Laser Frequency Stabilization

Laser Frequency Stabilization

François BONDU

CNRS – ARTEMIS, Nice

• How it works (Design)

• Make it work (experimental work)

• Noise performances

• Next activities

Complete Design in 2002 - early 2003

Designed to meet the requirements in the 10 Hz – 10 kHz band, Starting from the requirements.

Design written, in Virgo Notes:

Requirements for second stage VIR-NOT-OCA-1390-227 Advanced digital filters VIR-NOT-OCA-1390-233 Multiple servo loop systems VIR-NOT-OCA-1390-234 Description of the loops VIR-NOT-OCA-1390-247 THIS design functionalities successfully tried on the north arm this month

HOW IT WORKS - DESIGN

HOW IT WORKS - DESIGN

MC

laser

C

C

C

C

OVERVIEW

ITF Common mode

- or north arm -

HOW IT WORKS - DESIGN

MC

laser

C

C

C

2

stage

C

Pr_B1(p)_ACp Sc_IB_SSFS_Corr

Bs_IMC_DR_AC1

Second stage of frequency stabilization

HOW IT WORKS - DESIGN

MC

laser

C

C

C

Pre-stabilization

C

Bs_IMC_DR_AC1

First stage of frequency stabilization

(pre-stabilization)

HOW IT WORKS - DESIGN

MC

laser

C

C

C

IMC

C

Sc_IB_SSFS_Corr

Sc_MC_zCorrR

Bs_IMC_DR_AC1

IMC locking

HOW IT WORKS - DESIGN

MC

laser

C

C

C

C

Sc_NE_zGc Sc_IB_zErrGC

ITF Common mode DC control

HOW IT WORKS - DESIGN MODEL

« frequcncy noise » inputs

Stabilized laser frequency

Free running laser noise

IB or MC motion

Reference cavity length noise

HOW IT WORKS - DESIGN SIMULINK MODEL

With radiation pressure

HOW IT WORKS - DESIGN Note on alternative topology (‘topo #1’)

MC

laser

C

C

C

C

Sc_NE_zGc

Sc_IB_zErrGC

ITF DC control necessary

reference cavity can not change its frequency easily

=> ‘bad’ configuration (two loops sharing a same error signal)

=> second stage can not have DC gain

 Is not demonstrated to meet the requirements See talks of april 2002, nov 2003

 Would have been a deceiving solution, according to today knowledge HOW IT WORKS - DESIGN

Note on alternative topology (‘topo #1’)

- « 170 Hz » issue (december 2003) - analysis work

- lock acquisition

- transient during the lock acquisition / unlock of IMC - transient for lock with B1 / algo in Gc

- resonances in the MC suspension

Experimental work

Oscillation – saturation at about 170 Hz

zGc not « hold »

Experimental work (1)

Leak-through due to phase modulation via IMC motion (cf. H. Trinquet)

VIR-NOT-OCA-1390-178

Cf. also gouaty_09Nov03_sens.ppt

Stability of second stage:

crossing of these two TFs

Experimental work (1)

SUSPICION ^:

Experimental work (1)

Leak-through

• Made the calculations taking into account this effect:

=> loop stable

• The problem was a loop oscillation

too much gain, due to the « horns » in the laser frequency noise

Experimental work (1)

Experimental work (2)

Experimental work (2)

SSFS_763560321_1.gif SSFS_763560321_2.gif SSFS_763561763_1.gif SSFS_763561763_2.gif SSFS_763561763_3.gif SSFS_763561763_4.gif SSFS_763566197_1.gif SSFS_763566896_1.gif SSFS_763575387_1.gif SSFS_763575387_2.gif SSFS_763575387_3.gif SSFS_763633937_1.gif SSFS_763633937_2.gif SSFS_763633937_3.gif SSFS_763633937_4.gif SSFS_763633937_5.gif SSFS_763643300_1.gif SSFS_763643300_2.gif SSFS_763643300_3.gif SSFS_763643301_1.gif SSFS_763647965.gif SSFS_763649282.gif SSFS_763649282_1.gif SSFS_763650999.gif SSFS_763726087.gif SSFS_763726087_2.gif SSFS_763726087_3.gif SSFS_763730743_1.gif SSFS_763730743_2.gif SSFS_763730743_3.gif SSFS_764023082_1.gif SSFS_764023082_2.gif SSFS_764023082_3.gif SSFS_764023082_4.gif SSFS_764023082_5.gif SSFS_764025420_1.gif SSFS_764025420_2.gif SSFS_764107509_1.gif SSFS_764107509_2.gif SSFS_764109486.gif SSFS_764109945.gif SSFS_764109945_2.gif SSFS_764110790.gif SSFS_764112941_1.gif SSFS_764114400_1.gif

SSFS_764114409.gif SSFS_764184989.gif SSFS_764191160.gif SSFS_764191160_2.gif SSFS_764191160_3.gif SSFS_764191160_4.gif SSFS_764198517_1.gif SSFS_764237811.gif SSFS_764237811_2.gif SSFS_764240101.gif SSFS_764240101_2.gif SSFS_764242950_1.gif SSFS_764242950_2.gif SSFS_764255788.gif SSFS_764255788_3.gif SSFS_764257611_1.gif SSFS_764257611_2.gif SSFS_764257611_3.gif SSFS_764257611_4.gif SSFS_764332843_1.gif SSFS_764343670_1.gif SSFS_764343670_2.gif SSFS_764344270.gif SSFS_764344270_2.gif SSFS_764344706.gif SSFS_764344733.gif SSFS_764589302.gif SSFS_764589459_1.gif SSFS_764589459_2.gif SSFS_764589459_3.gif SSFS_764589459_4.gif SSFS_764589459_5.gif SSFS_764665294.gif SSFS_765302566_1.gif SSFS_765302566_2.gif SSFS_765302705_1.gif SSFS_765308183_1.gif SSFS_765308183_2.gif SSFS_765311654_1.gif SSFS_765311710_1.gif SSFS_765311732_1.gif SSFS_765318312_1.gif SSFS_765320955.gif SSFS_765327588.gif SSFS_765398187.gif

SSFS_765398187_2.gif SSFS_765398305_1.gif SSFS_765398305_2.gif SSFS_765398306_1.gif SSFS_765398306_2.gif SSFS_765398306_3.gif SSFS_765398306_4.gif SSFS_765398306_5.gif SSFS_765475670.gif SSFS_765479098.gif SSFS_765484281.gif SSFS_765485180.gif SSFS_765485500.gif SSFS_765485500_2.gif SSFS_765536392.gif SSFS_765536392_2.gif SSFS_765545485.gif SSFS_765545485_2.gif SSFS_765549504.gif SSFS_765549517.gif SSFS_765549540.gif SSFS_765552348.gif SSFS_765968820.gif SSFS_765971138.gif SSFS_765971138_2.gif SSFS_766144998.gif SSFS_766149536.gif SSFS_766150426.gif SSFS_766150430.gif SSFS_766150558.gif SSFS_766151638.gif SSFS_766151639.gif SSFS_766417300.gif SSFS_766417306.gif SSFS_766417306_2.gif SSFS_766417436.gif SSFS_766426403.gif SSFS_766426460.gif SSFS_766426497.gif SSFS_766426552.gif SSFS_766430777.gif SSFS_766431882.gif SSFS_766466742.gif SSFS_766590206.gif SSFS_766709295.gif

SSFS_766709295_2.gif SSFS_766712859.gif SSFS_766772044.gif SSFS_766779324.gif SSFS_766783915.gif SSFS_766784545.gif

SSFS_COHE_765321000_1.gif

SSFS_COHE_766713000_SSFSCorr_GxIB.gif SSFS_COHE_766713000_SSFSCorr_GxMC.gif SSFS_COHE_766713000_SSFSCorr_IBLCtx.gif SSFS_COHE_766713000_SSFSCorr_IBLCtytz.gif SSFS_COHE_766713000_SSFSCorr_ISYSAA.gif SSFS_COHE_766990000_PrRFCACp_ScNEzGC.gif SSFS_Calib_HG.gif

SSFS_Calib_HG2.gif SSFS_Calib_LG.gif SSFS_FFT_763633900.gif SSFS_FFT_763633937_1.gif SSFS_FFT_763633937_2.gif SSFS_FFT_763633937_3.gif SSFS_FFT_763643302_1.gif SSFS_FFT_764113906.gif SSFS_FFT_764114430_1.gif SSFS_FFT_764114500_1.gif SSFS_FFT_764114500_B1pd1ACp.txt SSFS_FFT_764114500_SSFSCorr.txt SSFS_FFT_764589420.gif

SSFS_FFT_764590250.gif SSFS_FFT_764597900.gif

SSFS_FFT_765321000_B1pd1ACp.txt SSFS_FFT_765321000_SSFSCorr.txt SSFS_FFT_765489291_B1pd1ACp.txt SSFS_FFT_765489291_SSFSCorr.txt SSFS_FFT_765549800.gif

SSFS_FFT_765549800_10.gif SSFS_FFT_765549800_11.gif SSFS_FFT_765549800_12.gif SSFS_FFT_765549800_2.gif SSFS_FFT_765549800_3.gif SSFS_FFT_765549800_4.gif SSFS_FFT_765549800_5.gif SSFS_FFT_765549800_6.gif SSFS_FFT_765549800_7.gif SSFS_FFT_765549800_8.gif SSFS_FFT_765549800_9.gif

SSFS_FFT_765549800_RFCACp.txt SSFS_FFT_766456300.gif

SSFS_FFT_766456300_B1pd1ACp.txt SSFS_FFT_766456300_SSFSCorr.txt SSFS_FFT_766467000.gif

SSFS_FFT_766467000_B1pd1ACp.txt SSFS_FFT_766467000_SSFSCorr.txt SSFS_FFT_766713000_B1ACp.txt SSFS_FFT_766713000_B1d6ACp.txt SSFS_FFT_766713000_B1d6_SSFSCorr.gif SSFS_FFT_766713000_RFC.gif

SSFS_FFT_766713000_RFC_2.gif SSFS_FFT_766713000_RFC_3.gif SSFS_FFT_766713000_RFC_4.gif SSFS_FFT_766713000_SSFSCorr.txt SSFS_FFT_766713000_zErrGC_2.txt SSFS_FFT_766713000_zErrGCx30_2.txt SSFS_FFT_766713100_zErrGC.txt SSFS_FFT_766990000.gif

SSFS_FFT_766990000_SSFSACp.txt SSFS_FFT_766990000_SSFSCorr.txt SSFS_FFT_766990000_zErrGCx30.txt SSFS_FFT_767040722_zErrGC_B1pACp.gif SSFS_JPC1.gif

SSFS_JPC2.gif

SSFS_RMS_765549800_RFCACp.txt

… careful analysis of 206 plots

Experimental work (4)

Large transient at lock acquisition

• Algo in Gc: start SSFS when correction signal is zero, and holds this zero

=> error signal is generally not null

=> excites the step response (large transient)

• NEW Algo in Gc to be implemented: start SSFS when error signal is null

=> then switch off the filter

=> decay slowly correction signal to zero

Experimental work (5)

Transient / algo in Gc

• While working, we found in-loop resonances in the MC suspension at 18 Hz, 25 Hz, 53 Hz, 128 Hz, 333 Hz, …

• The reference mass is not holded by its wires, but by the coil wires!

MC reference mass wires are not stresses and may have random resonance frequencies, varying in time.

Experimental work (6)

MC SUSPENSION RESONANCES

Noise performances

• 1 laser, 1 FP cavity, lock with high unity gain

• error signal = readout noise

• correction signal = beating note of laser cavity and FP cavity,

Usually dominated by laser cavity (« free running laser noise »)

• 1 laser locked on a FP cavity, lock of another cavity

2nd cavity error signal = beating note of the 2 FP cavities ex: experiment in Orsay in 95,

ex: « ISYS »: IMC and reference cavity

• 1 laser, 3+ FP cavities

ex: « north arm » (C2)

North arm, IMC, reference cavity

cf VIR-NOT-OCA-1390-260

Noise performances

Noise performances

Next activities (1/2)

SSFS on north arm

* update saturation in SSFS rack (+ RAMPEAUTO)

* new algo in Gc (trigger condition)

* check of IB-DSP/MC-DSP DOL

* Gc: trigger=0 in the sequence

* Gc: lock of RFC without intermediate steps

* test Coulon’s filter (NI+NE)

* increase of UG of SSFS

* test with Virgo final corrector stage

* SSFS -> ISYS reliability

SSFS on « recombined » interferometer

* ready

Input MC control

* TF measurement (again) (z,Tx,Ty)

* control from 2 coils

* change coil wires of MC suspension?

1-2 shifts 1 shift

2 shifts 1 shift 4 shifts 2-3 shifts

2-3 shifts

½ shift

???

(tentative durations)

Next activities (2/2)

Lock of recycling cavity

* improve rms of IMC length noise?

* improve 50 Hz level on reference cavity

* detune recycling cavity of 25 nm (1/10

linewidth)

SSFS on virgo

Noise and requirements

* cabling of « pertub » input

* calibration sequence

Simulation

* Siesta simulation of SSFS

(fs = 200 kHz, does not include IMC error signal saturation)

3-4 weeks 2-3 shifts 1 shift 2 weeks

1-2 weeks