Output mode cleaner status

1
Output mode cleaner status

• June 7th, 2004
• (Keita, Luca, Joe B, Stefan B, Valera, Bill K,
  Virginio, Jay, PeterK, Robert S, Nergis, PeterF)

2
OMC setup
OMC
3
The bottom lineexcess noise introduced
?!
4
OMC locking error signals

• Dither MC length (40 kHz)
• Maximizing total power pulls the resonance
  toward the stronger SB, creates an even larger SB
  asymmetry
• Maximizing 2fmoc power SBs pass symmetrically on
  either side of resonance center carrier is on
  resonance center
• AS_I
• Makes SBs at OMC output be balanced
• In practice, electronic AS_I servo is still
  connected (30 Hz BW), and the error signal for
  the OMC lock is taken from the AS_I correction
  signal, and applied to OMC with low bandwidth
  (0.1 Hz BW)
• All 3 have been tried, dont know whether AS_I or
  dither/2fmoc is better

5
Effect on couplings (locked using AS_I)

• When DARM is locked using AS3 (non-OMC beam),
  there is an AS2_Q offset corresponding to 10-12
  m!
• Laser amplitude noise sensitivity is reduced at
  high frequencies
• Presumably, coupling through frequency noise
  becomes important at low freqs isnt affected
  by the OMC

6
Effect on couplings (contd)

• Oscillator phase noise w/ OMC locked on AS_I

Green original measurement Blue AS3 Red AS2
(OMC)
7
Effect on AS_I signal
Factor of 7 decrease with OMC
8
The excess noise
9
Acoustic in nature
With injections
ASPD2
No injections
ASPD3
10
What can it be?

• The same whether the OMC is
• In vacuum or not (0.1mTorr) with its suspension
• Mounted on rigid block in air
• Beam clipping?
• Seems not to be the case aperture test on
  optical elements and OMC
• Back scattering?
• Unlikely since ASPD3 is not contaminated
• OMC length fluctuations?
• Noise contribution is not limiting now (measured
  using AS2I to estimate the OMC length
  fluctuations)
• Beam jitter?
• High order modes (TEMxy) transmitted by OMC a
  problem?
• ?

11
Beam jitter injection

• Highly sensitive to tapping of optics
• With significant acoustic injections, the excess
  noise could be accounted by beam jitter

12
Beam jittercouplings

• Significant angle-to-displacement coupling
• With pitch/yaw stabilization (5Hz)
• Theta_div3mrad
• Estimate of vertical displacement at OMC waist
• (RMS at OMC waist) 30 of waist, no pitch/yaw
  stabilization
• Beam stabilization (lt5Hz, pitch and yaw of one
  steering mirror) no effect on noise
• Repeat with 4 DOFs? Increase BW (how)? Increase
  noise and then stabilize?
• Angle dithering, minimizing 1 omega coupling no
  effect
• Repeat with 4 DOFs?

13
High order modes transmitted?

• Idea could an high order mode (TEM) be
  transmitted by the OMC and significantly
  contribute to the noise?

14
Modal composition of dark fringe
TEM00
TEM00
Dangerous?

• 8th order mode (nm8, odd) close to TEM00 is it
  a problem?
• Could experiment with a different output mode
  cleaner -- different g factor (g0.3 -gt 0.41)

15
Recent observations

• Misalign OMC transmitted beam so that the PD2 DC
  voltage drops by ½
• Compare ASD of AS2_Q with and without the change
• Account for gain change
• Light level dropped by 2x (as seen by the DC
  voltage reading)
• But calibration line dropped by 5x (?)
• Adjusted spectra by 5x
• Spectra very similar (apart from a few peaks that
  disappeared in the misaligned spectra)

16
Recent Observations

• Placed Iris in front of OMC
• Closed the Iris around beam until spectrum
  changed
• Calibration line was reduced by 4.7x, spectra was
  scaled and compared with no Iris
• Result
• lt80Hz the spectra are similar
• gt80Hz significant change
• Note the Iris is spatially filtering the beam,
  reducing the power of higher order modes
• Note plexiglass in beam to OMC visibly modified
  the beam on transmission (CDD) but no perceivable
  change in AS2Q (using headset)

17
Recent Observations

• New mode-matching
• moved OMC by 1.5cm
• Michelson bright fringe efficiency 70 -gt 80
• Table/electronics clean-up
• Placed 2 PZT actuated steering mirrors
• 4x light attenuation on both AS2 and AS3, OMC not
  locked, locked on AS3
• Overall noise floor the same but structures are
  different (60Hz due to PZTs)
• 1200Hz broad structure gone
• 4-5kHz structure appeared

18
Other tests?

• Could dither 4 angles, minimizing 1 omega and
  tracking the noise floor
• But one axis of PZT actuated mirror broken
• Replace OMC end mirror with new one?
• Could inject significant acoustic noise so that
  quad diode senses jitter at high frequency,
  stabilize the beam at high frequency and measure
  AS2_Q?
• But running out of ideas.

19
Considerations for new design

• Cavity bandwidth could be narrower to attenuate
  the RF SBs
• Limited by sensitivity to OMC length fluctuations
• Gregg Hs coating thermal noise estimate 10-17
  m/rtHz, versus 10-16 m/rtHz reqd for present
  design
• Decrease BW by factor of 2 -gt SB power decreased
  by factor of 2
• Keep finesse the same (40) and reduce BW by
  lengthening cavity ?
• PZT could have much smaller range (presumably
  could be stiffer)
• Bigger mirrors, bigger bore in spacer

20
Next ?

• Need to understand noise source, but even so,
  dont see how acoustic couplings and/or noise is
  going to be reduced by a factor of 1000.
  Options
• Entire AS port LSC detection chain in vacuum
• Forget the OMC, and make many more LSC
  photodetectors
• Pursue fabrication of our own OMC, or is it too
  early?
• Callum Janeen have started on a suspension
  design, soft in all directions

21
(No Transcript)