LIGO astrophysics

1
Frequency noise of 40m PSL and Mode
Cleaner Osamu Miyakawa Feb. 10, 2003 LIGO
committee meeting

• PSL frequency noise budget measurement
• Evaluation of PMC frequency noise spec

2
40meter Prototype interferometer for Advanced LIGO
BS
ITM
PMC
PSL
ITM
FSS
MOPA126
Detection bench
Mode Cleaner

• 10W MOPA126
• Frequency Stabilization Servo (FSS)
• Pre-Mode Cleaner (PMC)
• 13m Mode Cleaner with digital controlled
  suspension
• Stable and good noise performance

South Arm
Completed
3
Free running Frequency noise of40m 10W laser
MOPA126
09/25/2002

• 10 times better than design.
• Free running noise is estimated by closed FSS
  error signal and FSS closed loop gain.
• 1/f behavior

4
Control of Pre-Stabilized Laser (PSL)
PSL
PMC
MOPA126
TO 13m MC
FSS
Pockels Cell
Fast PZT
Thermal
Open loop gainGFSS
5
Open loop gain of 40m FSS
09/26/2002

• Unity Gain Frequency 309kHz
• (design 600kHz)
• Phase margin 28degree
• Gain is limited by the noise performance.
• This low UGF causes a phase delay on the MC loop
  and limits the UGF of MC later.

6
Frequency noise of 40m PSL
11/18/2002

• Frequency noise of MOPA126 is suppressed by the
  open loop gain of FSS.
• Frequency noise meets the requirement in all
  frequency.
• Noise is measured with MC lock.
• Frequency noise is estimated by the in-loop error
  signal of FSS.
• Length fluctuation of FSS is not included.
• Need another cavity -gt13m MC.

7
13m Mode Cleaner
MC1,MC3
13m
8
Measured open loop gain of 40m MC
12/23/2002

• Calculated gains obtained by Ranas model on
  Matlab agree with measurements.
• Unity gain frequency67.2kHz
• (designover 100kHz)
• Phase margin28.4deg
• Phase delay on total loop at high frequency
  limits the unity gain frequency.
• Cross over frequency26.6Hz
• Phase delay on Mirror loop by A/D converter
  (ICS110B).

9
Lock stability of 40m MC with detection mode
Jan.10/2003 17PM Jan.12/2003 8AM

• Digital controlled suspensions.
• Smooth Lock acquisition (within 5sec).
• Robust lock.
• It would be longer if not intentionally broken.

39 hours lock with detection mode
MC reflected power
MC transmitted power
10
Residual Frequency noise of MC

• Residual Frequency noise
• In-loop noise
• Out-loop noise

Fres
Suppression factor by MC gain, estimated by
another measurement

• Actual Measurement
• Calibration

11
Residual Frequency noise of MC

• Residual Frequency noise
• In-loop noise
• Out-loop noise

Fres
Ne

Suppression factor by MC gain, estimated by
another measurement
ltltNe
x EpafEl

• Actual Measurement
• Calibration

• To avoid the electronic noise of feedback filter

12
Noise improvement of MC

• Meets the requirement except for low frequency
  and bump around 600Hz.

1
2
3
5
4
6
7
13
MC Frequency noiseas a Monitor of PSL Frequency
noise
Fres

• Residual noise is higher than the frequency noise
  of PSL.

14
Noise budget of 40m MC

• Low frequency noise is limited by seismic noise
  of small stack of MC2.
• Unknown noise around 100Hz (saturation of VCO
  loop?).
• Beam jitter noise on PSL and Frequency noise
  limit the high frequency.
• Coil driver noise, Intensity noise, Feedback
  filter noise, Shot noise and Detector noise are
  much lower than total noise.

Saturation on VCO loop?
Beam jitter
Seismic Noise
Frequency noiseBeam jitter
15
40m PMC frequency noise

• HEPA filter above the PSL shakes the optical
  table.
• The optical table shakes mirrors and periscope on
  PSL.
• The fluctuation of mirrors causes a beam jitter.
• PMC detects the beam jitter as a length
  fluctuation of PMC.
• PMC frequency noise is much better than
  requirement.

16
Noise improvement of MC

• HEPA filter also causes PSL frequency noise.
• Beam jitter induces frequency noise.

17
Contribution of PMC frequency noiseto PSL
frequency noise
Beam jitter after PMC

• PMC length fluctuation noise still limits the PSL
  frequency noise, even if PMC frequency noise much
  better than requirement !

18
Open loop T.F. of 40m PMC servo

• The contribution depends on the PMC open loop
  gain.
• 31kHz notch filter.
• 71kHz notch filter to avoid the oscillation by
  resonance of PZT.
• UGF7.6kHz
• (Design UGF600Hz)
• Phase Margin55.7degree
• Gain is 12 times larger than design.

19
PMC Gain increment for PSL frequency noise

• With increasing gain, measured PSL frequency
  noise meets the requirement above 1kHz.
• No PMC length noise contribution to PSL frequency
  noise except for several peaks.
• Beam jitter noise added downstream of PMC still
  remains.
• For reasonable PMC gain, design requirement of
  PMC frequency noise is underestimated 30 times.

20
Noise improvement by increment of Length gain

• Increment of the length gain decreases PSL
  frequency noise.
• This slope around 100Hz is fluctuating up and
  down.
• Length gain is limited by phase delay on length
  loop.
• Saturation in VCO loop?

Length gain16dB
Length gain30.4dB
21
Measured cross over frequency of MC
Gl
VCO loop
Gm
Mirror loop
V2
V1

• Cross over frequency26.6Hz
• Big phase delay on Mirror loop by 2msec time
  delay of A/D converter (ICS110B).
• gtgtChange ICS110B to Pentek for increasing the
  length gain.

22
Measured open loop gain of 40m MC
12/23/2002

• The frequency stabilization by Mode Cleaner is
  performed by
• area.
• Increasing the length gain will decrease the
  frequency stabilization for interferometer in
  future.
• gtgtElliptic filter or more boost.
• MC transmitted frequency noise will be measured
  by single arm cavity.

23
Estimation of Seismic noise

• Measured ground motion by accelerometer.
• Transfer function of MC2 chamber including stack
  and suspension is assumed.
• Need to measure the stack transfer function for
  precise estimation.
• gtgtWill be measured in the next venting.

24
Estimation of PSL frequency noisefor 40m
interferometer

• -40dB Common Mode Rejection Ratio (CMRR) is
  assumed.
• Measured residual frequency noise of MC and
  measured OLTF are used.
• Need more CMRR or more OLTF to reach the target
  sensitivity.
• gtgtActual estimation will be performed by single
  arm cavity in future.

25
Summary

• PSL frequency noise measured by MC almost meets
  the requirement.
• Almost all noise sources of PSL frequency noise
  have been estimated.
• Design requirement of PMC frequency noise should
  be re-evaluated.
• Recommendation
• Turn off the HEPA filter of PSL while
  observation.
• Acoustic shield for PSL optical table.
• Change MC2 chamber including small stack.
• Next Plan
• Install BS, ITMs, ETMs, PRM and SRM.
• Measurement of MC transmitted frequency noise
  using single arm cavity
• Measurement of Common Mode Rejection Ratio.