[JC, Yuta]

We have found that MC1 coils are causing 60 Hz noise.
Tripping watchdogs for MC1 coils reduced 60 Hz noise seen in YARM by a factor of 100.

Method:
 - Locked YARM with POY11 and measured YARM sensitivity to use it for 60 Hz frequency noise monitor0.187493**0.5
 - Tripped MC1, MC2, MC3 coil output watchdogs to see if they are causing this 60 Hz frequency noise. IMC WFS were turned off.

Result:
 - Attachment #1 is YARM sensitivity and MC_F in Hz with MC1,2,3 untripped (dotted) and MC1 tripped (solid).

YARM (PSL locked vs Yarm), MC1,2,3 untripped: 6.0e2 Hz/rtHz (2.6e2 Hz RMS)
MC_F (sum of noises in IMC loop), MC1,2,3 untripped: 4.8e4 Hz/rtHz (2.1e4 Hz RMS)
YARM (PSL locked vs Yarm), MC1 tripped: 6.6e0 Hz/rtHz (2.9e0 Hz RMS) -- reduced by a factor of 100
MC_F (sum of noises in IMC loop), MC1 tripped: 4.7e4 Hz/rtHz (2.0e4 Hz RMS)

 - We have also tried tripping MC2 and MC3 coils, but they didn't make much difference.
 - Untripping only one of MC1 face coils created 60 Hz frequency noise, so all the face coils seem to have the same level of 60 Hz noise.

Next:
 - Inspect around MC1 coil driver

[Anchal, Yuta]

We think we have narrowed down the source of 60 Hz noise to one fo the following possibilities:

• Ground loop present along the MC1 suspension damping loop
• 60 Hz DAC noise on inputs of MC1 coil driver
• 60 Hz noise injected at dewhitening board before the dewhitening filter

The second and third cases are unlikely because we see 60 Hz noise present only in MC1 coils, not MC3 coils while they both share the same connection from DAC to SOS dewhitening filter boards as they share the SOS dewhitening board D000316-A. So it is unlikely that only the MC1 channels have this noise while the MC3 channels do not.

This inference was made from following observations:

Note: Turning ON analog dewhitenign on MC1 coil is done by turning off FM9 switch which is the simulated digital dewhitening filter. Also note that theanalog dewhitening filter has an attenuation of 30 dB at 60 Hz.

MC1 has an unconvetional setup where the satellite amplifier is from the new generation while the coil driver and dewhitening boards are from the old generation. The new generation satellite amplifiers sen PD signal through differential ended signals but the old generation PD whitening interface expects single ended inputs, so we ahve been using PD monitor outputs from the satellite amplifier which connects the ground of the two boards to each other. Maybe this is the reason for the ground loop.

Below are summary of electronics around MC1 and cable disconnection tests.
These suggest that the 60 Hz noise is probably from somewhere between DAC and the coil driver.
For now, we can work on IFO with SimDW off.

MC1 local damping electronics diagram:
Vacuum Flange
|| DB25 cable x2
Satellite Amp Chassis (LIGO-S2100029, LIGO-D1002818)
|| DB9 split cable
Suspension PD Whitening and Interface Board (LIGO-D000210)
||| 4pin LEMO x3
Anti-aliasing filter
|
ADC
|
CDS
 (SimDW is zpk([35.3553+i*35.3553;35.3553-i*35.3553;250],[4.94975+i*4.94975;4.94975-i*4.94975;2500],1,"n") gain(1.020); InvDW is the inverse)
|
DAC
|
SOS Dewhitening and Anti-Image Filter (LIGO-D000316) Shared with MC3
 (has 2ea. 800 Hz LPF & 5th order, 1 dB ripple, 50 dB atten, 28Hz elliptic LPF that can be turned on or bypassed)
||||| SMA-LEMO cable x5
 ("test in" are used; inputs can be disconnected with watchdogs)
SOS Coil Driver Module (LIGO-D010001, LIGO-D1700218)
 (HV offsets from Acromag are added at the output (independent from watchdogs))
|| DB9 split cable
Satellite Amp Chassis

Disconnecting cables:
 - Disconnecting cables between Satellite Amp Chassis and Suspension PD Whitening and Interface Board didn't help reducing 60 Hz noise.
 - Disconnecting LEMO cables between Suspension PD Whitening and Interface Board and Anti-aliasing filter didn't help reducing 60 Hz noise.
 - Turning off C1:SUS-MC1_SUSPOS/PIT/YAW/SIDE outputs didn't help reducing 60 Hz noise.
 - Turning off SimDW reduced 60 Hz noise.
 - Turning off watchdogs reduced 60 Hz noise.

Dewhitening filters:
 - When 60 Hz frequency noise was high, SimDW was on, but InvDW was off, which is in a weired state.
 - Now, all the MC suspensions have SimDW turned off and InvDW turned on (which supposed to turn on analog dewhitening filter, which is probably 28 Hz ELP which has a notch at 60 Hz)
 - Probably, when realtime model modifications for BH44 was made on Jan 17, coil dewhitening filter situation was not burt restored correctly, and we started to notice 60 Hz noise (which was already there but didn't notice because of dewhitening).
 - See 40m/17431 for the timeline, possibly related elogs 40m/17359, 40m/17361 about MC1 dewhitening switching on Dec 14-16.

Next:
 - Check if analong dewhitening filter actually has 28 Hz ELP by measuring transfer functions
 - Design SimDW and InvDW to correctly take into account of real dewhitening filters

Overtime the coil output filters on IMC optics have drifted into a bad configuration. Today at the meeting, Rana told us the correct configuration for these filters. I'll summarize this here and we have changes the filters on all IMC optics, MC1, MC2, and MC3 to match this configurations:

MC1 and MC3

Analog side:

Both MC1 and MC3 have a 28 Hz 5th order elliptical low pass filter as teh dewhitening filter in LIGO-D000316

Digital side:

At the coil output filters named as C1:SUS-MC1_ULCOIL, the filter module FM9 is connectd in RTCDS to the analog dewhitening filter such that only one of the two can remain ON. So For MC1 and MC3, we put a ellip("LowPass", 5, 1, 50, 28) filter on FM9 for all 5 coil output filters.

Note: We do not add a inverse dewhitening filter at FM10 like most other optics as inverting this filter will create resonant peaks at the dips of the elliptical filter which we want to avoid and we anyways do not use MC1 and MC3 optics for any kind of actuation above 20 Hz.

MC2

Analog side:

For MC2, the dewhitening filter is a 10 Hz pole, 30Hz zero like most other suspended optic.

Digital side:

At the coil output filters named as C1:SUS-MC2_ULCOIL, the filter module FM9 is connectd in RTCDS to the analog dewhitening filter such that only one of the two can remain ON. So For MC2 we put a SimDW filter which is matched to the anlog filter. We also put a InvDW filter on FM10, which is the analytical inverse of the SimDW filter. This filter does the anti-dewhitening required on the digital side and should be always ON.

To MC2 equivalent to other IMC optics in terms of overall transfer function for the local damping loops and ASC loops, we need additional 28 Hz elliptical low pass filter in these loops. But such a filter should not be in the path of LSC feedback when MC2 is used for locking CARM with a bandwidth of ~100 Hz. Thus, we put a ellip("LowPass", 5, 1, 50, 28) filter on FM6 of the following filters, which should be always ON as well:

• C1:SUS-MC2_ASCPIT
• C1:SUS-MC2_ASCYAW
• C1:SUS-MC2_SUSPOS
• C1:SUS-MC2_SUSPIT
• C1:SUS-MC2_SUSYAW
• C1:SUS-MC2_SUSSIDE

Effect on 60 Hz Noise

With the above changes, we see that the 60 Hz noise is same as the previous levels when we use the analog dewhitening filter (28 Hz elliptical filter) for MC1. We can move forward with our science experiments with that configuration but there is still something fishy about MC1 in comparison to MC3 which does not have this behavior. So this still needs to be looked at in future.

Wiki page for filter details and configurations

Information of this kind should be stored in a wiki page in my opinion. We should have a page where we list all common filter configurations for our suspensions and other loops, that can be generally classified and is useful for understanding legacy configuration for future folks who work here. I'm starting such a wiki page here, where I'll dump more information as I collect it and get time. Everyone is encouraged to update this in there free/procrastination times.

FPMI BHD with LO phase locked using BH55 is recovered after 60 Hz frequency noise saga.
Attachment #1 shows the calibrated FPMI spectrum with RF(AS55_Q) readout and BHD, compared with those taken on January 13 (40m/17400, before BH44 installation).
There is unknown excess noise at round 30-40 Hz. This is not from MC2 DAC noise, as turning on/off dewhitening filters didn't make a difference.
Attachment #2 shows the samething but zoomed at 60 Hz. 60 Hz noise is actually reduced by an order of magnitude compared with what we had before BH44 installation.
Note that RF amp for BH55 which was there on January 13 was removed now (40m/17413).
LO_PHASE is locked with BH55_Q, under whitening gain of 45 dB, whitening filter on, C1:LSC-BH55_PHASE_R=-110 deg, C1:HPC-LO_PHASE_GAIN=-10, FM5 and FM8. This gives UGF of ~20 Hz (we used to be able to get ~100 Hz, but not possible now).

Locking LO_PHASE with BH44 is not stable now, probably due to small optical gain. We might have to install RF amp for BH44.

Next:
 - Check the beam alignment to BH44 RF PD
 - Install RF amp for BH44
 - Re-install RF amp for BH55

Since we need more signal for both BH55 and BH44 to compare LO phase locking scheme, BH55 and BH44 RF outputs are now amplified with ZFL-1000LN+ and ZFL-500HLN+ respectively (see Attachment #1).
The amplifiers each draw ~0.1 A current of 15V DC power supply, and Sorensen power supply now reads 6.9 A (see Attachment #2).
With ITMX single bounce and LO beam fringing, BH55_Q (45 dB whitening gain, C1:LSC-BH55_PHASE_R=-110 deg) gives ~500 counts in amplitude, and BH44_Q (24 dB whitening gain, C1:LSC-BH44_PHASE_R=4.387 deg) gives ~100 counts in amplitude (and they are orthogonal) (see Attachment #3).

Ideally, BH55 and BH44 should give orthogonal signals to lock LO phase (40m/17302).
This was checked with various interferometer configurations.
BH55 and BH44 are indeed orthogonal in ITM single bounce and MICH, but was not measurable in FPMI.
Maybe we should investigate BH44 in MICH BHD configuration first to see why BH44 is very noisy in FPMI.

Method:
 - X-Y plotted BH55_Q and BH44_Q and fitted with an ellipse to derive amplitude of each signal and phase difference between them.
 - Amplitude and phase differences are calculated using the following equations, where (ap, bp) are the semi-major and semi-minor axes, respectively, and phi is the rotation of the semi-major axis from the x-axis. (Thanks to Tomohiro for checking the calculations!)

 xAmp = np.sqrt((ap * np.cos(phi))**2 + (bp * np.sin(phi))**2)
 yAmp = np.sqrt((ap * np.sin(phi))**2 + (bp * np.cos(phi))**2)
 phaseDiff = np.arctan(bp/ap*np.tan(phi)) + np.arctan(bp/ap/np.tan(phi))

Jupyter notebook: /opt/rtcds/caltech/c1/Git/40m/scripts/CAL/BHD/MeasurePhaseDiff.ipynb

 - This was done un following 4 configurations.
  - ITMX single bounce vs LO
  - ITMY single bounce vs LO
  - AS beam in MICH locked with AS55_Q vs LO
  - AS beam in FPMI locked with REFL55/AS455 vs LO
 - For each configuration, RF demodulation phases were tuned to minimize I.
 - Statistical error was estimated by calculating the standard deviation of 3 measurements.

 - Also, FPMI BHD sensing matrix was measured when FPMI is locked with REFL55/AS55, and LO_PHASE is locked with BH55_Q or BH44_Q.

Jupyter notebook: /opt/rtcds/caltech/c1/Git/40m/scripts/CAL/SensingMatrix/ReadSensMat.ipynb

Result of BH55/BH44 orthogonality check:
 - ITMX single bounce vs LO
            Demod phase         Amplitude       Phase Diff
    BH55_Q  -94.4 +/- 0.2 deg   600.4 +/- 0.6   
    BH44_Q  -9.0 +/- 0.2 deg    124.3 +/- 0.2   -86.7 +/- 0.1 deg
 - ITMY single bounce vs LO
            Demod phase         Amplitude       Phase Diff
    BH55_Q  -92.9 +/- 0.3 deg   588.0 +/- 0.3   
    BH44_Q  -8.9 +/- 0.3 deg    123.0 +/- 0.1   -87.2 +/- 0.1 deg
 - AS beam in MICH locked with AS55_Q vs LO
            Demod phase         Amplitude       Phase Diff
    BH55_Q  -68.7 +/- 0.8 deg   44 +/- 1   
    BH44_Q  -28.5 +/- 1.7 deg   10.3 +/- 0.1   -84 +/- 2 deg
 - AS beam in FPMI locked with REFL55/AS455 vs LO
            Demod phase         Amplitude       Phase Diff
    BH55_Q  35 +/- 3 deg        257 +/- 4   
    BH44_Q  -16 +/- 3 deg       44 +/- 1        -77 +/- 3 deg
 - Attachmented pdf contain example X-Y plots from each configuration. For ITM single bounce and MICH, BH55 and BH44 seems to be orthogonal, but for FPMI, ellipse fit does not go well.
 - Difference in the BH55 demodulation phase for ITMX single bounce and ITMY single bounce (1.5 +/- 0.4 deg) agrees with past measurement and agree marginally with Schnupp asymmetry (40m/17274).
 - Maybe we can derive some length differences using these demodulation phases.

Result of FPMI sensing matrix measurements:
 - Below is the sensing matrix when FPMI is locked with REFL55/AS55, and LO_PHASE is locked with BH55_Q. BH44 is noisier than BH55, and the response to LO1 is consistent with zero. This is also consistent with BH44 being orthogonal to BH55, but the error bar is too large to say.

Sensing matrix with the following demodulation phases (counts/m)
{'AS55': -168.5, 'REFL55': 92.32, 'BH55': -110.0, 'BH44': -8.93097234187195}
Sensors       DARM @307.88 Hz           CARM @309.21 Hz           MICH @311.1 Hz           LO1 @315.17 Hz           
AS55_I       (-2.49+/-8.35)e+10 [90]    (+2.36+/-0.85)e+11 [0]    (-0.64+/-3.99)e+10 [0]    (+0.57+/-4.07)e+09 [0]    
AS55_Q       (-3.50+/-0.08)e+11 [90]    (+0.09+/-1.20)e+11 [0]    (-0.79+/-8.66)e+09 [0]    (-0.70+/-5.96)e+08 [0]    
REFL55_I       (+0.72+/-8.09)e+11 [90]    (-1.42+/-2.75)e+12 [0]    (+0.00+/-1.37)e+11 [0]    (-0.38+/-2.78)e+09 [0]    
REFL55_Q       (+0.19+/-1.93)e+11 [90]    (-2.14+/-6.92)e+11 [0]    (+0.00+/-3.16)e+10 [0]    (+0.17+/-1.19)e+09 [0]    
BH55_I       (-1.41+/-0.55)e+11 [90]    (+1.46+/-2.28)e+11 [0]    (-1.60+/-3.72)e+10 [0]    (-0.07+/-3.05)e+09 [0]    
BH55_Q       (+2.05+/-3.10)e+10 [90]    (-1.72+/-4.86)e+10 [0]    (-0.31+/-2.19)e+10 [0]    (-3.06+/-0.87)e+09 [0]    
BH44_I       (-0.41+/-2.03)e+11 [90]    (+0.10+/-2.39)e+11 [0]    (+0.06+/-1.31)e+11 [0]    (-0.01+/-2.71)e+10 [0]    
BH44_Q       (+0.14+/-3.23)e+12 [90]    (+0.02+/-3.67)e+12 [0]    (+0.07+/-2.03)e+12 [0]    (-0.02+/-4.22)e+11 [0]    
BHDC_DIFF       (+8.49+/-0.47)e+11 [90]    (-0.06+/-2.93)e+11 [0]    (-0.16+/-1.01)e+10 [0]    (-0.27+/-2.04)e+09 [0]    
BHDC_SUM       (-2.30+/-0.11)e+11 [90]    (+0.68+/-7.92)e+10 [0]    (-0.44+/-3.33)e+09 [0]    (-0.63+/-5.63)e+08 [0]    

 - Below is the sensing matrix when FPMI is locked with REFL55/AS55, and LO_PHASE is locked with BH44_Q. BH44 response to LO1 is again consistent with zero. Locking LO_PHASE with BH44 is not robust. Also, BHDC_DIFF response to DARM is less, compared with LO_PHASE locked with BH55_Q. This means that BH55 is somehow better than BH44 in our FPMI BHD, which contradicts with simulations (with no contrast defect and DARM offset).

Sensing matrix with the following demodulation phases (counts/m)
{'AS55': -168.5, 'REFL55': 92.32, 'BH55': -110.0, 'BH44': -8.93097234187195}
Sensors       DARM @307.88 Hz           CARM @309.21 Hz           MICH @311.1 Hz           LO1 @315.17 Hz           
AS55_I       (-7.56+/-4.89)e+10 [90]    (+1.61+/-1.05)e+11 [0]    (+0.51+/-2.48)e+10 [0]    (+0.88+/-8.02)e+08 [0]    
AS55_Q       (-3.62+/-0.05)e+11 [90]    (+0.02+/-1.23)e+11 [0]    (+0.67+/-3.73)e+09 [0]    (+0.02+/-1.28)e+08 [0]    
REFL55_I       (+1.09+/-8.12)e+11 [90]    (-1.47+/-2.82)e+12 [0]    (+0.01+/-1.34)e+11 [0]    (+2.20+/-5.29)e+08 [0]    
REFL55_Q       (+0.22+/-1.93)e+11 [90]    (-1.83+/-7.23)e+11 [0]    (+0.02+/-3.18)e+10 [0]    (+0.56+/-1.17)e+08 [0]    
BH55_I       (-1.21+/-0.08)e+12 [90]    (+0.17+/-4.31)e+11 [0]    (-1.24+/-3.02)e+10 [0]    (-0.30+/-3.55)e+09 [0]    
BH55_Q       (-3.83+/-0.30)e+11 [90]    (-0.12+/-1.42)e+11 [0]    (-0.61+/-1.96)e+10 [0]    (-0.21+/-1.49)e+09 [0]    
BH44_I       (-0.22+/-2.01)e+11 [90]    (-0.07+/-2.30)e+11 [0]    (-0.02+/-1.27)e+11 [0]    (+0.08+/-2.62)e+10 [0]    
BH44_Q       (-0.77+/-8.27)e+11 [90]    (-0.13+/-9.51)e+11 [0]    (-0.04+/-5.23)e+11 [0]    (+0.02+/-1.08)e+11 [0]    
BHDC_DIFF       (+1.94+/-0.81)e+11 [90]    (-0.58+/-1.84)e+11 [0]    (+0.18+/-3.53)e+10 [0]    (-0.49+/-3.84)e+09 [0]    
BHDC_SUM       (-2.22+/-0.12)e+11 [90]    (+0.66+/-7.70)e+10 [0]    (-1.04+/-3.97)e+09 [0]    (+0.31+/-6.10)e+08 [0]

Other notes:
 - TRX and TRY are noisier at ~28 Hz when locked with REFL55/AS55 than when locked with POX/POY. DARM signal seems to be contaminated with broad 28 Hz noise. Needs investigation of the cause.
 - BS oplev loops seem to be close to unstable. When FPMI is unlocked, BS is kicked significantly.

Next:
 - Repeat measurement in 40m/17351 with BH44.
 - Compare LO phase noise in MICH configuration when LO_PHASE is locked with BH44 and BH55.
 - Investigate 28 Hz noise in FPMI
 - Tune BS local damping loops

I tried some LO PHASE noise coupling measurements today. With MICH locked using AS55_Q, I control the LO phase using the single RF (BH55_Q) or double RF (BH44_Q) demodulation error signals. The calibrated error and control points for single RF sideband sensing are shown in Attachment #1. In either case feedback loop is closed using FM5, FM8 first with a gain of 1.5 and then a "boost" using FM4. The actuation point is LO1 POS and the UGF was measured to be ~ 35 Hz for both.

** While doing this measurement, I noticed our LO_PHASE dark noise is significantly contributing 180 Hz, 300 Hz and other high line harmonics into the control signal rms so that may be something to look into soon.

I first thought I could use the remaining sensor to measure the noise coupling (e.g. BH44 locks LO phase and BH55 senses injected noise or viceversa), but these two sensing schemes give two different LO phase sensitivities so I decided to just use the calibrated control signals.

-- Noise coupling for BH55_Q --

After locking the LO_PHASE I inject 2 Hz wide uniform noise into three different frequency bands *within the control bandwidth* through C1:SUS-LO2_LSC_EXC, C1:SUS-AS1_LSC_EXC, and C1:SUS-AS4_LSC_EXC. The injected noise settings are captured by Attachment #2 (the screenshot of the excitation settings in diaggui).

I read back the calibrated C1:HPC-LO_PHASE_OUT_DQ representing the true LO_PHASE noise within the control bandwidth and also calibrate the injected noise spectra with the help of the actuation coefficients in [elog40m:17274]. The result is summarized in Attachment #3.

The diaggui template and data for this measurement are saved under /opt/rtcds/caltech/c1/Git/40m/measurements/BHD/BH55Q_NoiseCoupling.xml

-- Noise coupling for BH44_Q -- 

I repeat the same procedure as above and the injected noise settings, and the result is summarized in Attachment #4. 

The diaggui template and data for this measurement are saved under /opt/rtcds/caltech/c1/Git/40m/measurements/BHD/BH44Q_NoiseCoupling.xml

- Discussion -

It seems that noise injected along the AS beam path (AS1-AS4 dither) couples more into the control point of the LO phase. I also seem to be off in terms of calibrating the noise excitation (even though I scaled using the suspension actuation from [elog40m:17274]. General feedback on the methods used for this measurement are welcome of course. 

- Next steps - 

    - Extend this to single RF + audio dither scheme and double audio dither schemes (although it's hard because the control bandwidth is pretty low already)
    - Investigate line noise in RFPD + demod chain (present on the dark noise). 
    - Investigate other more interesting noise couplings, e.g. angular degrees of freedom, RIN, laser freq noise, etc...
    - Repeat under more relevant IFO configurations (e.g. FPMI)

FPMI BHD sensing matrix was measured by an updated method with updated RF demodulation phases for REFL55 and AS55.
Now audio demodulation phase for CARM components is 90 deg to make the sign correct.
Also, oscillators are turned on one by one to reduce contamination between DoFs (especially between MICH and CARM).
These helped a lot in reducing errors.

Sensing matrix with FPMI locked in RF, LO_PHASE locked with BH55_Q using LO1

Sensing matrix with the following demodulation phases (counts/m)
{'AS55': -177.9, 'REFL55': 77.06, 'BH55': -110.0, 'BH44': -8.9}
Sensors       DARM @307.88 Hz           CARM @309.21 Hz           MICH @311.1 Hz           LO1 @315.17 Hz           
AS55_I       (+3.25+/-0.67)e+11 [90]    (-8.63+/-0.41)e+11 [90]    (-1.02+/-1.49)e+09 [0]    (+0.44+/-1.39)e+07 [0]    
AS55_Q       (-6.04+/-0.05)e+11 [90]    (+0.92+/-3.10)e+10 [90]    (+9.10+/-6.78)e+08 [0]    (+0.12+/-2.08)e+07 [0]    
REFL55_I       (+1.18+/-0.03)e+11 [90]    (+2.78+/-0.12)e+12 [90]    (-0.35+/-2.34)e+09 [0]    (-0.94+/-2.38)e+07 [0]    
REFL55_Q       (+5.85+/-0.43)e+09 [90]    (-2.34+/-0.13)e+10 [90]    (+2.39+/-0.38)e+08 [0]    (+3.56+/-7.44)e+06 [0]    
BH55_I       (-3.51+/-3.45)e+10 [90]    (-6.65+/-0.82)e+10 [90]    (-4.91+/-3.03)e+08 [0]    (-1.82+/-0.09)e+09 [0]    
BH55_Q       (+7.86+/-0.29)e+11 [90]    (+2.99+/-0.42)e+11 [90]    (-2.87+/-7.76)e+08 [0]    (+2.81+/-0.15)e+09 [0]    
BH44_I       (-0.34+/-1.99)e+12 [90]    (+0.02+/-1.49)e+12 [90]    (-0.42+/-8.53)e+10 [0]    (-0.01+/-3.08)e+10 [0]    
BH44_Q       (-0.60+/-3.95)e+13 [90]    (-0.01+/-3.00)e+13 [90]    (+0.00+/-1.68)e+12 [0]    (-0.15+/-5.77)e+11 [0]    
BHDC_DIFF       (-9.18+/-0.29)e+11 [90]    (-4.11+/-4.66)e+10 [90]    (+1.46+/-0.10)e+09 [0]    (-1.70+/-0.41)e+08 [0]    
BHDC_SUM       (+2.97+/-0.21)e+11 [90]    (+0.44+/-1.57)e+10 [90]    (-1.01+/-0.06)e+09 [0]    (+2.68+/-0.84)e+07 [0]

 - AS55_Q now has 70% more gain to DARM for some reason (see 40m/17478). Whitening gain haven't changed from 24 dB.
 - There's still some room to tune AS55 RF demodulation phase to maximize DARM response.
 - CARM to REFL55_Q is 100 times smaller than that to REFL55_I; this is good.
 - There's still some room to tune BH55 RF demodulation phase to maximize LO1 response.
 - BH44 doesn't have much response to LO1, probably because LO_PHASE is locked with orthogonal BH55.

Sensing matrix with FPMI locked in RF, LO_PHASE locked with BH44_Q using LO1

 Sensing matrix with the following demodulation phases (counts/m)
{'AS55': -177.9, 'REFL55': 77.06, 'BH55': -110.0, 'BH44': -8.9}
Sensors       DARM @307.88 Hz           CARM @309.21 Hz           MICH @311.1 Hz           LO1 @315.17 Hz           
AS55_I       (+3.94+/-0.52)e+11 [90]    (-1.00+/-0.05)e+12 [90]    (-1.61+/-1.17)e+09 [0]    (+0.45+/-1.52)e+07 [0]    
AS55_Q       (-5.52+/-0.24)e+11 [90]    (+1.19+/-2.99)e+10 [90]    (+1.10+/-0.43)e+09 [0]    (-1.06+/-2.30)e+07 [0]    
REFL55_I       (+8.97+/-0.49)e+10 [90]    (+2.71+/-0.11)e+12 [90]    (-0.38+/-2.28)e+09 [0]    (-0.97+/-2.10)e+07 [0]    
REFL55_Q       (+6.30+/-0.65)e+09 [90]    (-2.01+/-0.12)e+10 [90]    (+2.26+/-0.69)e+08 [0]    (-2.61+/-6.97)e+06 [0]    
BH55_I       (+4.46+/-0.52)e+11 [90]    (-1.52+/-0.27)e+11 [90]    (-1.82+/-0.56)e+09 [0]    (+0.68+/-1.24)e+08 [0]    
BH55_Q       (+9.59+/-0.44)e+11 [90]    (+2.79+/-0.52)e+11 [90]    (+2.75+/-2.49)e+08 [0]    (+2.45+/-1.06)e+08 [0]    
BH44_I       (-0.40+/-2.42)e+12 [90]    (-0.03+/-1.88)e+12 [90]    (-0.03+/-1.13)e+11 [0]    (+0.12+/-4.18)e+10 [0]    
BH44_Q       (-0.19+/-1.09)e+13 [90]    (+0.70+/-7.91)e+12 [90]    (-0.09+/-4.65)e+11 [0]    (+0.11+/-1.34)e+11 [0]    
BHDC_DIFF       (+3.90+/-0.46)e+11 [90]    (+1.06+/-0.18)e+11 [90]    (-4.62+/-1.89)e+08 [0]    (+3.60+/-0.40)e+08 [0]    
BHDC_SUM       (+1.96+/-0.18)e+11 [90]    (-1.08+/-1.29)e+10 [90]    (-8.93+/-1.41)e+08 [0]    (-8.67+/-0.81)e+07 [0]

 - BHDC_DIFF sensitivity to DARM is less than that with LO_PHASE locked with BH55.
 - BH44 sensing matrix has too much error. Requires more averaging time and oscillator amplitude.

Jupyter notebook: /opt/rtcds/caltech/c1/Git/40m/scripts/CAL/SensingMatrix/ReadSensMat.ipynb

Next:
 - Tune AS55, BH55, BH44 RF demodulation phases
 - Try measuring sensing matrix for BH44 with more averaging time, oscillator amplitude, and PD whitening gain
 - Repeat measurement in 40m/17351 with BH44 under MICH configuration.
 - Compare LO phase noise in MICH configuration when LO_PHASE is locked with BH44 and BH55.
 - Make a noise budget in MICH BHD.
 - Investigate 28 Hz noise in FPMI
 - Tune BS local damping loops

[Anchal, Yuta]

We have measured LO phase noise in ITMX single bounce, simple MICH and FPMI configurations with LO phase locked with BH55 or BH44.
We found that BH55 and BH44 have almost exactly same noise in ITMX single bounce, but BH44 is noisier than BH55 in MICH and FPMI configurations.
In any case, LO phase can be locked within 0.1 rad RMS, so optical gain fluctuations in BHD_DIFF should be fine for BHD locking.

Method:
 - We have locked ITMX single bounce vs LO, AS beam under MICH locked with AS55_Q vs LO, and AS beam under FPMI locked with REFL55 & AS55 vs LO, using BH55_Q or BH44_Q
 - In each IFO configuration, we have minimized I phase to set RF demodulation phases for BH55 and BH44.
 - In each IFO configuration, optical gain of BH55_Q and BH44_Q was measured by elliptic fit of X-Y plot for BH55_Q vs BHDC_A or BH44_Q vs BH55_Q.
 - For each LO_PHASE lock, feedback gain was adjusted to set the UGF to around 50 Hz, and actuator used was LO1.
 - LO_PHASE_IN1 was calibrated using the measured optical gain, and LO_PHASE_OUT was calibrated using LO1 actuator gain of 26.34e-9 /f^2 m/counts measured in 40m/17285.
 - To convert meters in radians, 2*pi/lambda is used (which means dark fringe to dark fringe is pi).
 - Below summarizes the result of RF demodulation phases and optical gains (whitening gains were 45 dB for BH55 and 39 dB for BH44). RF demod phases aligns well with previous measurement, but optical gain for BH44 seems higher by an order of magnitude compared with 40m/17478 (whitening gain changed??). Optical gain for BH55_Q is consistent with previous measurement in 40m/17506 (note the demodulation phase change).

LO_PHASE lock in ITMX single bounce
        Demod phase  Optical gain     filter gain
BH55_Q  -99.8 deg    7.6e9 counts/m   -0.3
BH44_Q  -6.5 deg     1.3e10 counts/m  -0.15

LO_PHASE lock in MICH
        Demod phase  Optical gain     filter gain
BH55_Q  -67.7 deg    6.1e8 counts/m   -3.9
BH44_Q  -31.9 deg    8.5e8 counts/m   -3.1

LO_PHASE lock in FPMI
        Demod phase  Optical gain     filter gain
BH55_Q  35.7 deg     3.4e9 counts/m   -0.65
BH44_Q  -9.3 deg     4.3e10 4.3e9 counts/m  -0.84  (Typo fixed on Apr 18, 2023 by YM)

Result:
 - Attached are calibrated LO phase noise spectrum in different IFO configurations.
 - In ITMX single bounce, LO phase noise estimated using BH55 and BH44 are almost equivalent, and LO phase noise in-loop is ~0.04 rad RMS.
 - In MICH, LO phase noise estimated using BH44 is noisier than BH44 at around 20-60 Hz for some reason. LO phase noise in-loop is ~0.04 rad RMS for both cases.
 - In FPMI, LO phase noise estimated using BH44 is noisier than BH44 above ~20 Hz for some reason. LO phase noise in-loop is ~0.03 rad RMS for both cases. Dark noise is not limiting the measurement at least below 1 kHz.

Jupyter notebook: /opt/rtcds/caltech/c1/Git/40m/measurements/BHD/BH55_BH44_Comparison.ipynb

Next:
 - Lock MICH BHD with BH55 and BH44, and compare LO phase noise contributions to MICH sensitivity
 - Investigate why BH44 is noisier than BH55 in MICH and FPMI (offsets? contrast defect? mode-matching?)
 - Reduce 60 Hz + harmonics in BH55 and BH44

[Paco, Yuta]

MICH was locked with balanced homodyne readout with LO phase locked using BH55_Q and BH44_Q.
It turned out that BH44_Q gives better LO phase in MICH configuration (in FPMI, BH55_Q is better; see 40m/17506).
LO phase noise seems to contribute to MICH sensitivity in 30-200 Hz region in BH55 case, and 30-100 Hz in BH44 case (this was not the case in FPMI BHD, see 40m/17392).
The mechanism for this coupling needs investigation.

MICH BHD sensing matrix:
 - MICH BHD sensing matrix was measured when MICH is locked with AS55_Q and LO_PHASE is locked with BH55_Q or BH44_Q.
 - MICH UGF was at around 50 Hz, and LO_PHASE UGF was at around 10 Hz.
 - BHDC_DIFF had better sensitivity to MICH when LO_PHASE was locked with BH44_Q.
 - BH44 component was not measured well.

MICH sensing matrix with MICH locked with AS55_Q and LO_PHASE locked with BH55_Q

Sensing matrix with the following demodulation phases (counts/m)
{'AS55': 2.1, 'REFL55': 76.01784975834194, 'BH55': -63.16236453101908, 'BH44': -39.01036239539396}
Sensors       MICH @311.1 Hz           LO1 @315.17 Hz           
AS55_I       (+0.40+/-6.23)e+07 [0]    (-0.83+/-3.01)e+07 [0]    
AS55_Q       (+1.38+/-0.26)e+09 [0]    (+0.76+/-6.58)e+07 [0]   
BH55_I       (-3.22+/-0.37)e+09 [0]    (-0.81+/-8.42)e+07 [0]    
BH55_Q       (+4.03+/-0.52)e+09 [0]    (-4.01+/-1.05)e+08 [0]    
BH44_I       (-0.06+/-4.22)e+10 [0]    (+0.29+/-4.63)e+10 [0]    
BH44_Q       (-0.03+/-3.21)e+11 [0]    (+0.21+/-3.12)e+11 [0]    
BHDC_DIFF       (-1.07+/-0.39)e+09 [0]    (-3.35+/-7.47)e+07 [0]    
BHDC_SUM       (+2.07+/-0.57)e+08 [0]    (+0.32+/-1.65)e+07 [0]

MICH sensing matrix with MICH locked with AS55_Q and LO_PHASE locked with BH44_Q

Sensing matrix with the following demodulation phases (counts/m)
{'AS55': 2.1, 'REFL55': 76.01784975834194, 'BH55': -63.16236453101908, 'BH44': -39.01036239539396}
Sensors       MICH @311.1 Hz           LO1 @315.17 Hz           
AS55_I       (+0.22+/-5.36)e+07 [0]    (+0.91+/-3.10)e+07 [0]    
AS55_Q       (+1.43+/-0.08)e+09 [0]    (-0.78+/-7.45)e+07 [0]   
BH55_I       (+4.92+/-5.18)e+08 [0]    (-5.20+/-7.93)e+07 [0]    
BH55_Q       (-1.45+/-0.75)e+09 [0]    (+1.76+/-0.59)e+08 [0]    
BH44_I       (+0.01+/-1.14)e+11 [0]    (+0.02+/-1.08)e+11 [0]    
BH44_Q       (+0.03+/-1.95)e+11 [0]    (+0.07+/-1.98)e+11 [0]    
BHDC_DIFF       (+3.05+/-0.17)e+09 [0]    (+1.70+/-2.51)e+07 [0]    
BHDC_SUM       (-2.33+/-0.23)e+08 [0]    (+0.19+/-1.53)e+07 [0]

  - Jupyter notebook: /opt/rtcds/caltech/c1/Git/40m/scripts/CAL/SensingMatrix/ReadSensMat.ipynb

MICH BHD locking:
 - MICH lock with AS55_Q was handed over to BHD_DIFF using following ratio:
C1:LSC-PD_DOF_MTRX_3_4 = 1 (AS55_Q to MICH_A)
C1:LSC-PD_DOF_MTRX_4_34 = -1.34 (BHDC_DIFF to MICH_B, when BH55_Q is used)
C1:LSC-PD_DOF_MTRX_4_34 = 0.47 (BHDC_DIFF to MICH_B, when BH44_Q is used)

MICH BHD noise budget:
 - FM2 of C1:CAL-MICH_CINV was updated to 1/1.4e9 = 7.14e-10 to use measured optical gain.
 - Dark noise was measured at C1:CAL-MICH_W_OUT with PSL shutter closed, PD DOF matrix at various settings for various readout scheme.
 - Attachment #1 shows MICH sensitivity with MICH locked using AS55_Q (green), BHD_DIFF under BH55_Q (blue), BHD_DIFF under BH44_Q (red). BH44 case gives the least noise due to larger optical gain. However, there are excess noise at around 100 Hz, when MICH is locked with BHD_DIFF. The excess noise (bump at around 50 Hz) was similar to what we saw in LO phase noise estimate (40m/17511).
 - At low frequencies below ~30 Hz, the MICH sensitivity is probably limited by seismic noise, as it alignes with FPMI DARM sensitivity (orange curve; measured in 40m/17468).
 - Attachemnt #2 and #3 show estimate of LO phase noise contribution to MICH sensitivity in BH55 case and BH44 case. The coupling was estimated by measuring a transfer fuction from BH55_Q/BH44_Q to MICH_W_OUT. As there was significant coherence in 30-200 Hz region in BH55 case, and 30-100 Hz in BH44 case, transfer function value in that regions was used to estimate the coupling.
 - The coupling was estimated to be the following

 2e-10 m/count for BH55_Q to MICH_W_OUT (0.035 m/m using BH55_Q calibration factor to LO1 motion of 1.76e8 counts/m)
 2e-11 m/count for BH44_Q to MICH_W_OUT

 - Diaggui file: /opt/rtcds/caltech/c1/Git/40m/measurements/LSC/MICH/MICH_Sensitivity_Live.xml

Next:
  - Calibrate BH44_Q to LO1 motion
  - Measure transfer function from LO1 motion to BHD_DIFF under BH44 and BH55
  - Find out the cause of 50 Hz bump in LO phase noise
  - Compare LO phase noise coupling with simulations

[Paco, Yehonathan]

I took over the finesse calculations Yehonathan had set up for BHD. The notebook is here and for this post I focused on simulating what we might expect from our single RF vs dual RF sensors (55 MHz and 44 MHz respectively) in terms of LO phase control.

The configuration is simple, only MICH is included (no ETMs, no PRC, no SRC). The LO phase is changed by scanning LO1, the differential loss is changed by scanning the ITMXHR loss parameter (nominally at 25 ppm), and the microscopic DARM offset is changed by scanning the BS position by +- 6 nm.

Finesse estimates the sensor response by taking the demodulated sideband magnitude (BH55, BH44) with respect to a 1 Hz LO1 signal modulation. This can be done for a set of LO phase angles so as to get the nominal LO phase angle where the response is maximized.

I first replicated the plots from [elog17170] for the two sensors in question. This is just done as a sanity check and is shown in Attachment #1. This plot summarizes our expectation that the single RF sideband sensor should have a peak response to the LO phase around 90 deg away from the nominal BHD readout phase angle (0 deg in this plot). In contrast, the double RF demodulation scheme has a peak response around the nominal LO phase angle.

Attachment #2 looks at a family of similar plots representing differential loss changes between the two MICH arms. We tune this by changing the ITMX loss in finesse, and then repeat the calculation as described above. It seems that for the simple MICH, differential loss of ~ 10000 ppm does not impact the nominal LO phase angle where the responses are maximized for either sensor (note however that the response magnitude maybe changes for single RF sideband sensing at extremely high differential loss).

Finally, and most interestingly Attachment #3 looks at a family of similar plots representing a set of microscopic DARM offsets (+- 6 nm). This is tuned by changing the BS position ever so slightly, and the same calculation is repeated. In this case, the nominal LO phase angle does change, and it changes quite a lot for the single RF demod. It looks like this might be enough to explain how we can sense the LO phase angle with a single RF sideband, but I think the next interesting point would be to simulate the effect of contrast defect by changing the ITM RoCs (to scatter into HOMs) or the non-thermal ITM lenses (to probe the TEM00 contrast defect effect). Any comments / feedback at this point are welcome, as we move forward into other configurations where more serious thermal effects might be introduced (PRMI).

This is interesting. With the FPMI, the DARM phase shift is enhanced by the cavity. Therefore, I suppose the effect on the BH55 is also going to be enhanced (i.e. a much smaller displacement offset causes a similar LO phase rotation).

Yuta pointed out that the BH55 signal was weirdly never going to zero, so I actually tuned the demod angle and made sure I was reading the right (Q) quadrature. This doesn't affect our previous qualitative conclusion about DARM offsets, but here's an updated gif which also makes visualization easier (?).

but what about including the DC reflectivity imbalance of the arms? there would be another BH55 term from that field maybe.

I repeated the calculations but with FPMI (last case was all MICH). The qualitative behavior is the same, the BH55 sensing is mostly affected by residual darm offset. If the darm offset is of a couple of nm, the single RF sideband may sense the LO phase at as much as > 20 deg away from the nominal phase angle. This is not too different from the MICH case; so maybe I overlooked something about how I define FPMI in the calculation.

Attachments #1-3 show the plots of the BH55 (single RF sideband) and BH44 (double RF sideband) sensitivity to LO phase fluctuations around various nominal LO phase angles. Attachment #1 looks at the effect of differential loss, Attachment #2 looks at the effect of differential dc reflectivity (of the ITMs), and Attachment #3 looks at the effect of residual darm offsets. Dashed lines show the orthogonal quadrature (I) of the demodulated RF signals (always minimized).

[Anchal, Yuta]

We have repeated LO phase noise measurement done in elog 40m/17511.
Method we took was the same, but this time, we used (1+G)*[C1:HPC-LO_PHASE_IN1]/[optical gain] to estimate the free-running noise, instead of using [C1:HPC-LO_PHASE_OUT] multiplied by LO1 actuator gain.
We confirmed that both method agrees down to ~ 10 Hz (at lower frequencies, OLTF measurement is not robust; we used interpolated measured OLTF (Attachment #1) for compensation).
Below is the summary of optical gains etc measured today.
Filter gains were adjusted to have UGF of 50 Hz for all.

LO_PHASE lock in ITMX single bounce
        Demod phase  Optical gain     filter gain
BH55_Q  -102.7 deg    6.9e9 counts/m  -0.34
BH44_Q  -5.7 deg     1.3e10 counts/m  -0.17

LO_PHASE lock in MICH
        Demod phase  Optical gain     filter gain
BH55_Q  -72.6 deg    8.7e8 counts/m   -4.4
BH44_Q  -27.6 deg    8.8e8 counts/m   -2.2

LO_PHASE lock in FPMI
        Demod phase  Optical gain     filter gain
BH55_Q  24.2 deg     3.7e9 counts/m   -0.67
BH44_Q  2 deg        5.3e8 counts/m   -4.4  (An order of magnitude smaller than elog 40m/17511)

The values are consistent with elog 40m/17511, except for BH44 in FPMI.
It took sometime to robustly rock LO_PHASE with BH44_Q in FPMI today.
After some alignment, offset tuning and demod phase tuning, it finally worked.
Demod phase of BH44 was tuned to have more DC signal when LO_PHASE was locked with BH55_Q, considering that BH55 and BH44 are orthogonal.
It actually created BH44_I having more amplitude (some noise?) than BH44_Q, but BH44_Q was more coherent to LO_PHASE fringe in BH55_Q.
It might be related to why we are not dark noise limited for BH44_Q, while BH55_Q is dark noise limited in FPMI, and why we cannot lock FPMI BHD with BH44.

First, simple stuff. We estimate the noise budget with total input and output noises. Later, we will break it down (ADC, DAC, whitening, dewhitening noises etc.):

We take the dark noise of AS55, REFL11 and make sure that the whitening and "unwhitening" software filters are on (attachment 1)

To convert cts to Watts we use the values from previous MICH noise budgeting for AS55:

PD_responsivity = 1e3*0.8 #V/W
ADC_TF = 3276.8 #cts/V
demod_gain = 2 #6.77 #According to https://wiki-40m.ligo.caltech.edu/Electronics/LSC_demoddulators
whitening_gain = 10**(24/20) #24 dB

We are not sure why the demod gain was chosen to be 2 and not 6.77 as in the Wiki, maybe it was chosen to match the measurements back then. The demod gain for AS55 was actually measured to be 2.4 in elog 16961.

For now, for lack of time, we use the PD responsivity and demod gain of REFL11 from the wiki:

PD_responsivity = 4.08e3*0.8 #V/W
ADC_TF = 3276.8 #cts/V
demod_gain = 4.74 #According to https://wiki-40m.ligo.caltech.edu/Electronics/LSC_demoddulators
whitening_gain = 10**(18/20) #18 dB

Using the Finesse model for PRMI (should push to git) we calculate the sensing matrix (attachment 3). We turn off the HOMs as it gives us strange results for now.

We take the output noise that was measured at the output of the BS coil driver measured in elog 16960.

Attachment 2 shows the estimated PRMI noise budget. Notice that the dark noise contribution is an order of magnitude better than MICH (elog 16984) due to PRG.

Yuta and I had a discussion last week about the signal strength between the configurations. Here are some naive calculations.
=== Please check the result with a more precise simulation ===

Michelson: Homodyne (HD) phase signal @44MHz is obtained from the combination of LO11xAS55 and LO CAxAS44. SBs at AS rely on the Schnupp asymmetry, the signal is weaker than the one with a single bounce beam from an ITM.

PRMI Carrier resonant:
- Despite the non-resonant condition of the sidebands, the HD phase signal @44MHz is expected to be significantly stronger (~x300) compared with the MI due to the resonance of the carrier and the 44MHz sidebands (the 2nd-order SBs of 11 and 55) in the PRC. Thus, the LO CAxAS44 term dominates the signal.
- The MICH signal @55MHz is enhanced by the resonant carrier by a factor of ~5.5, in spite of the non-resonant 55MHz SBs.
- The MICH signal @BHD is enhanced by the resonant carrier by a factor of ~300. This is the comparable phase sensitivity to PRFPMI case.

PRMI Sideband resonant:
- Despite the non-resonant condition of the carrier, the HD phase signal @44MHz is expected to be even stronger (~x400) compared with the MI due to the resonance of the 11MHz and 55MHz sidebands in the PRC. Thus, the LO11xAS55 term dominates the signal.
- The level of the MICH signal @55MHz is expected to be comparable to the one with PRMI carrier resonant as the resonant condition for the CA and 55MHz SBs are interchanged.
- The MICH signal @BHD is expected to be negligibly small due to non-resonance of the carrier.

PRFPMI: Now the carrier and the 11 and 55MHz sidebands are resonant.
- The HD phase signal @44MHz is expected to be the same level as the SB resonant PRMI, and the LO11xAS55 term dominates the signal.
- The level of the MICH sensitivity @AS 55MHz shows x300 of the MICH signal of the MI and x50 of the MICH with PRMI.
- The MICH signal @BHD is going to be the same level as the one with PRMI Carrier resonant.
- The DARM signal shows up at the dark port signal enhanced by x300 from the MICH level due to the finesse of the arms.

Simple assumptions
1) PRM has a transmission of T_PRM = 0.05
2) PRG is limited by the transmission of PR2 (T_PR2=0.02 per bounce).
    If the IFO is lossless, PRG is 25 (i.e. theoretical maximum). In reality, the IFO loss is 2~3% -> PRG is ~15.
    The asymmetry of 30mm has a negligible effect.
3) For the anti-resonant fields, APRG is ~T_PRM/4 = 0.0125
4) Arm finesse is 450. Therefore the phase enhancement factor N is ~300.
5) Modulation depth is ~0.1. J0=1, J1=0.05, J2=0.00125
6) Sideband leakage by the asymmetry is ɑ=l_asym wm / c = 0.008 for 11MHz and 5ɑ for 55MHz.

Single Bounce

The numbers are power transmission to each port
   Carrier              11MHz                    55MHz
LO TPRM TPR2 = 1.0e-3   J1^2 TPRM TPR2 = 2.5e-6  J1^2 TPRM TPR2 = 2.5e-6
AS TPRM/4    = 1.3e-2   J1^2 TPRM/4    = 3.1e-5  J1^2 TPRM/4    = 3.1e-5

LO phase @44MHz: LO 11 x AS 55 = Sqrt(2.5e-6 * 3.1e-5) = 8.8e-6

Michelson

   Carrier               11MHz                     55MHz                      44MHz
LO TPRM TPR2 = 1.0e-3    J1^2 TPRM TPR2 = 2.5e-6   J1^2 TPRM TPR2   = 2.5e-6
AS TPRM ε^2  = 0.05 ε^2  ɑ^2 J1^2 TPRM  = 8.0e-9   25 ɑ^2 J1^2 TPRM = 2.0e-7  16 ɑ^2 J1^4 TPRM = 3.2e-10

LO phase @44MHz: LO 11 x AS 55 = Sqrt(2.5e-6 * 2.0e-7)  = 7.1e-7
                 LO CA x AS 44 = Sqrt(1.0e-3 * 3.2e-10) = 5.7e-7

AS MICH  @55MHz: AS CA x AS 55 = Sqrt(0.05 * 2.0e-7) ε  = 1.0e-4 ε
AS MICH  @BHD:　 LO CA x AS CA = Sqrt(1.0e-3 * 0.05) ε  = 7.1e-3 ε

PRMI (Carrier Resonant)

   Carrier              11MHz                     55MHz                      44MHz
LO PRG TPR2 = 0.3       J1^2 APRG TPR2 = 2.5e-7   J1^2 APRG TPR2   = 2.5e-7  J1^4 PRG TPR2 = 1.9e-6
AS PRG ε^2  = 15 ε^2    ɑ^2 J1^2 APRG  = 8.0e-10  25 ɑ^2 J1^2 APRG = 2.0e-8  16 ɑ^2 J1^4 PRG = 9.6e-8

LO phase @44MHz: LO 11 x AS 55 = Sqrt(2.5e-7 * 2.0e-8)  = 7.1e-8
                 LO CA x AS 44 = Sqrt(0.3 * 9.6e-8)     = 1.7e-4
AS MICH  @55MHz: AS CA x AS 55 = Sqrt(15 * 2.0e-8) ε    = 5.5e-4 ε
AS MICH  @BHD:　 LO CA x AS CA = Sqrt(0.3 * 15) ε       = 2.1 ε
 

PRMI (Sideband Resonant)

   Carrier               11MHz                     55MHz                      44MHz
LO APRG TPR2 = 1e-4      J1^2 PRG TPR2 = 7.5e-4    J1^2 PRG TPR2   = 7.5e-4  J1^4 APRG TPR2 = 6.3e-10
AS APRG ε^2  = 5e-3 ε^2  ɑ^2 J1^2 PRG  = 2.4e-6    25 ɑ^2 J1^2 PRG = 6.0e-5  16 ɑ^2 J1^4 APRG = 3.2e-11

LO phase @44MHz: LO 11 x AS 55 = Sqrt(7.5e-4 * 6.0e-5)  = 2.1e-4
                 LO CA x AS 44 = Sqrt(1e-4 * 3.2e-11)   = 5.7e-8
AS MICH  @55MHz: AS CA x AS 55 = Sqrt(5e-3 * 6.0e-5) ε  = 5.5e-4 ε
AS MICH  @BHD:　 LO CA x AS CA = Sqrt(1e-4 * 5e-3) ε    = 7.1e-4 ε

PRFPMI

   Carrier             11MHz                     55MHz                      44MHz
LO PRG TPR2 = 0.3      J1^2 PRG TPR2 = 7.5e-4    J1^2 PRG TPR2   = 7.5e-4   J1^4 APRG TPR2 = 6.3e-10
AS PRG ε^2  = 15 ε^2   ɑ^2 J1^2 PRG  = 2.4e-6    25 ɑ^2 J1^2 PRG = 6.0e-5   16 ɑ^2 J1^4 APRG = 3.2e-11

LO phase @44MHz: LO 11 x AS 55 = Sqrt(7.5e-4 * 6.0e-5)  = 2.1e-4
                 LO CA x AS 44 = Sqrt(0.3 * 3.2e-11)    = 3.1e-6
AS MICH  @55MHz: AS CA x AS 55 = Sqrt(15 * 6.0e-5) ε    = 3.0e-2 ε  ==> DARM@55MHz 9.0 ε
AS MICH  @BHD:　 LO CA x AS CA = Sqrt(0.3 * 15) ε       = 2.1 ε     ==> DARM@BHD   6.3e2 ε

{Mayank, Paco, Yehonathan}

Dewhitening noise curves were taken using SR785+SR560 for the PRMI noise budget. One representative channel was measured at each board, suspensions were tripped before work was done. The input pins to the dewhitening boards were shorted using an exposed ribbon cable.

At each board, the measurement was taken with and without dewhitening filter on. The toggling of the dewhitening filter was done by turning on and off the SimDW filters at the coil filter screen of each suspension.

Attachment 1 summarizes the results.

ITMX dewhitening noise is much higher than the rest.

ITMY measurement turned out to be bogus since we mostly measured dark noise. The reason we made the gain so low in that measurement is that it was saturating the SR560 whenever we used gain>1.

We measured the IQ demodulation board gains for REFL11 and AS55.

To do this, we replaced the PD input on the demod board with an RF signal at near the nominal frequencies of 11.066195 MHz and 55.330975 MHz using a Marconi 2024A identical to the one which sources the PM sidebands in our PSL. Even though we matched the modulation frequencies we found the two marconis were in practice offset by ~ 3 Hz. After tuning the frequency around a bit, we managed to get them to within 450 mHz.

REFL11

We started with REFL11 IQ demod board. After sourcing 11.066198 MHz into the PD input port, we took the I and Q outputs and looked at them using an osciloscope. We measured the Vpp levels on both as well as the Marconi output. The resulting levels were

• Source = 44.4 mVpp, I = 8.8 mVpp and Q = 10.8 mVpp ==> Gains are therefore 0.19 and 0.24. The amplitude gain of this board is sqrt(0.19 ** 2 + 0.24 ** 2) = 0.153. This is in stark disagreement with the wiki. Has the wiki finally failed us?

AS55

We then moved on to AS55 IQ demod board. After sourcing 55.330975 MHz* into the PD input port, we took the I and Q outputs and looked at them using an osciloscope. We measured the Vpp levels on both as well as the Marconi output. The resulting levels were

• Source = 16.8 mVpp, I = 50.8 mVpp and Q = 56.3 mVpp ==> Gains are therefore estimated to be 3.3 and 3.7. The amplitude gain of this board is sqrt(3.3**2 + 3.7**2) = 4.74. This is in slight contrast with the previously measured gain of 2.8, but we think a factor of 2 may have been misplaced in either calculation since one typically estimates AMP = 2 * sqrt(I**2 + Q**2).

* Note that in the second test, we didn't match up the frequency, which caused I and Q outputs to have significant gains (instead of just I).

I'm checking Koji + Yuta's not-so-naive calculations using finesse.

PRMI Carrier resonant:
- The HD phase signal @44 MHz is estimated to be 386.5 times stronger.
- The MICH signal @BHD_DIFF is estimated to be enhanced by a factor of 299.75.

PRMI Sideband resonant:
- The HD phase signal @44MHz is estimated to be () stronger.
- The MICH signal @BHD_DIFF is estimated to be suppressed by a factor of

PRFPMI: Now the carrier and the 11 and 55MHz sidebands are resonant.
- The HD phase signal @44MHz is estimated to be ().
- The MICH signal @BHD_DIFF is estimated to be the same level as the one with PRMI Carrier resonant.

I included the output electronic noises into the PRMI carrier noise budget (attachment 1).

The coil driver noise was calculated using the Johnson noises of the coil driver resistor:

PRM 430 ohm

BS 100 ohm

ITMX/Y 400 ohm

For the dewhitening noises I use the measurements from yesterday. As expected fro yesterday's measurements, the ITMX dewhitening noise is dominating. For the coil driver gain I use the recently measured actuation calibration (elog   17522 ) to extract it. I find that these gain values:

PRM 1.009

BS 1.333

ITMX/Y 0.24

For the DAC noise I assume 1uV/sqrtHz and use the simDW filters from the coil outputs MEDM screens as the DW filters TFs.

Next:

1. Break down input noises.

2. Measure how much light is reaching REFL11 to correct the sensing matrix and get the right shot noise.

{Mayank, Yehonathan}

Yesterday, we measured AS55 and REFL11 dark noises at the IQ demod boards outputs (attachment 1) using SR560+SR785 setup.

We also measured the whitening board noise of REFL11 using an improvised adapter (picture will come later). The measurement result is shown in attachment 2. Didn't have time to measure the whitening noise for AS55

Also, after realizing the Finesse model doesn't account for the REFL port attenuation I measure how much DC power at the REFL11 PD to be 0.8mW by aligning PRM and misaligning the rest of the optics.

For some reason, the power before the attenuation is only ~ 360mW. The Finesse model predicts around 700mW. Where is the rest of the light going?

I added the PDs Dark noises (using the recently measured IQ demod gains) and shot noise to the PRMI carrier noise budget (attachment 3). ADC and whitening noises coming soon.

Measurements and PRMI noise budget notebooks were uploaded to the 40m git.

{Paco, Yehonathan, Yuta}

Paco and Yuta locked PRMI carrier and I took the MICH OLTF measurement (attachment 1).

I downloaded 300secs of C1:LSC-MICH_IN1_DQ from when the PRMI was locked yesterday and calibrated it with the OLTF. I plot it together with the noise budget (attachment 2).

{Mayank, Yehonathan}

We measured the noise at the WF1 (REFL11) and WF2 (AS55) boards at the LSC rack with and without whitening filter. We switch the filter on and off by switching off and on the unwhitening in the PDs filter bank.

Attachment 1 shows the measurements.

Attachment 2 shows the ratio between the noise with and without whitening filter. I also plot the inverse of the unwhitening MEDM filter (all the unWhite filters were the same). I tune the gain of that filter to match the ratio of the AS55 whitening noises.

This is because I couldn't match the ratio of the REFL11 noises.

Moreover, the overall gain doesn't make sense to me. AS55 whitening has a gain of 24db and REFL11 has a gain of 18db. I'm not entirely sure where these values should show up. Also seems like REFL11 whitening has more gain than AS55 whitening. Will have to investigate more tomorrow.

{Mayank, Yehonathan}

We measured today the TFs of the whitening boards. We measured in particular REFL11 I/Q and AS55 I/Q channels using SR785.

There seems to be an issue with turning on whitening gain bigger than 18dB. In all our measurements, when the whitening filter was off the TF was flat and had the right gain. However, when we turned the whitening on, the measured TFs for gains higher than 18 dB would like exactly like as if the whitening gain was 18 dB. This happened in all channels that were measured and across two separate whitening filter boards.

Also, it was hard to measure both low and high-frequency parts of the TFs when the gain was high. The gain difference should be normally 40 dB but for higher gains it seems smaller. We verified that at higher gain level the high-frequency response was dependant on the ecxitation level meaning we had some saturation there.

We forgot to take a reference TF measurement by looping the SR785 on itself using the same BNC cables used for the actual measurement. I took this measurement today (attachment 1). As can be seen, there is a significant delay in the SR785 + cables themselves.

I also retook some measurements on the AS5_I whitening channel for various gains. Being careful with the excitation level and the channel range on the SR785 to avoid saturation I was also able to see low-frequency gains higher than 18dB so that problem is gone too. The results are shown in attachment 2 with the reference phase subtracted from the measurements.

I added input noises and angle to length coupling to the noise budget.

I added ADC and whitening filters noise contributions. The ADC noise is assumed to be 1uV/sqrtHz and the whitening noises were measured before in elog 17582. I use the measured whitening filter (elog   17584 ) to get the signal referred noise and calibrate.

The angle-to-length coupling is computed by taking the suppressed OpLev noise spectra of ITMX, ITMY, and BS and converting them to length noise by using the recently measured coupling coefficients in ELOG   17583 

Is the A2L coming from the optical lever feedback? If so, we can make a 30 Hz ELP to cut it off by 60 Hz.

We lowered the BH44 and BH55 DC transimpedances to ~ 50 V/A

[Paco, Yuta]

Background

When locking the homodyne phase angle using BH44_Q or BH55_Q error signals, we notice the orthogonal quadrature (BH44_I, BH55_I) sometimes appears too noisy. The origin of this useless signal is not known, but we have recently attenuated these beams by placing ND filters before the two RFPDs to avoid saturation effects which become obvious when we lock PRMI. We decided to investigate further by the following tests:

• Remove ND filters and lower the DC transimpedances to ~ 50 V/A
• Check for scattering from suspended optics, e.g. by injecting a line at ITM PIT/YAW and look at the BH44/BH55 demodulated spectra
• Check for PRCL sensing by BH44/BH55, e.g. by measuring the transfer function and/or running a simulation in finesse.
• Check the RF spectra for the signals entering the IQ demod boards (including the 44 and 55 MHz LOs).

DC transimpedance modifications

The first thing we did was change the DC transimpedances of both RFPDs. After removing them from the table, we checked the schematics for 40m RFPDs on the wiki. The DC transimpedance for these gold RFPDs (D980454-v1-C) is estimated as (R22*(1+R13/R23), where these resistors are located around the follower and non-inverting amplifier stages along the DC output traces. After opening the two RFPDs and taking photos of the circuits before any changes (Attachment #1-2), we estimated the DC transimpedances from the measured values for R22, R23 and R13 and summarized them below:

The changes were made on R13 (photos in Attachments #2-3) and the final values summarized below:

All changes have been summarized and recorded in the wiki. The ND filters were set to 0.04 (minimum attenuation) and RFPDs reinstalled.

Next steps:

Continue investigating these items:

• Remove ND filters and lower the DC transimpedances to ~ 50 V/A
• Check for scattering from suspended optics, e.g. by injecting a line at ITM PIT/YAW and look at the BH44/BH55 demodulated spectra
• Check for PRCL sensing by BH44/BH55, e.g. by measuring the transfer function and/or running a simulation in finesse.
• Check the RF spectra for the signals entering the IQ demod boards (including the 44 and 55 MHz LOs).

BH44 is sensitive to PRC alignment noise

[Paco, Yuta]

We investigated the content of BH44 demodulated signals under PRMI configuration. We had a few ideas of what was being sensed by BH44_I but we wanted to test this. Attachment #1 shows a timeseries screenshot of the DCPDs and BH44 error signals during PRMI lock stretch. It is pretty clear how BH44_I is sensing the same as REFLDC. To understand what REFLDC is sensitive to, we locked PRY (this is like having a lossy PRC) and looked at REFLDC, and BH44 error signals again. When PRY is aligned nicely, BH44 error signals show clean LO fringes and we could lock LO_PHASE stably (Attachment #2). Dithering the PRM YAW at 0.5 Hz (amplitude of 150 counts) is sensed by the REFLDC output, so we can attribute its fluctuations to the PRC misalignment (Attachment #3). Now we saw that the zero crossing of the homodyne phase angle changes following REFLDC, and LO_PHASE could not be locked stably. These suggest that alignment of PRC is sensed by BH44, and we might need alignment control to stably lock LO_PHASE in PRMI.
To get the idea of what is causing alignment fluctuations of PRC, we checked the spectrum of SUSPIT/YAW of PRM, PR2, PR3, BS, ITMX, and ITMY. It was not clear what is causing REFLDC fluctuations. (But we found that ITMX and ITMY has huge bounce mode at 16.2 Hz; see Attachment #4).

Next:
 - Check FINESSE to see what BH44 sees. PRCL? PRG?
 - Commission REFL WFS for alignment control of PRC?
 - Commission dither loops (add option to demodulate PRCL, modulate PR2 and PR3) for alignment control of PRC?
 - Check RF spectrum of BH44 and RF LO for 44 MHz (sidebands other than 44 MHz might be contaminating the signal).
 - Check ITMY scattering. Dither ITMY in YAW and check BH44.
 - Move on to PRMI sideband BHD

I calculated the sensing matrix for PRMI carrier using the Finesse model (git updated) using MAXTEM=2. PRG is calculated to be 11.14, consistent with observations.

LO Phase is chosen such that it maximizes MICH signal on BHD_DIFF.

The RFPDs were assumed to have a demodulation angle that maximizes the signal they are intended to sense (using MAXTEM=2).

That is,

BH44/55 maximized for HPC

REFL11 maximized for PRCL

AS55 maximized for MICH

Values are in uW/nm

50.00

73.67

0.68

17.28

440.80

3.81e-14

0.38

4.22

0.4

3.17

13.46

Some interesting numbers here. First, BHD_SUM is sensitive to MICH. It's not surprising because PRM reflects the MICH signal into POP.

Also interesting, BHD_SUM is super sensitivef to PRCL. Much more than REFL11. We can use it to enhance the PRCL lock.

Unfortunately, although BH44/55 are sensitive to HPC (LO phase), they are swamped by MICH and PRCL. This issue needs to be addressed in order to gain robust LO Phase locking in PRMI.

it is great to see a sensing matrix without 900 digits of precision!

for choosing what sensor to use, we don't necessarily care about W/m, but more like the equivalent noise of the sensor in m/rtHz, taking into account the real noise floor. In that case, we would possibly rotate the demod phase to maximize the SNR rather than maximize the S or minimize the N.

[Yuta, Hiroki]

BHD alignment has been restored

We aligned the AS beam (reflection of ITMX) and LO beam and maximized the fringing of the BHD differential signal (Attachment 1).
We used LO1, SR2 and AS4 for the alignment and the result parameters are shown in Attachment 2.
 

Procedure log of BHD alignment

Alignment of LO beam:

• ETMY, ETMX and ITMY were misalinged during this BHD alignment
• Misaligned SR2 to have only the LO beam in BHD DCPDs
• Tuned LO1 so that the LO beam comes to the previous position in the video

Alignment of AS beam:

• Restored SR2
• Tuned AS4 so that the AS beam comes to the position of LO beam
• Repeated the followings for the pitch and the yaw until the fringing was maximized:
  • Misaligned AS beam using SR2
  • Restore the alignment of AS beam using AS4
  • If the fringing gets worse, it means that you moved SR2 in the wrong direction. Move the SR2 in the opposite direction next and repeat the procedures above.
  • If the fringing gets better, it means that you moved SR2 in the correct direction. Continue the procedures above.

Dean made the OMC cables for the BHD Platform. They are going through the C&B process.

From left to right: QPD cable, PZT cable with Picomotor etc, DCPD cables

Currently, I'm constructing the mode-matching telescope on a breadboard for the alignment of two OMCs for BHD.

What I did on July 25th:

• Constructed the input optical system before the fiber collimator (Attachment  1).
• Measured the mode-matching efficiency to the fiber collimator: 0.878 +/- 0.006

Next:

• Fix the end of the fiber on the mode-matching breadboard.
• Arrange two lenses to match the output beam with the OMC eigen mode (Horizontal: 489.6 um, Vertical: 490.5 um)

I almost finished constructing the mode-matching breadboard today.

What I did on July 26th:

• Constructed the mode-matching telescope on a breadboard (Attachment  1).
  • Height of optical axis: 3.5" (from the bottom of the breadboard)
  • Used two lenses of f = 200 mm @ ~6 cm and ~24 cm from the surface of the collimator mount
    (Used JamMt to obtain the solution)
  • Tuned the position of the two lenses monitoring the resulting beam waist so that the waist becomes ~ 500 um
    (OMC eigen mode: 489.6 um (horizontal), 490.5 um (veritical))

Next:

• Measure the dimensions of the resulting mode-matching telescope
• Measure the beam profile of the resulting beam
• Estimate the upper limit of the mode-matching efficiency to the OMC

Supplementary

Output beam profile from collimator:
I measured the beam profile of the output from the collimator (Attachment 2).

Beam waist X: 44.3 +/- 0.5 um
Beam waist Y: 46.1 +/- 0.6 um
Waist position X (from the front surface of the collimator mount): 1.64 +/- 0.04 cm
Waist position Y (from the front surface of the collimator mount): 1.70 +/- 0.05 cm

I found that the resulting beam is not collimated but focused.
The model number is not written on the collimator so I was not able to find its specification.

BHD Platform has been washed and set to dry over the weekend
 

Yesterday, I began to wipe down the BHD platform in the C&B lab. While I was cleaning, I wanted to be very particulate of how clean it was here because of how long it was sitting in Don’s office. There was dust that was built up in the crevasses of the underside of the platform and it was be difficult to get out. I spent ~30 min clean and I felt like I was better off washing the entire piece. 

I contacted Maty and this morning we decided to place the platform inside the Machine Washer. We had to remove the 2 basket and shelves in order to fit the Platform inside. We decided to stand the platform vertically as shown in Attachment #1 and hold in place using Zipties. After a few soft tugs on the platform, Matt and I agreed that it was pretty secure. We started the parts washer and ran it for 10 minutes. The Machine washer was set to use water at 120°F. 

After the machine washer was done, Maty removed the platform from the machine and we were rinsed of using water. 

What’s next ? 

We agreed to leave the platform to dry over the weekend. This will set us back a day, but we did not want to insert to platform to be baked while it was still wet. We left the platform under the vent to dry over the weekend.
 

[Maty, JC]
This morning, Maty and I proceeded to work on the BHD Platform work. We rinsed off the BHD platform one more time with water and placed onto the work bench. Here, we wiped down using IPA and removed a ton of the metal shavings from the platform. (Attachment #2 & #3 ) The threads in the platform had a ton of shaving, but we were able to get majority, if not all, off the part. 

Next, we garbed up and prepared the large tub for the BHD Platform. The tub contained 107 L of water, so we added 1000 mL of Liquinox into the tub. The Platform was a bit too wide, so we had to put it in at an angle. We used a beaker to pour the solution over all of the part. Along with this, we used the Sonicating probe to clean the threads. ( Attachment #3 ) We did this for the entire part and then drained the tub. 

I lifted and placed the platform into the blue machine parts washer to rinse off the part. We went thread by thread to make sure all of the Liquinox was rinsed off of the part, this to ~30 minutes. ( Attachment #4 )After rinsing the part off, we brought it over to the clean room flow bench. Here, Maty placed large clothes under to platform. She then went through each hole of the platform with the TopGun to remove the loose excess water. ( Attachement #5 )

We have stopped here and are leaving the BHD Platform to dry overnight. We will insert the BHD Platform into the Large Air Bake Oven TOMORROW! We plan to have it ready by the end of the week. 

[JC]
I began to prepare the temporary clean room where we will be putting together the BHD Platform Assembly - D2100085. I wiped down the table with the Acetone and IPA multiple times. After this, I used Mylar sheets to use as the cleanroom walls. I slit the sheet into separate pieces to allow easy access and held these sheets up by using Kapton tape. Next, I connected the HEPA booth to an extension cord which was plugged into the Wall outlet of 1X3. 

Mode-matching breadboard has been constructed

I have constructed the mode-matching breadboard for aligning the BHD OMCs.
I also measured the profile of the resulting beam:

Beam waist X: 496 +/- 2 um
Beam waist Y: 504 +/- 3 um
Waist position X (from the front surface of the collimator mount): 224 +/- 1 cm
Waist position Y (from the front surface of the collimator mount): 236 +/- 2 cm
Maximum mode-matching efficiency to OMC: 99.78 +/- 0.07 % (if the waist of the OMC eigen mode is place at 230 cm)
(OMC eigen mode: 489.6 um (horizontal), 490.5 um (veritical))

The details are shown in the following attachments:

• Attachment 1: Current configuration and summary of beam profile
• Attachment 2: Resulting beam profile
• Attachment 3: Photo of mode-matching bread board

BHD platform is estimated to stop baking by the end of the day. We are following the LIGO Procedure and baking at 150°C By tomorrow morning, I will have the platform out of the Air Bake Oven.

· I have began to prepare the smaller components for the assembly for baking. I am trying to get all of these components cleaned and together by the end of next week.

OMC Interface Aligner - (It's upside down...)

BHD OFI arrangement

A bit improved the design of OMC Interface Aligner

The idea is...The OMC I/F aligner covers the OMC for aligning the kinematic mounts (3 pairs of a V-groove and a ball) on the OMC. This makes the kinematic mount of two OMCs identical.

However, the OMC kinematic mount can't be adjusted because all the fasteners of the kinematic mounts are hidden by their counterparts.
We can copy the alignment of the OMC to the aligner, but the opposite is not possible.

Ready / Soon Ready
- BHD OMC Cables ready
- OMC#1 / OMC#4 ready
- BHD Platform parts being cleaned
- Assembly area HEPA being built
=> We will be soon ready to assemble BHD Platform and test with the OMC

In progress
- OMC locking setup (Moku)
- Connectors being attached to the BHD Platform actuators (picos & rotation stage)
- BHD Platform OFI parts drawing/procurement

- 40m BHD Electronics (BHD Adapter / DCPD TIA / Actuator driver I/F)

Other vent items
- In-vac ribbon cable holder (JC)

- Connector holder

- Scattered light control

- Pre-vent work
    * ASS recovery / extension

    * ETMX tuning
    * Vertex Eletronics upgrade
    * Fix PZT amps / PZT
    * Acromag

- Vent work items
    * New PR2
    * Alignment
 

I went to the Solidworks model of the ITMY invac table and checked where the center of mass is.

The center of mass (COM) of the loaded items is at (+34.0mm, +8.8mm) from the center of the table. (BTW, it is above the table by 115mm). The total mass is 86.2kg.

To bring the COM at the center of the table, we need to place the balance mass(es).
An example solution is shown in the attachment.
It involves two masses (8kg and 2.5kg).
The 8kg one is fixed on the BHD platform behind AS3. The other one will be placed on the table, well away from everything.

The optical fiber that Hiroki set up turned out that long enough for the use in the new HEPA optical bench. (Attachment 1)

The fiber was rolled and placed beneath the PSL table. The end is capped. (Attachment 2)

The BHD OFI path consists of a few optical components: (Attachment 1)

1. HWP ROtator
2. Thin Film Polarizer (x2)
3. Faraday Rotator
4. HWP (fixed) Mount
5. Beam Dump (x2)

They use custom length posts because they have different vertical sizes and also, the beam elevation on the platform is nonstandard (4.2inch).

We are going to make these posts using Newport's 4inch pedestal posts, which we already have.
The posts are machined to have the proper lengths. Additionally, we want to make venting holes and gaps to keep the gas escapting paths.

The 3D models of the posts can be found as D2300352_1inch_post.SLDPRT in the SW Vault (C:\llpdmpro\RnD\40m\BHD Platform\Beam dump) - Attachment 2
This model has four configurations that corresponds to the posts for 1/2/4/5 of the above.

Config: -01 beam dump (Attachment 3)
QTY 5 (2 + 3 spares)
The total height to be shortened to be 3.2inch.

Config: -02 Thin Film Polarizer  (Attachment 5)
QTY 3 (2 + 1 spare)
The total height to be shortened to be 3.3125inch

Config: -03 HWP Rotator  (Attachment 6)
QTY 1
The total height to be shortened to be 3.413inch.

Config: -04 HWP Mount  (Attachment 7)
QTY 1
The total height to be shortened to be 3.075inch.
The top part is made thinner (D=0.7") to fit with the recess on the mount. The height of this part is 0.15".

General Remarks:

• The top 8-32 hole should be re-drilled and re-tapped. The depth should be ~0.5inch.
• The upper and lower surfaces have the vent grooves. The original hole may interfere with the new top surface??? Please check. It's OK if we can mount the beam dump head on it.
• Appropriate vent holes must be made from the side.
• The top and bottom surfaces may need to be filed (deburred) to recover the flat contact of the top/bottom surfaces.
• Appropriate chamfer should be applied to remove the sharp edges.

Not a fanatic of the "-02 TFP -3.3125 [ D2300352_1inch_post ]" configuration. The previous hole that comes with the original design could leave some semi-sharp edges. (Even if we ask the machinist the break them down).

Koji recommended to check the if we can modify the 6in Pedestal and avoid the issue of the hole. I went ahead and made these modifications and uploaded the new sketch to the [D2300352](https://dcc.ligo.org/cgi-bin/private/DocDB/ShowDocument?.submit=Identifier&docid=D2300352&version=) as D2300352_1inch_post_V2.SLDPRT

I still have a couple of modifications to make such as add vent groove at the top.