u VSR 2 stability: » 6 Months of data » all samples within +- 0. 2µs » No sensitive to reboot l New control system based on Linux PCs u More CPU cycles, Flexibility u Connection with other PC/ADC/DAC using optical fibers l New ADC boards Remark: DAQ software used also for LIGO-Virgo online analysis

ADC 7674 board l l 16 differential channels 800 k. Hz sampling rate u Enable digital demodulation like for the OMC, or LVDTs l 18 bits; Noise floor: u 100 n. V/sqrt(Hz) without whitening u 30 n. V/sqrt(Hz) with on board whitening l l 4 DSPs for filtering and down sampling Absolute GPS timing Optical fiber output to PCs Cost: 2 k. Euros/boards