[Paco, Koji, Yuta]

We managed to lock MICH using REFL55_Q by setting the demodulation phases and offsets right.
The following is the current FPMI locking configuration we achieved so far.

DARM: POX11_I / gain 0.007 / 0.5*ETMX-0.5*ETMY (or 1*ETMX) / UGF of ~100 Hz
CARM: POY11_I / gain 0.018 / 1*MC2 / UGF of ~200 Hz
MICH: REFL55_Q / gain -10 / 0.5*BS / UGF of ~30 Hz

Transitioning DARM error signal from POX11_I to 0.5*POX11_I+0.5*POY11_I was possible with FM4 filter off in DARM filter bank, but not to AS55_Q yet.

REFL55 and AS55 demodulation phase tuning:
 - We found that both AS55 and REFL55 are contaminated by large non-MICH signal, by making a ASDC vs RF plot (see 40m/16929).
 - After both arms are locked with POX and POY, MICH was locked with AS55_Q. ASDC was minimized by putting an offset to MICH filter.
 - With this, REFL55 offsets were zeroed and demodulation phase was tuned to minimize REFL55_Q.
 - Locked MICH with REFL55_Q, and did the same thing for AS55_Q.
 - Resulting ASDC vs RF plots were attached. REFL55_Q now looks great, but REFL55_I and AS55 are noisy (due to signals from the arms?).

Jupyter notebook: https://git.ligo.org/40m/scripts/-/blob/main/CAL/MICH/MICHOpticalGainCalibration.ipynb

Sensing matrix:
 - With FPMI locked using POX/POY, DARM and CARM lines were injected at around 300 Hz to measure the sensing gains. For line injection, C1:CAL-SENSMAT was used, but for the demodulation we used a script. The following is the result.

 Sensors              DARM (ETMX)         CARM (MC2)        
C1:LSC-AS55_I_ERR    3.10e+00 (-34.1143 deg)    1.09e+01 (-14.907 deg)    
C1:LSC-AS55_Q_ERR    9.96e-01 (-33.9848 deg)    3.30e+00 (-27.9468 deg)    
C1:LSC-REFL55_I_ERR    6.75e+00 (-33.7723 deg)    2.92e+01 (-34.0958 deg)    
C1:LSC-REFL55_Q_ERR    7.07e-01 (-33.4296 deg)    3.08e+00 (-33.4437 deg)    
C1:LSC-POX11_I_ERR    3.97e+00 (-33.9164 deg)    1.51e+01 (-30.7586 deg)    
C1:LSC-POY11_I_ERR    6.25e-02 (-20.3946 deg)    3.59e+00 (38.4207 deg)

Jupyter notebook: https://git.ligo.org/40m/scripts/-/blob/main/CAL/SensingMatrix/MeasureSensMat.ipynb

 - By taking the ratios of POX11_I and AS55_Q for DARM, POY11_I and REFL55_I for CARM, we tried to find the correct gains for REFL55 and AS55 for DARM and CARM. x3.96 more gain for AS55_Q than POX11_I and x0.123 less gain for REFL55_I than POY11_I.

Next:
 - Try locking the arms with no triggering, and then try locking FPMI with REFL/AS without triggering. No FM4 for this, since FM4 kills gain margin.
 - Lock single arm with AS55_Q and make a noise budget. Make sure to misalign ITMX(Y) completely when locking Y(X)arm.
 - Lock single arm with REFL55_I and make a noise budget.
 - Repeat Xarm noise budget with Yarm locked with POY11_I and MC2 (40m/16975).
 - Check IMC to reduce frequency noise (40m/17001)

There was a EQ in Ridgecrest (approximately 200 km north of Caltech). It was around 6:20 PM local time.

All the suspensions tripped. I have recovered them (after some struggle with the weird profusion of multiple conflicting scripts/ directories that have appeared in the recent past...)

ETMY is still giving me some trouble. Maybe because of the HUGE bias on that within the fast CDS system, it had some trouble damping. Also the 'reenable watchdog' script in one of the many scripts directories seems to do a bad job. It re-enables optics, btu doesn't make sure that the beams are on the optical lever QPD, and so the OL servo can smash the optic around. This is not good.

Also what's up with the bashrc.d/ in some workstations and not others? Was there something wrong with the .bashrc files we had for the past 15 years? I will revert them unless someone puts in an elog with some justification for this "upgrade".

This new SUS screen is coming along well, but some of the fields are white. Are they omitted or is there something non-functional in the CDS? Also, the PD variances should not be in the line between the servo outputs and the coil. It may mislead people into thinking that the variances are of the coils. Instead, they should be placed elsewhere as we had it in the old screens.

It looks like Tomislav's measurements of the WFS demod board noise were actually of the cable that goes from the whitening to the ADC. So the huge low frequency excess that he saw is not due to wind, but just the inverse whitening of the digital system?

In any case, today, I looked at the connections from the Whitening to the ADC. It goes through an interface chassis to go from ribbon to SCSI. The D-Sub connectors there have the common problem in many of the LIGO D-sub connectors: namely that the strain relief nuts are too tall and so the D connector doesn't seat firmly - its always about to fall out. JC, can you please take a look at this and order a set of low profile nuts so that we can rework this chassis? Its the one between the WFS whitening and the SCSI cables which go to the ADCs.

After pushing them in, I confirmed that the WFS are working, by moving all 6 DoF of the MC mirrors via bias slider, and looking at the step responses (attached). As you can see, all sensors see all mirrors, even if they are noisy.

Next up: get a breakout for the demod output connector and measure the noise there.

For today, I aligned the IMC by hand, then centred the WFS beams by unlocking the IMC and aligning the bright beam. I noticed that the WFS1 beam was being dumped randomly, so I angled the WFS1 by ~3 deg and dumped the specular reflection on a razor blade dump. To handle the sign change in the MC1 actuation (?), I changed the sign in the MC1 ASC filter banks. MCWFS loops still nto closing, but they respond to mirror alignment.

Of the 7 Sorenson Power Supplies I tested, 5 are working fine, 1 cannot output voltage more than 20 Volts before shorting, and other does not output current. Six Sorensons are behind the X-Arm.

I have temporarily abandoned vectfit and aaa since I've been pretty unsuccessful with them and I don't need poles/zeroes to find the unity gain frequency. Instead I'm just fitting the transfer function linearly (on a log-log scale). I've found the UGF at about 5.5 kHz right now, using old data - next step is to get the Red Pitaya working so I can take data with that. Also need to move this code from matlab to python. Uncertainty's propagated using the 95% confidence bounds given by the fit, using curvefit - so just from the standard error, and all points are weighted equally. Ideally would like to propagate uncertainty accounting for the coherence data too, but haven't figured out how to do that correctly yet.

[UPDATE 7/22/2022: added raw data files]

[Yuta, Paco]

• We first zero the offsets in ASDC, AS55, REFL55, POX11, and POY11 when PSL shutter is closed.
  • After this, we checked the offsets with only ITMX aligned. Some of RFPDs had ~2 counts of offsets, which indicate some RFAM of sidebands, but we decided not to tune Marconi frequencies since the offsets were small enough.
• We went over the demod phases for AS55, REFL55, POX11, and POY11.
  • For POX11/POY11 first we just minimized the Q in each locked XARM/YARM individually. The newfound values were
    • C1:LSC-POX11_PHASE_R = 106.991
    • C1:LSC-POY11_PHASE_R = -12.820
  • Then we misaligned the XARM by getting rid of the MICH fringe in the ASDC port with ITMX yaw offset, and locked YARM using AS55_Q and REFL55_I and found the demod phase that minimized the AS55_I and REFL55_Q. The newfound values were
    • C1:LSC-AS55_PHASE_R = -65.9586
    • C1:LSC-REFL55_PHASE_R = -78.6254
  • Repeating the above, but now misaligning YARM with ITMY yaw offset, locking XARM with AS55_Q and REFL55_I, we found the demod phases that minimized AS55_1 and REFL55_Q. The newfound values were
    • C1:LSC-AS55_PHASE_R = -61.4361
    • C1:LSC-REFL55_PHASE_R = -71.0434
• The above demod phases difference, Schnupp asymmetry between X and Y were measured. We repeated the measurement three times to derive the error.
  • Optimal demod phase difference between X arm and Y arm for both AS55 and REFL55 were measured to be -4.5 +/- 0.1 deg, which means that lx-ly = 3.39 +/- 0.05 cm (Marconi frequency: 11.066195 MHz).
• We measured the gain difference between AS55_Q and POX11/POY11 = -0.5
• We measured the gain difference between REFL55_I and POX11/POY11 = -2.5

After this, we locked DARM, CARM and MICH using POX11_I, POY11_I and AS55 error signals respectively, and actuating on ETMX, MC2, and BS with NO TRIGGERS (but FM triggers were on for boosts as usual). Under this condition, FM5 is used for lock acquisition, and FM1, FM2, FM3, FM6 are turned on with FM triggers. No FM4 was on. We also noticed:

• CARM FM6 "BounceRoll" is slightly different than "YARM" FM6 "Bounce". The absent roll resonant gain actually makes it easier to control the CARM, we just had to use YARM filter for locking it.
• When CARM is controlled, we often just kick the ETMX to bring it near resonance, since the frequency noise drops and we otherwise have to wait long.

Very nice!

DARM feedback should go to ETMY - ETMX, not just a single mirror: Differential ARM.

For it to work with 1 mirror the UGF of the CARM loop must be much larger than DARM UGF. But in our case, both have a UGF of ~150 Hz.

In principle, you could run the CARM loop with higher gain by using the CM servo board, but maybe that can wait until the X,Y -> CARM, DARM handoff.

I wasn't sure how the IMC servo was optimized recently. We used to have the FSS over all gain (C1:PSL-FSS_MGAIN) of +6dB a few years back. It is not 0dB. So I decided to do a couple of measurements.

1) Default setting:

C1:IOO-MC_REFL_GAIN +4
C1:IOO-MC_BOOST2 +3
C1:IOO-MC_VCO_GAIN +13
C1:PSL-FSS_MGAIN +0
C1:PSL-FSS_FASTGAIN +19

2) Looked at the power spectrum at TEST1A output (error signal)
TEST1A is the signal right after the input gain stage (C1:IOO-MC_REFL_GAIN). Prior to the measurement, I've confirmed that the UGF is ~100Hz even at +0dB (see next section). It was not too bad even with the current default. Just wanted to check if we can increase the gain a bit more.
The input gain was fixed at +4dB and the FSS overall gain C1:PSL-FSS_MGAIN was swept from +0 to +6.
At +5dB and +6dB, the servo bump was very much visible (Attachment 1).
I decided to set the default to be +4dB (Attachment 3).

3) Took OLTF at 0dB and 4dB for the FSS overall gain.

Now the comparison of the opel loop transfer functions (OLTF) for C1:PSL-FSS_MGAIN at 0dB and 4dB. The OLTF were taken by injectiong the network analyzer signal into EXCA and measure the ratio between TEST1A and TEST1B (A/B).

C1:PSL-FSS_MGAIN +0 -> UGF 100kHz / Phase Margin ~50deg
C1:PSL-FSS_MGAIN +4 -> UGF 200kHz / Phase Margin 25~30deg

The phase margin was a bit less but it was acceptable.

4) IMC FSR

Took the opportunity to check the FSR of the IMC. Connected a cable to the RF MON of the IMC REFL demod board. Looked at the peak at 40.56MHz (29.5MHz + 11.066MHz). The peak was not so clear at 11.066195MHz (see 40m ELOG 15845). The peak was anyway minimized and the new modulation frequency was set to be 11.066081MHz (new FSR). The change is 10ppm level and it is within the range of the temp drift.

We can calculate how much detector calibration uncertainty affects the estimation of astrophysical parameters using the following method:

Let be set of astrophysical parameters (like component masses, distance etc), be set of detector parameters (like detector pole, gain or simply transfer function vaue for each frequency bin). If true GW waveform is given by , and the detector transfer function is given by , then the detected gravitational waveform becomes:

One can calculate a derivative of waveform with respect to the different parameters and calculate Fisher matrix as (see correction in 40m/17017):

where the bracket denotes iner product defined as:

where is strain noise PSD of the detector.

With the gamma matrix in hand, the error propagation from detector parameter fractional errors to astrophysical paramter fractional errors is given by (eq 26 in Evan et al 2019 Class. Quantum Grav. 36 205006):

where and .

Using the above mentioned formalism, I looked into two ways of calculating error propagation from detector calibration error to astrophysical paramter estimations:

Using detector response function model:

If we assume detector response function as a simple DC gain (4.2 W/nm) and one pole (500 Hz) transfer function, we can plot conversion of pole frequency error into astrophysical parameter errors. I took two cases:

• Binary Neutron Star merger with star masses of 1.3 and 1.35 solar masses at 100 Mpc distance with a of 500. (Attachment 1)
• Binary black hole merger with black masses of 35 and 30 at 400 MPc distance with spin along z direction of 0.5 and 0.8. (I do not fully understand the meaning of these spin components but a pycbc waveform generation model still lets me calculate the effect of detector errors) (Attachment 2)

The plots are plotted in both loglog and linear plots to show the order of magnitude effect and how the error propsagation slope is different for different parameters. 'm still not sure which way is the best to convey the information. The way to read this plot is for a given error say 4% in pole frequency determination, what is the expected error in component masses, merger distance etc. I

Note that the overall gain of detector response is not sensitive to astrophysical error estimation.

Using detector transfer function as frequency bin wise multi-parameter function

Alternatively, we can choose to not fit any model to the detector transfer function and simply use the errors in magnitude and phase at each frequency point as an independent parameter in the above formalism. This then lets us see what is the error propagation slope for each frequency point. The hope is to identify which parts of the calibration function are more important to calibrate with low uncertainty to have the least effect on astrophysical parameter estimation. Attachment 3 and 4 show these plots for BNS and BBH cases mentioned above. The top panel is the error propagation slope at each frequency due to error in magnitude of the detector transfer function at that frequency and the bottom panel is the error propagation slope at each frequency due to error in phase of the detector transfer function.

The calibration error in magnitude and phase as a function of frequency would be multiplied by the curves and summed together, to get total uncertainty in each parameter estimation.

This is my first attempt at this problem, so I expect to have made some mistakes. Please let me know if you can point out any. Like, do the order of magnitude and shape of error propagation makes sense? Also, comments/suggestions on the inference of these plots would be helpful.

Finally, I haven't yet tried seeing how these curves change for different true values of the merger event parameters. I'm not yet sure what is the best way to extract some general information for a variety of merger parameters.

Future goals are to utilize this information in informing system identification method i.e. multicolor calibration scheme parameters like calibration line frequencies and strength.

Code location

1. In the error propogation equation, it should be \Delta \Theta = -H^{-1} M \Delta \Lambda, instead of the fractional error. 

2. For the astro parameters, in general you would need t_c for the time of coalescence and \phi_c for the phase. See, e.g., https://ui.adsabs.harvard.edu/abs/1994PhRvD..49.2658C/abstract.

3. Fig. 1 looks very nice to me, yet I don't understand Fig. 3... Why would phase or amplitude uncertainties at 30 Hz affect the tidal deformability? The tide should be visible only > 500 Hz. 

4. For BBH, we don't measure individual spin well but only their mass-weighted sum, \chi_eff = (m_1*a_1 + m_2*a_2)/(m_1 + m_2). If you treat S1z and S2z as free parameters, your matrix is likely degenerate. Might want to double-check. Also, for a BBH, you don't need to extend the signal much higher than \omega ~ 0.4/M_tot ~ 10^4 Hz * (Ms/M_tot). So if the total mass is ~ 100 Ms, then the highest frequency should be ~ 100 Hz. Above this number there is no signal. 

[Yuta, Paco]

In summary, we locked FPMI using REFL55_I, REFL55_Q, and AS55_Q. The key to success was to mix POX11_I and POY11_I in the right way to emulate CARM/DARM, and to find out the correct demodulation phase for AS55.

Procedure

1. Close PSL shutter and zero offsets in AS55, REFL55, POX11, POY11, and ASDC
   • For ASDC run python3 resetOffsets.py -c C1:LSC-ASDC_IN1, otherwise use the zer offsets on I and Q inputs from the RFPD medm screen.
2. Lock XARM/YARM using POX/POY to tune demodulation phase.
   • Today, the demode phase in POX11 changed to 104.801, and POY11 to -11.256 deg.
3. XARM and YARM are used in the following configuration
   • INMAT
     • 0.5 * POX11_I - 0.5 * POY --> XARM
     • 0.5 * POX + 0.5*POY --> YARM
     • REFL55_Q --> MICH (** this should be turned on after POX11/POY11)
   • LSC Filter gains
     • XARM = 0.012
     • YARM = 0.012
     • MICH = +40 (note the sign flip from last time)
   • OUTMAT
     • XARM --> 0.5 * ETMX - 0.5 * ETMY
     • YARM --> MC2
     • MICH --> BS
   • UGFs (sanity check)
     • XARM (DARM) ~ 100 Hz
     • YARM (CARM) ~ 200 Hz
     • MICH (MICH) ~ 40 Hz
4. Run MICHOpticalGainCalibration.ipynb to see if ASDC vs REFL55_Q looks nice (ellipse in the XY plot), and find any residual offset in REFL55_Q.
   • If the plot doesn't look nice in this regard, the IFO needs to be aligned.
5. Sensing matrix for CARM/DARM and MICH.
   • With the DARM, CARM and MICH lines on, verify the demod error signals look ok both in mag and phase.
   • For example, we found that CARM error signals were correctly represented by either 0.5 * POX11_I + 0.5 * POY11_I or 0.5 * REFL55_I.
   • Similarly, we found that DARM error signal was correctly represented by either 0.5 * POX11_I - 0.5 * POY11_I or 2.5 * AS55_Q.
   • To find this, we minimized CARM content in AS55_Q, as well as CARM content in REFL55_Q.
6. We acquired the lock by re-configuring the error point as below:
   • INMAT
     • 0.5*REFL55_I --> YARM (CARM)
     • 2.5 * AS55_Q --> XARM (DARM)
   • During the hand-off trials, we repeatedly ran the sensing matrix and UGF measurements while stopping at various intermediate mixed error points to check how the error signal calibrations changed if at all.
     • Attachment #1 shows the DARM OLTF using POX/POY (blue), only with CARM handoff (green), and after DARM handoff (red)
     • Attachment #2 shows the CARM OLTF using POX/POY (blue), only with CARM handoff (green), and after DARM handoff (red)
     • Attachment #3 shows the MICH OLTF using POX/POY (blue), only with CARM handoff (green), and after DARM handoff (red)
   • The sensing matrix after handoff is below:

Sensing Matrix with the following demodulation phases
{'AS55': 192.8, 'REFL55': 95.63177865911078, 'POX11': 104.80089727128349, 'POY11': -11.256509422276006}
Sensors          	           DARM     	           CARM     	            MICH     	
C1:LSC-AS55_I_ERR_DQ	5.09e-02 (89.6761 deg)	2.03e-01 (-114.513 deg)	1.28e-04 (-28.9254 deg)	
C1:LSC-AS55_Q_ERR_DQ	4.78e-02 (88.7876 deg)	3.61e-03 (-68.7198 deg)	8.34e-05 (-39.193 deg)	
C1:LSC-REFL55_I_ERR_DQ	5.18e-02 (-92.2555 deg)	1.20e+00 (65.2507 deg)	1.15e-04 (-102.027 deg)	
C1:LSC-REFL55_Q_ERR_DQ	1.81e-04 (59.0854 deg)	1.09e-02 (-114.716 deg)	1.77e-05 (-23.6485 deg)	
C1:LSC-POX11_I_ERR_DQ	8.51e-02 (91.2844 deg)	4.77e-01 (67.1709 deg)	7.97e-05 (-72.5252 deg)	
C1:LSC-POX11_Q_ERR_DQ	2.63e-04 (114.584 deg)	1.32e-03 (-113.505 deg)	2.10e-06 (118.146 deg)	
C1:LSC-POY11_I_ERR_DQ	1.58e-01 (-88.9295 deg)	6.16e-01 (67.6098 deg)	8.71e-05 (172.73 deg)	
C1:LSC-POY11_Q_ERR_DQ	2.89e-04 (-89.1114 deg)	1.09e-03 (70.2784 deg)	3.77e-07 (110.206 deg)	

Lock gpstimes:

1. [1342220242, 1342220260]
2. [1342220420, 1342220890]
3. [1342221426, 1342221574]
4. [1342222753, 1342223230]

Sensitivity estimate (NANB)

Using diaggui, we look at the AS55_Q error point and the DARM control point (C1:LSC-XARM_OUT). We roughly calibrate the error point using the sensing matrix element and actuation gain at the DARM oscillator freq 4.78e-2 / (10.91e-9 / 307.880^2). The control point is calibrated with a 0.95 Hz SUS pole. Attachment #4 shows the sensitivity estimate.

I measured the RIN by taking the spectrum of C1:MC_TRANS_SUMFILT_OUT and dividing it by the mean count on that channel (~13800 cts). Attachment 1 shows the result.

I updated the MICH AS55 noise budget but got a very low contribution (gold trace in attachment 2).

It seems too low I think. What could've gone wrong? Finesse calculates that the transfer function from laser amplitude modulation to AS55 is ~ 1.5e-9 at DC. If I turn off HOMs I get 1e-11 at DC, so this coupling is a result of some contrast defect. Should I include some RMS imbalances in the optics to account for this? Should I include it as a second-order effect due to MICH RMS deviation from zero crossing?

To balance the actuation on ETMX and ETMY, x4.12 was aded to C1:SUS-ETMX_(UL|UR|LR|LL|SD)COIL FM1. OSEM damping filter gains, oplev loop gains, and alignment offsets were divided by this factor.
C1:LSC-ETMX_GAIN is now 1.

To do:
 - Balance ETM and ITM. It should make ASS more sensible.
 - Re-commission Xarm ASS and Yarm ASS.

You should measure the coupling by noise injection. Noise budgeting does not need any modeling:

1) Measure the power spectrum density of the target signal (i.e. DARM) and the source noise (i.e. RIN this case)

2) Calibrate both using a calibration peak to convert 1) into the physical units (m/rtHz, 1/rtHz, etc)

3) Measure the transfer function from source to target using the noise injection. (i.e. RIN injection this case and look at the injection to RIN and injection to DARM)

4) Measure open-loop transfer functions if necessary. (i.e. DARM control open-loop transfer function to convert the error signal into the free running noise level)

Primarily, these are measured noise levels and noise couplings there is no room to involve a model there.
Once the noise budget was done, you can compare it with the model and say "the coupling is big/small/comparable".
 

Also, why don't you use C1:MC_TRANS_SUMFILT_IN1_DQ instead? Your _OUT signal seems affected by the bunch of comb notch filters to artificially remove the 60Hz harmonics. It's not a fair RIN measurement.

Now that you have found a working configuration, I suggest we update CARM and DARM filter banks so that they are used in locking those degrees of freedom instead of repurposing XARM/YARM banks. It would be bit easier to understand and leaves room for future changes for one configuration while keeping single arm lock configurations untouched.

Addressing the comments as numbered:

1. Yeah, that's correct, that equation normally but it is different if I define bit differently that I did in the code, correct my definition of to :
   
   then the relation between fractional errors of detector parameter and astrophysical parameters is:
   
   I prefer this as the relation between fractional errors is a dimensionless way to see it.
2. Thanks for pointing this out. I didn't see these parameters used anywhere in the examples (in fact there is no t_c in documentation even though it works). Using these did not affect the shape of error propagation slope function vs frequency but reduced the slope for chirped Mass by a couple of order of magnitudes.
   1. I used the get_t_merger(f_gw, M1, M2) function from Hang's work to calculate t_c by assuming  must be the lowest frequency that comes within the detection band during inspiral. This function is:
      
      For my calculations, I've taken as 20 Hz.
   2. I used the get_f_gw_2(f_gw_1, M1, M2, t) function from Hang's work to calculate the evolution of the frequency of the IMR defined as:
      
      where is the frequency at t=0. I integrated this frequency evolution for t_c time to get the coalescence phase phi_c as:
      
3. In Fig 1, which representation makes more sense, loglog of linear axis plot? Regarding the affect of uncertainties on Tidal amplitude below 500 Hz, I agree that I was also expecting more contribution from higher frequencies. I did find one bug in my code that I corrected but it did not affect this point. Maybe the SNR of chosen BNS parameters (which is ~28) is too low for tidal information to come reliably anyways and the curve is just an inverse of the strain noise PSD, that is all the information is dumped below statistical noise. Maybe someone else can also take a look at get_fisher2() function that I wrote to do this calculation.
4. Now, I have made BBH parameters such that the spin of the two black holes would be assumed the same along z. You were right, the gamma matrix was degenerate before. To your second point, I think the curve also shows that above ~200 Hz, there is not much contribution to the uncertainty of any parameter, and it rolls-off very steeply. I've reduced the yspan of the plot to see the details of the curve in the relevant region.

[Paco, Anchal-remote-support, Yuta]

We added fast channels to BHD DC PDs.
C1:LSC-DCPD_(A|B)_IN1 are now available, but C1:HPC-DCPD_(A|B)_IN1 still gives us zero.

c1hpc situation -> not good
 - We can see the slow signal at C1:X07-MADC1_EPICS_CH16 (DC PD A) and CH17 (DC PD B)
 - C1:HPC-DCPD_(A|B)_IN1 is there, but zero.
 - We have modified c1hpc model to add DCPD_(A|B) filters in front of the input matrix (see Attachment #1).
 - After modifying the model, we run
ssh c1sus2
rtcds make c1hpc
rtcds install c1hpc
ssh fb1
sudo systemctl restart daqd_*

 - After this, we got 0x2000 error. So, we ran the following. This removed 0x2000 error, but DCPD signals are still zero. They are also not available in C1HPC-MONITOR_ADC1.adl screen (see Attachment #3).
ssh c1sus2
rtcds restart c1hpc

c1lsc situation -> good
 - We could see the slow signal at C1:X04-MADC1_EPICS_CH4 (DC PD A) and CH5 (DC PD B), and also C1:LSC-DCPD_(A|B)_NORM after making C1:LSC-DCPD_(A|B)_POW_NORM=1. The ADC channel and DCPD channel are exactly the same.
 - After confirming the above, we modified the c1lsc model to add DCPD_(A|B) filters in front of the input matrix (see Attachment #2).
 - After modifying the model, we run
ssh c1lsc
rtcds make c1lsc
rtcds install c1lsc
ssh fb1
sudo systemctl restart daqd_*

 - After this, we also got 0x2000 error. We also noticed that, for example, C1:X04-MADC0_EPICS_CH31 and C1:LSC-ASDC_INMON are different, which used to be the same (ASDC_INMON was largely attenuated).
 - In the end, we run the following to remove 0x2000 error, but it crashed c1lsc, as well as c1sus, c1ioo.
ssh c1lsc
rtcds restart c1lsc

 - So, we did rebootC1LSC.sh. This made c1lsc, c1ioo and c1sus as green as before, except for RFM issue in TRX/TRY, like we saw in June. We followed the steps in 40m/16887 to hard reboot c1iscex/c1iscey and ran rebootC1LSC.sh again. This made C1CDS_FE_STATUS.adl screen as green as before (see Attachment #3).

 - Fast channels C1:LSC-DCPD_(A|B)_IN1 are now available. They are also available in C1LSC-MONITOR_ADC1.adl screen (see Attachment #3).

[Yehonathan, JC]

Yehonathan and I began to put the electronics on Rack 1X3. To do this, we had to move the monitor over the the PD testing table. Before mounting the Coil Drivers, we added numbers to the spaces to follow the rack plan Koji has provided. The drivers which have been mounted are PRM (Slots 10,11), BS (Slots 15, 16), ITMX (Slots 26, 27), and ITMY (34, 35).

[Paco, Yuta]

We measured contrast of Michelson fringe in both arms locked and mis-aligned. It was around 90%.
We also measured the contrast of ITM single bounce vs LO beam using BHD DC PDs. It was around 43%.
The measurement depends on the alignment and how to measure the maximum and minimum of the fringe. ITM-LO fringe was also not stable because motions of AS/LO mirrors are large. More tuning necessary.

Background
 - As measured in elog 40m/17012, we see a lot of CARM in AS, which indicates large contrast defect.
 - We want to check mode-matching of LO beam to AS beam.

BHD DC PD conditioning
 - We added DCPD_A and DCPD_B to /opt/rtcds/caltech/c1/scripts/LSC/LSCoffsets3 script, which zeros the offsets when shutters are closed.
 - We also set C1:LSC-DCPD_(A|B)_GAIN = -1 since they are inverted.

Contrast measurement
 - Contrast was measured using channels ['C1:LSC-ASDC_OUT','C1:LSC-POPDC_OUT','C1:LSC-REFLDC_OUT','C1:LSC-DCPD_A_OUT','C1:LSC-DCPD_B_OUT']. For LO, only DCPD_(A|B) are used.
 - We took 15%-percentile (40% for ITM-LO fringe) from the maximum and minimum of the data, and took the median to estimate the maximum value and the minimum value (see Attachment).
 - Contrast = (Imax - Imin) / (Imax + Imin)
 - We measured three times in each configuration to estimate the standard error.
 - Jupyter notebook: https://git.ligo.org/40m/scripts/-/blob/main/CAL/BHD/measureContrast.ipynb

Results
Both arms locked, MICH fringe (15% percentile)
Contrast measured by C1:LSC-ASDC_OUT is 89.75 +/- 0.17 %
Contrast measured by C1:LSC-POPDC_OUT is 79.41 +/- 0.86 %
Contrast measured by C1:LSC-REFLDC_OUT is 97.34 +/- 0.34 %
Contrast measured by C1:LSC-DCPD_A_OUT is 95.41 +/- 1.55 %
Contrast measured by C1:LSC-DCPD_B_OUT is 89.76 +/- 1.49 %
Contrast measured by all is 90.34 +/- 1.68 %

Both arms mis-aligned, MICH fringe (15% percentile)
Contrast measured by C1:LSC-ASDC_OUT is 89.32 +/- 0.57 %
Contrast measured by C1:LSC-POPDC_OUT is 94.55 +/- 0.62 %
Contrast measured by C1:LSC-REFLDC_OUT is 97.95 +/- 1.37 %
Contrast measured by C1:LSC-DCPD_A_OUT is 96.40 +/- 1.04 %
Contrast measured by C1:LSC-DCPD_B_OUT is 90.98 +/- 1.07 %
Contrast measured by all is 93.84 +/- 0.94 %

ITMY-LO fringe (40% percentile)
Contrast measured by C1:LSC-DCPD_A_OUT is 45.51 +/- 0.45 %
Contrast measured by C1:LSC-DCPD_B_OUT is 38.69 +/- 0.43 %
Contrast measured by all is 42.10 +/- 1.03 %

ITMX-LO fringe (40% percentile)
Contrast measured by C1:LSC-DCPD_A_OUT is 46.65 +/- 0.65 %
Contrast measured by C1:LSC-DCPD_B_OUT is 39.82 +/- 0.51 %
Contrast measured by all is 43.24 +/- 1.45 %

Discussion
 - As you can see from the attachment, REFLDC is noisy and over estimating the contrast. ASDC is reliable. We need to tune the threshold to measure the maximum value and minimum value. We should also use the mode instead of median.
 - Contrast depends very much on the alignment. We didn't tweak too much today.
 - ITM-LO fringe was not stable, probably due to too much motion in AS1, AS4, LO1, LO2. Their damping needs to be re-tuned.

Next:
 - Model FPMI sensing matrix with measured contrast defect
 - Estimate AS-LO mode-mismatch using the measured contrast
 - Lock ITM-LO fringe using DCPD_(A|B) as error signal, and ITM or LO1/2 as actuator
 - Lock MICH with DCPD_(A|B), and with LO beam
 - Get better contrast data with better alignment and better AS1, AS4, LO1, LO2 damping

[Paco, Yehonathan, JC]

We were trying to setup the Jenne laser to characterize the response of three 1811s that Yehonathan is using for his WOPA experiment (in QIL). We hooked up a ~ 5 VDC power supply to the bias tee and looked to see if there was any DC response in the REF PD. We used a DB9 breakout board and a DB9 cable, and saw some current being drawn. The DC current was a bit too high (500 mA), so we turned the DC voltage off, and realized the VDC power was reversed, probably along the DB9 cable which we didn't check before. As we flipped the power supply leads and turned power back on, we could no longer see any current even though the voltage was now right (or was it???). We would like to debug this laser, and continue using it if it still works (!), but there is negligible documentation either here or in the wiki, so if there are any known places to look at it would be helpful to know them.

[Koji, Yehonathan, Paco]

Koji pointed out that this laser was always driven with a current driver (which was not nearby), and after finding it on one of the rolling carts, we hooked up the system but found that the laser driver displayed open circuit near the usual 20mA operating point. We therefore have to conclude that this laser is no more. We will look for a reasonable replacement.

For troubleshooting, the proper laser driver (found beneath the AG network analyzer) was connected.
The current ~1mA was provided and the driver detected the "open circuit", which means the laser diode was busted.

https://dcc.ligo.org/LIGO-T060240

The laser diode in the parts list is: "GTRAN GaAs Strained QW Laser Diode, Part # LD-1060".

[Paco, Yuta, JC]

We did some easy tests on the BHD readout in preparation for BHD MICH. With the arm cavities and LO beam misaligned, but the MICH aligned, we measured the transfer function from C1:LSC-DCPD_A_OUT to C1:LSC-DCPD_B_OUT to get a rough estimate of the gain balance: 1.8 * DCPD_A = DCPD_B. We then locked MICH using REFL55_Q and looked at

• A=C1:LSC-DCPD_A_OUT
• B=C1:LSC-DCPD_B_OUT
• 1.8 * A - B (which we encoded using C1:LSC-PRCL_A_IN1)
• 1.8 * A + B (which we encoded using C1:LSC-PRCL_B_IN1)

namely the DCPD BHD signals. After turning the MICH_OSC on (2000 gain @ 311.1 Hz), we took some power spectra under the following three configurations:

1. LO misaligned, no MICH offset.
2. LO overlap, no MICH offset.
3. LO overlap and MICH offset.

For 1. the expectation was that since LO is misaligned and the AS port is dark, we would get no signal. In 2., however both A and B would might see some incoherent signal, but still no MICH. Finally in 3. all signals should be able to see MICH, including A-B. Attachment #1 shows the measurements 1, 2, and 3 (offset = -5.0). Then, with increasing offset values, the BHD MICH signals increased as well; discussion to follow.

IPC issue still unresolved.

Updated shared memory tag so that 'SUS' -> 'SU2' in c1hpc, c1bac and c1su2. Removed obsolete 'HPC/BAC-SUS' references from IPC file, C1.ipc. Restarted the FE models but the c1sus2 machine froze, so I did a manual reboot. This brought down the vertex machines---which I restarted using /opt/rtcds/caltech/c1/scripts/cds/rebootC1LSC.sh---and the end machines which I restarted manually. Everything but the BHD optics now have their previous values. So need to burtrestore these.
 

# IPC file:
/opt/rtcds/caltech/c1/chans/ipc/C1.ipc

# Model file locations:
/opt/rtcds/userapps/release/isc/c1/models/isc/c1hpc.mdl
/opt/rtcds/userapps/release/sus/c1/models/c1su2.mdl
/opt/rtcds/userapps/release/isc/c1/models/isc/c1bac.mdl

# Log files:
/cvs/cds/rtcds/caltech/c1/rtbuild/3.4/c1hpc.log
/cvs/cds/rtcds/caltech/c1/rtbuild/3.4/c1su2.log
/cvs/cds/rtcds/caltech/c1/rtbuild/3.4/c1bac.log

SUS overview medm screen :

• Reduced the entire screen width
• Revert to old screen style watchdog layout

[Tega, Yuta]

We were able to fix the shared memory issue by updating the receiver model name from ''SUS' to 'SU2' and the ADC zero issue by including both ADC0 and ADC1 in the c1hpc and c1bac models as well as removing the grounding of the unused ADC channels (including chn#16 and chn#17 which are actually used in c1hpc) in c1su2. We also used shared memory to move the DCPD_A/B error signals (after signal conditioning and mixing A/B; now named A_ERR and B_ERR) from c1hpc to c1bac.
C1:HPC-DCPD_A_IN1 and C1:HPC-DCPD_B_IN1 are now available (they are essentially the same as C1:LSC-DCPD_A_IN1 and C1:LSC-DCPD_B_IN1, except for they are ADC-ed with different ADC; see elog 40m/16954 and Attachment #1).
Dolphin IPC error in seding signal from c1hpc to c1lsc still remains.

[Koji Paco]

Koji went to Downs and found a CRT labeled "for 40m Rana?". So I decided to salvage it to the 40m after getting approval from Rich/Todd.

Paco and I tried this unit with the control room CCD signal and it just worked fine. So we can use this as a spare for any purpose in the lab.

[Tega, Yuta]

We have added C1:HPC-DCPD_A_ERR and C1:HPC-DCPD_B_ERR testpoints, which can be used as A+B, A-B etc.
Restarting c1hpc crashed c1sus2, and also made c1lsc/ioo/sus models red.
We run /opt/rtcds/caltech/c1/Git/40m/scripts/cds/restartAllModels.sh to restart all the machines. It worked perfectly without manually pressing power buttons! Wow!

We have also edited /opt/rtcds/caltech/c1/medm/c1su2/C1SU2_WATCHDOGS.adl so that it will use new /opt/rtcds/caltech/c1/Git/40m/scripts/SUS/medm/resetFromWatchdogTrip.sh instead of old /opt/rtcds/caltech/c1/scripts/SUS/damprestore.py.

Sorry I forgot to put tc & phic in the example. 

I modified astroFisherLib.py to include these parameters. Please note that their meaning is that we don't know when the signal happens and at which phase it merges.

It does not mean the time & phase from a reference frequency to the merger. This part is not free to vary because it is fixed by the intrinsic parameters.  

It might be good to have a quick scan through the Cutler & Flanagan 94 paper to better understand their physical meanings.

[Paco, Yehonathan]

==== Late elog from Friday ====

Koji provided us with a QFLD-950-3S (QPHOTONICS) salvaged from Aidan's junk pile (LD is alive according to him). We tested the Jenne laser setup with this just to decide if we should order another one, and it worked.

The laser driver anode and cathode pins (8/9, 4/5 respectively) on the rear DB9 port from the  ILX Lightwave LDX-3412 driver were connected to the corresponding anode and cathode pins in the laser package (5, and 9; note the numbers are reversed between driver and laser). Then, interlock pins 1 and 2 in the driver were shorted to enable operation. This is all illustrated in Attachments #1-2.

After setting a limit of 27.6 mA current in the driver, we slowly increased the actual current to ~ 19 mA until we could see light on a beam card. We can go ahead and get a 1060 nm replacement.

The DFD was setup to measure the change in beatnote when excited. A long long (128in) cable goes from the SR785 near the DFD all the way to the Xend AUX which it accordingly excites and the DFD is monitored by the oscilloscope at the other end. This was completed on Friday. The wires and stand have been moved to the side but the setup is still a bit chaotic. As of writing this post, there is still atleast some minor issue with the setup as we aren't getting the expected output. 

[I will shortly update this elog with more pictures]

Edit: the SR785 was replaced by the AG 4395, and pictures added

From the 40m wiki, I was able to use the instructions here to map out what to do to get the IPC issue resolved. Here is a summary of my findings.

I updated the /etc/dis/dishost.conf file on the frame builder machine to include the c1sus2 machine which runs the sender model, c1hpc, see below. After this, the file becomes available on c1sus2 machine, see attachment 1, and the c1sus2 node shows up in the dxadmin GUI, see attachment 2. However, the c1sus2 machine was not active. I noticed that the log file for the dis_nodemgr service, see attachment 3, which is responsible for setting things up, indicated that the dis_irm service may not be up, so I checked and confirmed that this was indeed the case, see attachment 4. I tried restarting this service but was unsuccessful. I restarted the machine but this did not help either. I have reached out to Jonathan Hanks for assistance.

[Paco, Tega, Yuta]

Today, we made a custom MEDM screen for the BHD Homodyne Phase Control, which is basically an overview of the c1hpc model. See Attachments 1 & 2 for details.

Measured the PZT beatnote using the setup mentioned in elog post 17031. Attached is the data taken from 10kHz to 1MHz, decadewise data was also taken that I'm not including in this post. A_R refers to the transfer function taken of channel A wrt the voltage reference (the swept sine we are inputting which has an IF of 30kHz). A and B correspond to the I and Q components of the signal taken from the DFD, respectively. I am currently working on plotting the data, and will shortly update this post with plots. Next steps - 

- quantify the uncertainty in the signal (I think)

- vectfit the data to find poles and zeroes

(and possibly find a better way to print/obtain data)

Edit: first pass of data plotted

Trying to vectfit to the data taken from the DFD previously but failing horribly. I will update this post as soon as I get anything semi-decent. For now here is this fit.

[Yuta, Paco]

TL;DR Successfully controlled LO phase, and did BHD-MICH readout with various MICH offsets and LO phases.

Today we implemented a DCPD based LO phase control. First, we remeasured the balancing gain at 311.1 Hz (the MICH oscillator freq) and combined C1:HPC-DCPD_A_OUT with C1:HPC-DCPD_B_OUT to produce the balanced homodyne error signal (A-B). We feed this error signal to C1:HPC-LO_PHASE_IN1 and for the main loop filters we simply recycled the LSC-MICH loop filters FM2 through FM5 (we also copied FM8, but didn't end up using it much). Then, we verified the LO phase can be controlled by actuating either on LO1 or LO2. For LO2, we added an oscillator in the HPC LOCKINS at 318.75 Hz (we kept this on at 1000 counts for the measurements below).

The LO phase control was achieved with a loop gain in the range of 10-30 (we used 20), no offset, and FM4, and FM5 engaged. FM2 can be added to boost, but we usually skipped FM3. Then, we went through a set of measurements similar to the ones described in a previous elog. A key difference with respect to the measurements from before is that we locked MICH using AS55Q (as opposed to REFL55Q). This allowed us to reach higher MICH offsets without losing lock. After turning on the MICH oscillator at 3000 counts, we looked at:

1. LO misaligned + MICH at dark fringe (offset = -21).
   • Here, we don't expect to see any MICH signal and indeed we don't, except for a small residual peak from perhaps a MICH offset or slightly imbalanced PDs.
2. LO aligned, but uncontrolled + MICH at dark fringe (offset = -21).
   • Here we would naively expect MICH to show up in A-B, but because of the uncontrolled LO phase, we mostly see the noise baseline (mostly from LO RIN? ...see measurement 3) under which this signal is probably buried. Indeed, the LO fringe increased noise in A, B, and A-B but not in A+B. This is nice.
3. LO aligned, but uncontrolled + MICH with dc readout (offset = +50).
   • Here we expected the MICH signal to show up due to the large offset, and we can indeed see it in A, B, and A+B, but not in A-B. Nevertheless we see almost exactly the same noise level even though we allow some AS light into the BHD readout, so maybe the noise observed in the A-B channel from measurements 2 and 3 is mostly from LO RIN. This needs further investigation...
4. LO aligned, controlled at no offset + MICH with dc readout (offset = +50).
   • In general here we expected to see a noise reduction in the A-B channel since the LO fringe is stable, and a MICH signal should appear. Furthermore, since LO phase is under control, we expect the LO2 Oscillator to appear which it does for this and the following measurements. Because of the relative freedom, we tried this measurement in two cases:
     1. When feeding back to LO1
        • We actually see MICH in the A-B channel, as expected, after the noise level dropped by ~ 5. We also observed small sidebands +- 1 Hz away from the MICH peak, probably due to local damping in either LO or AS paths.
     2. When feeding back to LO2
        • We also see MICH here, with a slightly better drop in noise (relative to feeding back to LO1). Sidebands persisted here, but around at +- 2 Hz.
5. LO aligned, controlled (offset = 10) + MICH with dc readout (offset = +50). *
   • Here, we expected the A-B MICH content to increase dramatically, and indeed it does after a little tuning of the LO phase . The noise level decreased slightly because LO phase noise is decreased around the optimal point.
6. LO aligned, controlled (offset = 20) + MICH with dc readout (offset = +30). *
   • Here, we naively expected A+B MICH content to decrease, but A-B remain constant. In order to see this we tried to keep the balance between the offsets, but this was hard. We don't really see much of this effect, so this also needs further investigation. As long as we keep controlling the LO phase using the DCPDs because the offsets tend to reduce the error signal we will have a harder time.

* For these measurements we actuated on LO2 to keep the LO phase under control.

Note that the color code above corresponds to the traces shown in Attachment #1.

What's next?

• Alignment of LO and AS might be far from optimized, so it should be tried more seriously.
• What's the actual LO power? How does it compare with AS power at whatever MICH offsets?
• Try audio dither LO phase control.
  • With MICH offset.
  • Without MICH offset, double demod (after dolphin fix )

I think the fit fails as the measurement quality is not good enough.

Paco and I messed around with the attenuation of the scope and bandwidth of the IF. We also replaced the BNC T's in the circuits with RF splitters. We saw some decent improvements to the data. The data is attached and a diagram of the experiment. [We analytically calculated the impedances to avoid any mismatch taking place]. Working on fitting the data.

We also moved around the wires so that the AG4395 is closer to the PZT.

[Paco, Yehonathan, Yuta]

We measured power and counts at BHD DCPDs with LO beam only and ITM single bounce.
We found that LO beam power is ~7 times less than the expected.
We also confirmed that AS beam is clipped somewhere inside vacuum and have 20-50% less power compared with the expected.
LO/AS beams going to DCPD A and B also have power imbalance by 30-40%.

What we did:
 - Run LSCoffsets.py to zero the offsets. I modified the old script so that it can handle new BHD PDs. Also, a bug was fixed (it didn't take into account the gains in filer modules, so INMON is now used instead of OUT16 for calculating offsets).
https://git.ligo.org/40m/scripts/-/blob/main/RFPD/LSCoffsets.py
 - Measured powers and counts in BHD DCPDs at ITMY table with LO beam only and ITMX/ITMY single bounce.
 - During the measurement, we found that power into DCPD A and DCPD B are quite different. One of the reason was a lens and an iris right after the viewport for A path. We removed both of them. Also, only A path have a pickoff which picks off ~20% of light to BHD camera (called SRMF; 40m/16880).
 - We also found that LO beam shape is ugly. ITM single bounce beam from X and Y have similar clipping (see Attached photos). We tried to reduce clipping with various suspensions (LO1, LO2, AS1, AS4, SR2, SRM, BS, ITMX, PR2), but was not possible by moving only single suspension.

Result:
 - Result of counts and power measurements are as follows. Power was measured right in front of DCPD, and also right after the viewport to estimate the loss in the in-air paths. Note that LSC channels have gain of 1, but HPC channels have gain of -1.9 for DCPD_A and -1 for DCPD_B.

                       Blocked  LO       ITMX      ITMY    
C1:LSC-DCPD_A_OUT16    -0.01    -17.89    -91.62    -86.21    
C1:LSC-DCPD_B_OUT16    +0.06    -17.72   -131.83   -131.98    
C1:HPC-DCPD_A_OUT16    +0.07    +34.12   +174.63   +164.24    
C1:HPC-DCPD_B_OUT16    +0.13    +17.60   +131.31   +131.49    
Power at DCPD_A        19 uW    63 uW    278 uW    290 uW    
Power at DCPD_B        19 uW    65 uW    393 uW    404 uW    
Power at viewport A    -- uW    82 uW    350 uW    337 uW    
Power at viewport B    -- uW    64 uW    436 uW    431 uW

DCPD calibration:
 - From the measurements above, counts/W in IN1 can be calculated as follows. Offset of 19 uW is substracted from the measured power to take into account for background light.

C1:LSC-DCPD_A_IN1     -3.59e+05 counts/W
C1:LSC-DCPD_B_IN1     -3.61e+05 counts/W
C1:HPC-DCPD_A_IN1     -3.60e+05 counts/W
C1:HPC-DCPD_B_IN1     -3.57e+05 counts/W

Discussion:
 - DCPD calibration shows that DCPD to ADC itself is quite balanced within 1%. A factor of 1.8-1.9 seen was from unbalanced light between A path and B path (40m/17037).
 - Power expected for ITM single bounce to one of DCPDs is ~520 uW, but was 350-430 uW as measured right after the viewport. Power at A is significantly less than that for B. Note that power at AS55 was as expected (40m/16952). Also, clipping cannot be reduced by moving suspensions. These could mean that clipping is happening after AS2.

950 mW * 0.9 (IMC transmission?) * 5.637%(PRM) * 97.8%(PR2) * 50%(BS) * 98.6%(ITM) * 50%(BS) * 10%(SRM) * 90%(AS2) * 50%(BHDBS) = 520 uW

 - Power expected for LO beam to one of DCPDs is ~530 uW, but was 60-80 uW as measured right after the viewport. Power at A is significantly more than that for B, which is opposite for ITM single bounce. This could mean that something is happening at BHDBS? We are not sure why the power is so low. Are we seeing some ghost beam? For PR2 transmission, 22000 ppm was used for calculation, from 40m/16541.

950 mW * 0.9 (IMC transmission?) * 5.637%(PRM) * 2.2%(PR2) * 50%(BHDBS) = 530 uW

 - As far as we remember, beam shapes were not as bad when we closed out the chambers...

Next:
 - Check if measured power explains the visitivity of LO-ITM single bounce (40m/17020)
 - If not, what is the mode mismatch? Is it possible to explain the mode mismatch with deviations from designed mode-matching telescope?
 - Measure POP power to see if PR2 actually have T=2.2%
 - Play with LO1 and LO2 to invesitate LO beam shape and power
 - Check coherence between LO/AS power fluctuations with suspension motions
 - What is the expected counts/W for these DCPDs?
 - Balance the optical paths in ITMX table for A and B (same lenses, same mirrors)
 - Install better lens in front of camera

The LO beam was found to have a power of 60uW, 10% of the power expected. We are pretty sure about the expectation because the AS beam has a power of 300uW, roughly the expected power. Additionally, the visibility of the MICH fringes in the BHDR is 40%.

If the mode-matching is perfect then we expect the visibility to be

which is roughly 74.5%.

If there is some mode-mismatch one can show that the visibility is , where is the mode-mismatch.

Using Finesse model I calculated \sqrt(M)=0.93 in the MICH configuration so the expected visibility is around 70%, far away from the observed 40%. To explain the observed visibility the mode mismatch would have to be ~ 30% which is very unlikely.

So it could either be a ghost beam or that the LO beam is clipped so badly that it also degrades its phase front (and therefore the mode-matching). The fact that we see fringes on the LO beam might suggest knife edge clipping on one of the auxiliary optics in the BS chamber.

{Yuta, Yehonathan}

We went to the BS table to check the POP beam power. We first notice that the POP beam has a nice gaussan profile on the viewing card. We traced it the beam to the viewing port and measured the power there. Before measuring the power we misalign ITMY/ITMX to get rid of interferences. We measure the beam to be 205uW in both cases.

The expected power is

950 mW * 0.9 (IMC transmission?) * 5.637%(PRM) * 97.8%(PR2) * 50%(BS) * 98.6%(ITM) * 50%(BS)  * 2.2%(PR2) = 260uW

which is reasonably close to what we measure which confirms that PR2 transmission is around what we think it is.

This strengthen our suspicion that LO beam gets clipped somewhere.

We also improved the clipping on the POP camera by one of the beamsplitters along the beam path and the alignment to the POPDC PD (~100 cts before, ~ 1000 cts after).

As a result of the following work, the file volume of /cvs/cds was reduced from 3.2TB to 2.2TB, and /opt/rtcds/caltech/c1/scripts was reduced from 10GB to 1.5GB

/cvs/cds/caltech/scripts_archive was cleaned up. Now the archive files are reduced to have:

• every month 1st day from 2005 to 2018/12
• every ten days (1, 11, 21) for 2019 and 2020
• everyday for 2021 and 2022

(scripts)/MEDMtab/image was deleted. I can be restore back from one of the script_archive files.

(scripts)/MC/logs/AutoLocker.log was just deleted and refreshed. For the past settings, we can refer autoburt snapshots or script_archive files.

(scripts)/Admin/n2Check.log

• It turned out that the frequency of the check was reduced to once per 10min on Sep 9th, 2021 (unelogged activity).
• The volume of the text since then was not much volume. So I deleted the lines before this date. And the file size is <7MB now.

(scripts)/ZI was moved to /cvs/cds/apps

/opt/rtcds/caltech/c1/burt/autoburt/snapshots

• 2018, 2019, 2020 snapshots were archived in tar.gz.
• These snapshots were then deleted

[Yuta, Tega, Yehonathan]

To investigate the BHD power imbalance and clipping issues, we did some shaking test of the mirrors in the LO path and AS paths. The results suggests the following:

• The clipping is happening after the BHD BS in DCPD_A path, as opposed to our initial guess of BHD BS transmission clipping in elog 17040.
• The LO beam we are seeing is probably a ghost beam from PR2

We performed both PIT and YAW shaking of all mirrors and looked at the output at DCPD_A and DCPD_B, see table below for details. Since we only see the dithering signal in DCPD_A, it suggests that the clipping is ocurring after the BHD BS and is also confined to the path between BHD BS and DCPD_A. We also swapped the camera location from DCPD_A to DCPD_B on ITMY table and confirmed that the beam was clipped for DCPD_A but not for DCPD_B.

This result discounts the possiblity of clipping being responsible for the power imbalance and therefore suggests that the power imbalance may actually be due to BHD BS not being 50:50. From the measurement in elog 17040, the transmission of BHD BS is 440.3% and the reflectivity is 560.3%. Note that DCPD_A is the transmission of BHD BS for AS beam, whereas DCPD_B is the reflection of BHD BS for AS beam, elog 16932.

We expect the shaking of PR2 to give no signal in either DCPD_A or DCPD_B when the LO beam is purely in trasmission, however, we see a signal in DCPD_A sugesting that the LO beam transit path through PR2 may not be as expected, i.e. the beam might be exiting the side of PR2 instead of the AR coated surface.

Finally, we measured the coherence between the dithering dof and DCPD_A/B & POP, see attachment 2, where we noticed that both DCPD_A/B have high coherence in the 1Hz-10Hz frequency band whereas ther was no coherence in POP as expected. This suggests that there may also be some small clipping in DCPD_B path.

LO Beam Shaking (LO1, LO2, PR2):

AS Beam Shaking (AS1 and AS4)

Some insights from the inside vacuum situation:

• The beam is an incident near normal on PR2 close to the center of the optic. It wasn't hard to align this part, I'm very confident that we aligned it to the center of PR2. So I do not think the LO beam is ghost beam from PR2.
• The place that is most susceptible to clipping is POP_SM5 mirror in front of LO1. The LO beam has little clearance from the edge of the mirror.
• Another possibility of clipping in LO beam is through the cage of LO2. LO2 is a 45-degree incidence mirror, so it is possible we are clipping off the cage or seeing a ghost beam mixed in LO beam here.
• The fact that moving PR2 is affecting LO beam is weird but doesn't necessarily mean it is a ghost from PR2.

[Yuta, Tega]

From our previous work (elog 17044) of shaking PR2 and seeing a signal in DCPD_A and the fact that LO beam power is far smaller than the expected nominal value, we decided to use TT1 and TT2 to realign the LO beam. This resulted in LO beam power going up by a factor of 6 and an improvement in the LO beam shape. We are still unable to find LO and AS alignment which realize BHD fringe with no clipping everywhere.

Deformed LO beam issue: Following the TT1 and TT2 alignments, used PR2 and PR3 to recover the transmission of the X and Y arms to 1. We also used LO1 and LO2 offsets to further reduce the beam deformation by eliminating the HOM concentric fringes that surounded the LO beam and to maximize the DCPD outputs. BHD optics in ITMY table was tweaked a lot to keep the LO beam centered on the BHD DCPDs and camera. The improved LO beam is still astigmatic in the yaw direction but at least now looks like a TEM00 mode. We also repositioned the DCPD_A path camera lens to remove the circular diffused fringes due to lens clipping. After the alignment, power was measured to be the following. It also reduced the coherence between DCPD outputs and suspension motions (see attached).

                       LO         ITMX
C1:HPC-DCPD_A_OUT16    +127.50    +96.24 (ITMX single bounce consistent to 40m/17040)
C1:HPC-DCPD_B_OUT16    +120.51    +141.52
Power at viewport A    504 uW (almost as expected 40m/17040)
Power at viewport B    385 uW

AP table AS beam clipping: We also noticed clipping in the AS beam in AP table which we removed by moving SR2 and AS1 in YAW and then used AS4 to keep the BHD AS beam centered in the BHD DCPDs.

BHD fringe: After overlaping the LO and AS beams, we saw diagonal fringes indicating beam tilt of LO wrt AS, so we tried to remove the AS beam tilt using AS1 and AS4 but failed to do so because the AS4 mirror seemed to completely distort the beam, so intead we decided to use SR2 and AS1 to remove the tilt between LO and AS beams, which realized BHD fringe. But the motion of SR2 and AS1 then moved the AS beam that it is no longer seen in AP table. The alignment to realize LO and AP AS beam without clipping, and that to realize BHD fringe are attached.

MCAutolocker/FSSSlow are not properly documented and not properly working.
Tidy up the script and documentation, or bring it back to megatron

I was aware that the FSSSlow was misbehaving since the shutdown upon the July power outage.
- FSS Slow servo did nothing even though the apparent settings in C1:PSL_SLOW screen looked fine and heart beat blinking
- Wanted to restart FSSSlow at megatron. Despite the login message showing how to do it, the system service does not exist anymore, because it was moved somewhere.
- Searched 40m wiki but found no info how to kill and restart it
- Found an elog. It was moved to docker on optimus ELOG 16480 . The restart procedure can only be found here. Please fix all the documentation inconsistency >> Anchal

According to this elog, the following commands need to be run for starting up MCAutolocker and FSSSlow on optimus:

cd /opt/rtcds/caltech/c1/Git/40m/scripts/MC
sudo docker-compose up -d

- Problem continues. Now FSSSlow is running but only when the IMC is locked. It does not stop even when the IMC lock is lost. How can we debug docker thing?
- This is minor but the MCAutolocker log (/opt/rtcds/caltech/c1/scripts/MC/logs/AutoLocker.log) is no longer updated even though MCAutolocker is running. Was it moved somewhere?

I wanted to check what's wrong with the WFS.

I played with MC2 misalignment to check the error signals.
MC2 pitch and yaw misalignment optically produce a vertical translation and horizontal rotation of the cavity axis at the waist, respectively. So it is thought to be a more separated excitation of the cavity axis.
Then I noticed that WFS2 error signals in general has high (~100%) pitch-yaw coupling. So it was suspicious.

I went to the rack and found that WFS2 SEG4 RF input (labeled "8") was not completely inserted. (Attachment 1)
It seemed that the LEMO connector or the receptacle didn't latch properly anymore and could be easily pulled.
I gave some elbow grease to fix this but in vain. I ended up to use LEMO-BNC adapters which somehow offered a robust connection.

Desipite the insightful discovery, this was not the intrinsic solution to the issue. I checked the past signal history, but I don't think this loose connection caused the missing signal.

Next, I needed to go a bit deeper. The WFS sensors are supposed to be adjusted to I phase where the PDH signal maximally shows up. And all the segments are supposed to have the same sign in terms of the PDH signal.

I've unlocked the IMC and turned on MC2 tickling. This swept the cavity over the resonances.
WFS1 SEG1I~3I showed about the same waveform, but SEG4 Q shows the PDH signal rather than SEG 4 I.
Then tried the same test for WFS2. The SEG4 I signal has the sign-flipped PDH signal compared to WFS2 SEG1I-SEG3I.

I quickly adjusted the demod phase of WFS1 SEG4 and WFS2 SEG4 to correct them,

WFS1 SEG4 103.9-> -20
WFS1 SEG4 -58  -> 120

This in fact made the pitch and yaw separated but flipped (Pitch signal shows up in WFS1Y and yaw signal shows up in WFS1P. Same for WFS2)

These modifications were reverted upon my leaving.

Now things are much more subtle now. And I need to do a more careful quantitative analysis of the demodulation phases / input matrix / output matrix.

Note: It seems that I had worked on IMCWFS on Dec 21, 2016

The FSSSlow script was not properly documented and it was not working, so I had to use one that I knew worked from CTN. This scripts lives in

/opt/rtcds/caltech/c1/Git/40m/scripts/PSL/FSS/PIDLocker.py

and uses a configuration file

/opt/rtcds/caltech/c1/Git/40m/scripts/PSL/FSS/PIDConfigFSS.yml

The script runs all the time inside docker container which keeps it running. The heart-beat blinker shows if the script is active or not, but it only starts working when C1:IOO-MC_LOCK is 1 and C1:PSL-FSS_SLOWLOOP is 1. The second channel is a button on C1PSL_SLOW screen to engage autolocker. It has to be turned on for FSS to work.

docker instructions:

The following message is displayed on login in optimus:

For checking log files of autolocker script, on optimus do:

For checking log files of FSS PID loop, on optimus do:

In the above commands, add < | tail -15> to just see the most recent 15 lines in the log file. change 15 to whatever number of lines you want to see from the end.

At any time, if you want to know how docker is running stuff, check out the /opt/rtcds/caltech/c1/Git/40m/scriptsdocker-compose.yml file for self-explanatory script usage.

I'll add some documentation on the wiki soon. That is indeed required and should have been done already.

Debugging scripts:

All scripts could be debugged by running them on rossa by directly using python command. You can stop the docker container on optimus using:

and then run the file on rossa to check it's behavior. After debugging and fixing any issues, please commit the file to gitlab repo and go back to optimus and restart the docker container:

I'll add this procedure to a wiki page as well.

Reverting back to systemd on megatron

The setup on megatron was not removed at all. All service files exist in same place and old scripts can be started in the old manner by doing following on megatron:

For FSSSlow:

For MC autolocker:

For diabling these services again, do:

Note that one should stop docker containers on optimus before starting these systemd services to avoid conflicting scripts running together.

I have added above instructions on megatron motd. So on loging into megatron, these updated instructions would come.

If someone wants to fix the old scripts and use systemd for managing those scripts, please do so but I won't be able to help in debugging those old scripts. The shell scripts are very complicated and beyond my knowledge and python scripts are lacking documentation.

I'm happy to help debug or extend the functionality of the new scripts that live in the git directory.

> it only starts working when C1:IOO-MC_LOCK is 1 and C1:PSL-FSS_SLOWLOOP is 1.

- OK. Your new MCAutolocker does not reflect the lock status to C1:IOO-MC_LOCK. This causes FSS Slow to go crazy when the IMC is not locked. Can you fix that?

- So C1PSL_SLOW.adl screen, which spawns by the "SLOW Servo" button on the FSS screen, has no effect on the FSS SLOW servo anymore. It is obsolete. At least the screen (or the link to the screen) should be removed. (Work on it once you are back.)

- Also, please make a wiki page and copy the description on the previous page.

Have successfully plotted data from the Red Pitaya on Jupyter Notebook! Have lost years of my life fighting with PyQt. Thanks to Deeksha for heavy contribution. Next task is to get actually good data (seeing mostly noise right now and haven't figured out how to change my input settings) and then to go to set up the RPi in the lab.

[Yuta, Tega]

We decided to give the dolphin debugging another go. Firstly, we noticed that c1sus2 was no longer recogonising the dolphin card, which can be checked using

or looking at the status light on the dolphin card of c1sus2, which was orange for both ports A and B.

We decided to do a hard reboot of c1sus2 and turned off the DAQ chassis for a few minutes, then restared c1sus2. This solved the card recognition problem as well as the 'dis_irm' driver loading issue (I think the driver does not get loaded if the system does not recognise a valid card, as I also saw the missing dis_irm driver module on c1testand). 

Next, we confirmed the status of all dolphin cards on fb1, using

It looks like the dolphin card on c1sus2 has now been configured and is availabe to all other nodes. We then restated the all FE machines and models to see if we are in the clear. Unfortunately, we are not so lucky since the problem persisted.

Looking at the output of 'dmesg', we could only identity two notable difference between the operational dolphin cards on c1sus/c1ioo/c1lsc and c1sus2, namely: the card number being equal to zero and the memory addresses which are also zero, see image below.

Anyways, at least we can now eliminate driver issues and would move on to debugging the models next.