I confirmed that we need to vent the chambers.

All of the mirrors have been aligned except for ETMX.

ETMX does not respond to the excitation by the UR and LR coils. Likely that the magnets are knocked off, or stuck in the coil.

PRM/SRM oplevs are too much off and can't be turned on. Need realignment of the beams on the QPDs.

- FB was down. FB restarted ("telnet fb 8087", then type shutdown)

- Aligned the MC mirrors.

- Aligned PRM. Look at the REFL. It was slightly mislisligned.

- AS has no beam. The Y arm was resonating with the green. So I determined that the TTs were the misaligned guys.

- Touched TT2 pitch with an increment of 0.1. Immediately the AS beam spot for ITMY was found. And the arm was resonating.

- The RM was further aligned. The bias sliders were saved and then the PRM was misaligned.

- Yarm was locked on TEM01. The ASS maximized the transmission for TEM01, and then the arm was locked on TEM00.
  The ASS aligned the arm and TTs. These values were saved.

- Yarm was aligned and I can see the AS spot. So I believe the BS is still well aligned.

- Aligned the PRM to reduce the ghost beams.

- Moved the ITMX to have Michelson fringes properly.

- Also aligned the SRM.

- Now ETMX was checked. Played with the alignment biases to see if the mirror was sticking on the coils. The mirror can rock a little, but it did not come back.

- Then, checked each magnets. 0.8Hz 1000cnt signals were injected to each coils (cf. C1:SUS-PRM_**COIL_EXC) to see how the mirror could react.
  The OSEM output and green spot on the ETMX cage were observed.

- Saw some response by actuating the UL, LL, SD coils.

- Saw no response from the UR and LR coils. They show the OSEM output of zero. Does this mean the magnets fell in the coils?

//Manasa// MC spot positions measured and they look alright with not much change from before the earthquake (attach)

 KroneCrane Fred inspected and certified the 3 40m cranes for 2014. The vertex crane crane was load tested at fully extended position.

 It looks like that ETMX have  2 sticky magnets.

Asymptotic cooling of the mirror motion with CESAR was tested.

With ALS and the full control bandwidth (UGF 80-100Hz), the actuator amplitude of 8000cnt_pp is required.

Varying control bandwidth depending on the noise level of the signal, the arm was brought to the final configuration with the actuator amplitude of 800cnt_pp.

 Ha!

 Annual crane inspection is scheduled for 8-11am Monday, March 17, 2014

The control room Smart UPS has two red extension cords that has to be removed: Nodus and Linux1

[Jenne, EricQ]

As Koji suggested in his email this afternoon, we looked at how much actuator range is required for various forms of arm locking:  (1) "regular" PDH lock aquisition, (2) ALS lock acquisition, (3) CESAR cooling.

To start, I looked at the spectra and time series data of the control signal (XARM_OUT) for several locking situations.  Happily, when the arm is at the half fringe, where we expect the 1/sqrt(TRX) signal to be the most sensitive (versus the same signal at other arm powers), we see that it is in fact more quiet than even the PDH signal.  Of course, we can't ever use this signal once the arm is at resonance, so we haven't discovered anything new here.

EricQ then made some violin plots with the time series data from these situations, and we determined that a limit of ~400 counts encompasses most of the steady-state peak-to-peak output from locking on the PDH signal.

[ericq: What's being plotted here are "kernel density estimates" of the time series data of XARM_OUT when locked on these signals. The extent of the line goes to the furthest outlier, while the dashed and dotted lines indicate the median and quartiles, respectively]

I tried acquiring ALS and transitioning to final PDH signals with different limiters set in the Xarm servo.  I discovered that it's not too hard to do with a limit of 400 counts, but that below ~350 counts, I can't keep the ALS locked for long enough to find the IR resonance.  Here's a plot of acquiring ALS lock, scanning for the resonance, and then using CESAR to transition to PDH, with the limit of 400 counts in place, and then a lockloss at the end.  Even though I'm hitting the rails pretty consistently, until I transition to the more quiet signals, I don't ever lose lock (until, at the end, I started pushing other buttons...).

After that, I tried acquiring lock using our "regular" PDH method, and found that it wasn't too hard to capture lock with a limit of 400, but with limits below that I can't hold the lock through the boosts turning on.

Finally, I took spectra of the XARM_OUT control signal while locked using ALS only, but with different limiter values. Interestingly, I see much higher noise between 30-300 Hz with the limiter engaged, but the high frequency noise goes down.  Since the high frequency is dominating the RMS, we see that the RMS value is actually decreasing a bit (although not much).

We have not made any changes to the LSC model, so there is still no smoothing between the ALS and IR signals.  That is still on the to-do list.  I started modifying things to be compatible with CARM rather than a single arm, but that's more of a daytime-y task, so that version of the c1lsc model is saved under a different name, and the model that is currently compiled and running is reverted as the "c1lsc.mdl" file.

I have modified the IFOconfigure scripts and the corresponding .req files for the X arm and Y arm in burt. I have also added configure scripts to save and restore LSC settings for locking the arms using ALS error signals.

The settings are yet to be saved and the scripts should also be checked if they are working as required.

Q and I have started to use the ALS slow servo for the end aux lasers while locking the arms using ALS. The servo prevents us from hitting the limits of the PZT range for the end lasers and a better PDH locking.

But keeping the servo ON causes the slow output to drift away making it hard to find the beat note everytime the arm loses lock. The extensive beat note search following the unlock can be avoided by clearing history of the slow servo.

Speaking of the whitening board, I had neglected to post details showing the the whitening was at least having a positive effect on the transmon QPD noise. So, here is a spectrum showing the effects that the whitening stages have on a QPD dark noise measurement like I did in ELOG 9660, at a simulated transmission level of 40 counts. 

The first whitening stages gives us a full 20dB of noise reduction, while the second stage brings us down to either the dark noise of the QPD or the noise of the whitening board. We should figure out which it is, and fix up the board if necessary. 

The DTT xml file is attached in a zip, if anyone wants it.

[Nic, Jenne, EricQ, and Koji]

We have used CESAR successfully to bring the Xarm into resonance.  We start with the ALS signal, then as we approach resonance, the error signal is automatically transitioned to 1/sqrt(TRX), and ramped from there to POX, which we use as the PDH signal.

In the first plot, we have several spectra of the CESAR output signal (which is the error signal for the Xarm), at different arm resonance conditions.  Dark blue is the signal when we are locked with the ALS beatnote, far from IR resonance.  Gold is when we are starting to see IR resonance (arm buildup of about 0.03 or more), and we are using the 1/sqrt(TRX) signal for locking.  Cyan is after we have achieved resonance, and are using only the POX PDH signal.  Purple is the same condition as cyan, except that we have also engaged the low frequency boosts (FM 2, 3, 4) in the locking servo.  FM4 is only usable once you are at IR resonance, and locked using the PDH signal.  We see in the plot that our high frequency noise (and total RMS) decreases with each stage of CESAR (ALS, 1/sqrt(TR) and PDH). 

To actually achieve the gold noise level of 1/sqrt(TR), we first had to increase the analog gain by swapping out a resistor on the whitening board. 

The other plots attached are time series data.  For the python plots (last 2), the error signals are calibrated to nanometers, but the dark blue, which is the transmitted power of the cavity, is left in normalized power units (where 1 is full IR resonance). 

In the scan from off resonance to on resonance, around the 58 second mark, we see a glitch when we engage FM4, the strong low frequency boosts.  Around the 75 second mark we turned off any contribution from 1/sqrt(TR), so the noise decreases once we are on pure PDH signal. 

In the scan through the resonance, we see a little more clearly the glitch that happens when we switch from ALS to IR signals, around the 7 and 12 second marks. 

We want to make some changes, so that the transition from ALS to IR signals is more smooth, and not a discrete switch.

Today we modified the CESAR block.

- Now the sign(X) function is in a block.

- We decided to use the linearization of the PDH.

- By adding the offset to the TR signal used for the switching between TR and PDH, we can force pure 1/sqrt(TR) or pure PDH to control the cavity.

True. But we first want to realize this for a single arm, then move onto the two arms case.
At this point we'll need to incorporate frequency dependence.

As part of our CESAR testing last night, we had a look at the noise of the 1/sqrt(TR) signal. 

Looking at the time series data, while we were slowly sweeping through IR resonance (using the ALS), Rana noted that the linear range of the 1/sqrt(TR) signal was not as wide as it should be, and that this is likely because our SNR is really poor. 

When a single arm is at a normalized transmission power of 1, we are getting about 300 ADC counts.  We want this to be more like 3000 ADC counts, to be taking advantage of the full range of the ADC.

This means that we want to increase our analog gain by a factor of 10 for the low gain Thorlabs PDs. 

Looking at the photos from November when I pulled out the Xend transmission whitening board (elog 9367), we want to change "Rgain" of the AD620 on the daughter board.  While we're at it, we should also change the noisy black thick film resistors to the green thin film resistors in the signal path. 

The daughter board is D04060, S/N 101.  The main whitening board for the low gain trans QPD is D990399, RevB, S/N 104.

We should also check whether we're saturating somewhere in the whitening board by putting in a function generator signal via BNC cable into the input of the Thorlabs whitening path, and seeing where (in Dataviewer) we start to see saturation.  Is it the full 32,000 counts, or somewhere lower, like 28,000? 

Actually the gain was changed. From gain of 2 (Rgain = 49.4kOhm) to 20 (Rgain = 2.10kOhm), Corresponding calibration in CDS was also changed by locking the Xarm, running ASS, then setting the average arm power to be 1. Confirmed Xarm is locking. And now the signal is used for CESAR.  We see emperically that the noise has improved by a factor of approximately 10ish.

 Very nice error signal. Still, I think we need to take into account the frequency shape of the transfer function TR -> CARM. 

Two weeks ago (Feb 26) I took the "Q MON" output of the demodulator that sends its Q output to the MC servo board as the error signal, and connected it to an SR785, so we can occasionally monitor the error signal noise. (Also, I did not appropriately ELOG the fact I touched things...)

I'm working on an automated script to do the monitoring, but the wireless router that the SR785 is connected is wicked slow. I should run an ethernet cable to it...

I'm just figuring I'll look at the full span (~100kHz) spectrum every ten minutes, and compare it to some nominal spectrum from a known-good time, and the last few hours.

The nominal gain of the XARM for the POX11 error signal is 0.03 (instead of previous 0.3)

This is due to my increase of the gain in FM4 by 20dB so that we can turn of FM4 without changing the UGF when it is at 150Hz.

The YARM servo was not yet touched.

After making the CDS modification, CESAR was tested with ALS.

First of all, we run CESAR with threshold of 10. This means that the error signal always used ALS.
The ALS was scanned over the resonance. The plot of the scan can be found in EricQ's elog.
At each point of the scan, the arm stability is limited by the ALS.

Using this scan data, we could adjust the gains for the TR and PDH signals. Once the gains were adjusted
the threshold was lowered to 0.25. This activates dynamic signal blending.

ALS was stabilized with XARM FM1/2/3/5/6/7/9. The resonance was scanned. No glitch was observed.
This is some level of success already.

Next step was to fully hand off the control to PDH. But this was not successfull. Everytime the gain for the TR was
reduced to zero, the lock was lost. When the TR is removed from the control, the raw PDH signal is used fot the control
without normalization. Without turning on FM4, we lose 60dB of DC gain. Therefore the residual motion may have been
too big for the linear range of the PDH signal. This could be mitigated by the normalization of the PDH signal by the TR.

The LSC model was modified for CESAR.

A block called ALSX_COMBINE was made in the LSC block. This block receives the signals for ALS (Phase Tracker output), TRX_SQRTINV, TRX, POX11 (Unnormalized POX11I).
It spits out the composite ALS signal.

Inside of the block we have several components:

1) a group of components for sign(x) function. We use the PDH signal to produce the sign for the transmission signal.

2) Hard triggering between ALS and TR/PDH signals. An epics channel "THRESH" is used to determine how much transmission
we should have to turn on the TR/PDH signals.

3) Blending of the TR and PDH. Currently we are using a confined TR between 0 and 1 using a saturation module. When the TR is 0, we use the 1/SQRT(TR) signal for the control,
    When the TR is 1, we use the PDH signal for the control.

4) Finally the three processed signals are combined into a single signal by an adder.

It is important to make a consideration on the offsets. We want all of ALS, 1/SQRT(TR), and PDH to have zero crossing at the resonance.
ALS tends to have arbitorary offset. So we decided to use two offsets. One is before the CESAR block and in the ALS path.
The other is after the CESAR block. Right now we are using the XARM servo offset for the latter purpose.

We run the resonance search script to find the first offset. Once this is set, we never touch this offset until the lock is lost.
Then for the further scanning of the arm length, we uused the offset in the XARM servo filter module.

Using Koji's mathematica notebook, and Nic's python work, I set out to run a time domain simulation of the error signal, with band-limited white noise added in. 

Basically, I sweep the displacement of the cavity (with no noise), and pass it to the analytical formulae with the coefficients Koji used, with some noise added in. I also included some 1/0 protection for the linearized PDH signal. I ran a sweep, and then compared it to an ALS sweep that Jenne ran on Monday; reconstructing what the CESAR signal would have looked like in the sweep. 

The noise amounts were totally made up. 

They matched up very well, qualitatively! [Since the real sweep was done by a (relatively) noisy ALS, the lower noise of the real pdh signal was obscured.]

Given this good match, we were motivated to start trying to implement it on Monday. 

At this point, since we've gotten it working on the actual IFO, I don't plan on doing much more with this simulation right now, but it may come in handy in the future...

In order to better understand how the composite signal would behave in the presence of noise, I decided to do a simple analysis of the cavity signals while sweeping through resonance.

My noise model was to just assume that a given signal has some rms uncertainty (error bars) and use linear error propagation to propagate from simple signals to more complicated ones.

I used the python package uncertainties to do the error propagation.

I assumed that the ALS signal, the cavity transmission, and the cavity PDH error signal all have some constant noise that is independent of the cavity detuning. Below is a sweep through resonance (x axis is cavity detuning in units of radians).

The shaded region represents the error on each signal.

Next I calculated the 'first order' calculated error signals. These being a raw PDH, normalized PDH, an inverse square root trans, and the normal ALS again. I tuned the gains so they match appropriately.

Here, one can see how the error in the trans signal propagates to the normalized and trans signals and becomes large are the fractional error in the trans signal becomes large.

Next I did some optimization of linear combinations of these signals. I told the code to maximize the total signal to noise ratio, while ensuring that the overall signal had positive gain. I did this again as a function of the cavity detuning.

Each curve represents the optimized weight of the corresponding signal as a function of detuning.

So this is roughly doing what we expect, it prefers ALS far from the resonance, and PDH close to the resonance, while smoothly moving into square root trans in the middle.

It's a little fake, but it gives us an idea of what the 'best' we can do is.

Finally I used these weights to recombine the signals into a composite, to get an idea of the noise of the overall signal. At the same time, I plot the weighting proposed by Koji's mathematica notebook (using trans and 1-trans, and a hard switch to ALS).

So as one can see, at least for the noise levels I chose, the koji weighting is not much worse than the 'optimal' weighting. While it is much simpler.

The code for all this is in the svn at 40mSVN/nicolas/workspace/2014-03-06_compositeerror

The ALS error (i.e. phase tracker output) is linear everywhere, but noisy.
The 1/sqrt(TR) is linear and less noisy but is not linear at around the resonance and has no sign.
The PDH signal is linear and further less noisy but the linear range is limited.

Why don't we combine all of these to produce a composite error signal that is linear everywhere and less-noisy at the redsonance?

This concept was confirmed by a simple mathematica calculation:

The following plot shows the raw signals with arbitorary normalizations

1) ALS: (Blue)
2) 1/SQRT(TR): (Purple)
3) PDH: (Yellow)
4) Transmission (Green)

The following plot shows the preprocessed signals for composition

1) ALS: no preprocess (Blue)
2) 1/SQRT(TR): multiply sign(PDH) (Purple)
3) PDH: linarization with the transmission (If TR<0.1, use 0.1 for the normalization). (Yellow)
4) Transmittion (Green)

The composite error signal

1) Use ALS at TR<0.03. Use 1/SQRT(TR)*sign(PDH)*(1-TR) + PDH*TR at TR>0.03
2) Transmittion (Purple)
 

I don't know why, but as you can see in Steve's plot from earlier this morning, the PMC transmission has been going down significantly all weekend.  The PMC refl camera was very bright.  I tweaked up the alignment (mostly pitch), and now we're back to normal. 

The IMC was having trouble staying locked all morning, and I'm hoping that this PMC adjustment will help - the MC already looks better, although it's only been a few minutes.

fb timing was off again.

 Morning seconds without adjustment.

Pd 76 and 77 are compared at 30 days of pumping. We spend 13-14 days at atmosphere  before each.

 Pump down 76 was with leaky ion pump gate valve. The ion pumps are not in use for years so they accumulated some higher pressure  PLUS the valve switching caos at computer reboot most likely

increased the ion pumps pressures to about 10-20 torr

I think one of the ion pump gate valve was not sealing well. This leak was holding back pump down speed at pd76

[EricQ, Manasa]

ALS common locked by actuating on MC2 and ALS Differential locked by actuating on ETMX and ETMY (Stable lock acquired for over an hour).

Common and Differential offsets were swept to obtain IR resonance in both the arms (arms stayed on resonance for over 15 minutes).

Procedure:

1. Configured LSC settings to allow locking using ALS error signals.

2. Locked common and differential using ALS error signals

Aux matrix
              ALSX    ALSY
------------------------------
XARM    1            -1
YARM    1              1
-----------------------------
X arm servo settings:
FIlters: FM1, FM5, FM6, FM7, FM9
Gain = -8.0

Y arm servo settings:
Filters: FM1, FM5, FM6, FM7, FM9
Gain = +8.0

Out matrix
    XARM    YARM
------------------------
ETMX    1    0
ETMY    0    1
------------------------

3. Transitioned CARM control output to actuate on MC2 instead of ETMX

SUS-MC2_LSC servo gain = 1.0
The transition was done in very small steps : actuating on MC2 in -0.01 steps at the outmatrix upto -1.0 while reducing the ETMX actuation to 0 simultaneously.

DARM still stayed locked only with actuation on ETMY.

4. Transitioned DARM control to ETMX and ETMY.

Used ezcastep to step up DARM control (Y arm output) actuation on ETMX and step down the actuation on ETMY.

Final output matrix
    Xarm    Yarm
-----------------------
ETMX      0    -0.5
ETMY      0     0.5
MC2    -1.0      0
-----------------------

Noise plot in attachment.

5. Finding arm resonance

Used ezcastep to gradually build up offsets in CARM (LSC-XARM_OFS) to find IR resoance in one arm (Y arm).
Introducing a small (order of 0.5) DARM offset (LSC-YARM_OFS) shifted the Y arm off-resonance.
Used CARM offset to get back the Y arm to resonance.
Changing CARM and DARM offsets alternately while tracking the Y arm resonance got us to a point where we had both the arms resonating for IR.

6. At this point the MC decided to give up and we lost lock.

Notes:
1. We found that the WFS2 YAW output filterbank had the output switched OFF (probably accidentally by one of us). This was reenabled. Please be careful while manually turning ON and OFF the MC WFS servos.

MC remained locked with WFS enabled all through last night and this morning. Koji dropped by and looked at the MC. The MC WFS servo, though stable, was at the edge of becoming unstable. This was because I did not touch the WFS pointing on the QPDs yesterday after realigning. So I recentered the WFS, reset the WFS filterbank offsets and reenabled the servo.

I measured the spot positions on MC mirrors for reference.

Spot positions in mm (MC1,2,3 pit MC1,2,3 yaw): [1.405767579680834, 0.79369009503571208, 1.3220430681427462, -1.2937873599406551, -1.1704264340968924, -1.2518046122798692]

 In addition, we have to make sure to not let the suspension DACs saturate and make sure that the impulse response time of the OL servo is short; otherwise the lock acquisition kicks or bumps can make it wiggle for too long.

CMG-40T handheld controller unit is missing its power supply. In order to zero the instrument one has to apply plus and minus DC voltage.

The wiring on this 10 pin Amphenol PT02E-12-10P is shown.

You don't need to make transition from ALS X/Y to ALS C/D. Just stabilize the arms with ALS C/D from the beginning.

I locked the arms using ALS error signals and the LSC filter modules. But when I try to acquire CARM and DARM using ALS, the arms lose lock when the matrix elements ALSX to Yarm and ALSY to X arm reach -/+0.9

What I did:

1. ALS locking of arms
(i) Found arm beat notes
(ii) Input matrix POX and POY elements set to '0'
(iii) Aux matrix elements ALSX to Xarm and ALSY to Y arm set to '1'
(iv) Power normalization matrix elements for TRX and TRY set to '0'
(v) Triggers for arm lock over ridden and the FM triggers were set to 'manual'
(vi) Arm servo gains set to '0'
(vii) All but FM5 were disabled
(viii) Phase tracker history reset and servo actuation set to ETMs
(ix) Servo gain increased in steps (+/-10 for the arms)
(x) FM1, FM6, FM7 enabled (see note 1 below)
(xi) FM9 enabled

Arms were locked with ~2000Hz rms

2. CARM and DARM locking
(i) Scanned the arms for IR resonance
(ii) Moved off-resonance (Stepped arm servo offsets by 30 counts)
(iiI) Stepped matrix elements ALSY to X arm and ALSX to Y arm ezcastep C1:LSC-PD_DOF_MTRX_6_29 +-0.1 C1:LSC-PD_DOF_MTRX_7_28 +0.1

Whenever the matrix elements reached -/+0.9, the arms were kicked out of lock. I don't see anything obvious as to why this is happening even after nearly 10+times of redoing.

Notes:
1. I found the filters for the arm servos different for X and Y. FM1 and FM8 were missing in one of the filter modules. Jenne remembered Rana modifying and removing the unnecessary filters in one arm. We put back FM1 (low pass filter) which might not be necessary for PDH lock but is necessary for ALS. FM8 is now added to FM7.
2. To self : Check ALS Y arm power outlets (60Hz frequency comb seen in the error signal)

The IMC has not been behaving well since this morning and totally not happy when Q was finishing his measurements. The WFS servo had large offsets in pitch. Looking back at the trend and using ezcaservo to restore the suspensions did not help.

I realigned the IMC and brought TRANS SUM to ~18000 and MCREFL to < 0.5. The spot positions are not very good; nearly 2 mm off in pitch on MC1 and MC3. But after the alignment of MC, the WFS servo offsets were below +/-20.

The MC has been locked stably with WFS servo ON for the last few hours.

P.S. I did not touch the WFS pointing or reset the WFS offsets.

 I have modified the script ALSchangeOffsets.py (in ..../scripts/ALS/) to also handle a "CARM" situation.  There is a new button for this on the ALS in LSC screen.  This script takes the desired offset, and puts half in the ALSX offset, and half in the ALSY offset.  Whatever offset you ask for is given the sign of the input matrix element in the ALS->CARM row of the input matrix.  For example, if you ask for a CARM offset of 1, and the matrix elements are ALSX->CARM=+1 and ALSY->CARM=-1 (because your beatnotes are on opposite sides of the PSL), you will get an offset of +0.5 in ALSX and -0.5 in ALSY, which should be a pure CARM offset. The offsets get set as expected, but I haven't had a chance to test it live while the arms are locked. 

I also want to write a script that will average the IN1 of the 1/sqrt(TR) signals, and put that number into the 1/sqrt(TR) offsets.  If this is run when we are at about half fringe, this will set the zero point of the 1/sqrt(TR) signals to the half fringe (or where ever we are).  Then, we need a script similar to the ALS CARM one, to put offsets into the CARM combination of 1/sqrt(TR)s. 

I think that putting the offsets in before the servo filters will mean that the signals coming out of the input matrices will already be at their zero points, so we won't have as much trouble shifting from ALS to IR.

At today's meeting, it was suspected that these noise levels were far too low. (ELOG 9660)

I've attached the math I did to get the conversions, as well as the dark noise SQRTINV spectra at various imitated transmission values and the python script that does the converting. 

I've gone over my calculations, and think they're self-consistent. However, a potential source of misestimation is the treatment of the Lorentzian profile simply existing with the coupled arm line width (38pm). The conversion to m/rtHz is directly proportional to the line width of the transmission peak, so if it is much broader in practice (because of imperfect PRC buildup or something), the noise will be that much worse.

I'm open to any other feedback about what I may have done wrong!

I measured the arm cavity losses as Kiwamu did way back in ELOG 5074.

I used the same logic as the ../scripts/LSC/armloss script, but did it manually. This meant:

1. Lock and ASS-Align both arms. 
2. Misalign the ITM of the arm that I'm not measuring, to get its spot off of AS
3. Take 10 seconds of ASDC_OUT data while the arm is locked. 
4. Unlock, misalign ETM of arm of interest, take another 10 seconds of ASDC_OUT
5. Relock, run ASS, goto #3

Analysis was done similar to ../scripts/LSC/armloss.m. This uses the nominal T values (.014 and 15e-6) to estimate the input power from the unlocked ASDC data, and the cavity reflectivity from the locked ASDC / input power. Then, loss is calculated by:

• Pin = ASDC(unlocked) / R1
• Rc = ASDC(locked) / Pin
• rc=sqrt(Rc), etc.
• Loss = 1 - (( 1 / r1r2)) * ( 1 - t1^2 r2 / (r1 - rc)) ^2

I did this for pairs of locked / unlocked data stretches. (Subsequent pairs maybe have slightly different things going on, but each pair was taken within a minute or so of each other)

Unfortunately, during the X Arm measurements, the MC was misbehaving with large REFL fluctuations, so I don't have confidence the results.

The Y Arm data seems fine, however. 

The Y arm loss is 123.91 +/- 10.47 ppm 

(Trial-to-Trial fluctuations dominate the fluctuations within each trial by far, and their standard deviation is what I report as the random error above)

This seems roughly in agreement with old values I've seen in the ELOG. I'll remeasure the x arm tomorrow during the day. Here's a plot showing the ASDC values of the Y Arm measurements. 

Step by step description of transition from 2arm ALS to Common/Differential LSC for FPMI

- Step 0: Place the frequencies of the arm green beams at the opposite side of the carrier green.

- Step 1: Activate stablization loops for ALSX and ALSY simultaneously.
  (Use LSC filter modules for the control. This still requires correct handling of the servo and filter module triggers)

- Step 2: Activate stablization loops for ALS Common and Differential by actuating ETMX and ETMY

- Step 2 (advanced): Activate stabilization loops for ALS Common by actuating MC2 and ALS Differential by ETMX and ETMY

- Step 3: Transition from ALS Common to 1/SQRT(TRX)+1/SQRT(TRY). Make sure that the calibration of TRX and TRY are matched.
  The current understanding is that the offset for 1/SQRT(TRX)+1/SQRT(TRY) can't be provided at the servo filter. Figure out
  what is the correct way to give the offsets to the TR signals.

- Step 4: Lock Michelson with AS55Q and then POP55Q (PD not available yet) or any other PD, while the arms are kept off-resonant using ALS.

- Step 5: Reduce the TR offsets. Transition to RF CARM signals obtained from POP55I or REFL11I in the digital land.

- Step 5 (advanced): Same as test6 but involve the analog common mode servo too.

- Step 6: Transition from ALS Differential to AS55Q

Independent test: One arm ALS (To be done everyday)

- ALS resonance scan

- Measurement of out-of-loop displacement (or frequency) stability 

- Check openloop transer function

Independent test: Common Mode servo for one arm

- Reproduce Decmber CM servo result of transition from one arm ALS to CM servo
  Insert 1/sqrt(TRY) servo in between?

- How can we realize smooth transition from ALS to POY11?

Not a whiteboard, but here's a cartoon of my oplev cost function cartoon.  For the "maximize this area" and "minimize this area", I plan to use ratios between the curves, and then give those ratios to a sigmoid function.

 Keven, Steve

1X6, 1X7 and 1X9 instrument racks floor space were cleaned today

 I was getting the Y Arm ready for Eric Q's loss measurements and so I looked at the noise and loop shape. The loop shape was strange:

You can see that the gain margin is too low at high frequencies. That's why we have >15 dB of gain peaking. Way too much! I think this is from Masayuki and Manasa increasing the phase margin at some point in the past. I lowered the gain by 3 dB from 0.1 to 0.07 and now the awful gain peaking is less. But what about the low frequency gain? Is there enough?

I calibrated the OUT channel with 14 nm/count (1/f^2) with a Q = 10 pole pair at 1 Hz. The error signal is done to cross over at 180 Hz. It looks like the resonant gain at 25 Hz is a little too much and the in-loop RMS is 10 pm. Jenne says the linewidth is ~1 nm, so this seems sort of OK. Except that the LIGO-I DARM RMS had to be <0.1 pm for ~the same linewidth. Do we need to do better before trying to bring the arms into resonance?

I've remove FM1 and FM8. I put the RollRG of FM8 into the BounceRG and renamed it BounceRoll. Also changed the Y-arm restore so that RollRG and the 5,5:0,0 are no longer triggered automatically since the double integrator was overkill and we already have a 1:0 in FM2. I also lowered the peak gain for the roll mode RG from 30 to 10 dB because it was also overkill. We've gained a few more degrees at the UGF.

Successful PRMIsb locking with REFL165I/Q

My previous entry suggested that somehow the REFL165 signals show reasonable separation between PRCL and MICH, contrary to our previous observation.
I don't know what is the difference now. But anyway I took this advantage and tried to lock sideband resonant PRMI.

REFL165I was adjusted so that the signal is only sensitive to PRCL. Then REFL165I and Q were mixed so that the resulting signal shows.
(Next time, we should try to optimize the Q phase to eliminate PRCL and just use the I phase for PRCL.

At first, I used AS55Q for lock acquisition and then switched the MICH input matrix to REFL165.
Later I found that I can acquire PRMI just turning on AS55Q without turning off REFL165.

The REFL165 MICH signal had an offset of 15cnt. The lock was more robust and the dark port was darker once the MICH input offset was correctly set.

MICH OFS = 0
Turn on AS55Q only / or AS55Q + REFL156I/Q
Once it is locked and all of the FMs are activated, give -15.0OFS to MICH.
Turn off AS55Q.

Input ports:
AS55       WHTN: 21dB  demod phase -5.5deg
REFL165 WHTN: 45dB demod phase -156.13deg

Input matrix:
AS55Q x1.00 MICH
REFL165I x-0.035 + REFL165Q -0.050 MICH

REL165Q x+0.14

Triggers:
MICH POP110I 100up/10down / FM Trig FM2/3/6/7/9 35up 2down 5sec delay
PRCL POP110I 100up/10down / FM Trig FM2/3/6/9 35up 2down 0.5sec delay

Servo:
MICH OFS -15.0 / Gain -10 / Limitter ON
PRCL OFS 0 / Gain -0.02 / Limitter ON

Output matrix:
MICH ITMX -1.0 / ITMY +1.0
PRCL PRM 1.0

I have installed Dropbox on the 40m workstations, using the foteee account. 

I put it in /users/Dropbox.

As a side note, I did the install while sitting on Pianosa, but since I put the folder on the mounted hard drive, I think we should be able to use it from any workstation, although I have not yet confirmed this.

I wanted to check how the refl signals looked like.
I decided to measure the demod phase where PRCL and MICH appear, one by one.

The method I used is to actuate PRCL or MICH at a fixed frequency and rotate the demod phase such that
the signal at the actuating frequency disappears.

For the PRCL actuation, PRM was actuated by the lock-in oscillator with the amplitude of 100cnt.
For MICH, the ITMX and ITMY was actuate at the amplitude of 1000cnt and 1015cnt respectively.

The script I used was something like this

ezcaread C1:LSC-REFL11_PHASE_R
ezcaservo -r C1:CAL-SENSMAT_CARM_REFL11_Q_I_OUTPUT C1:LSC-REFL11_PHASE_R -g 100 -t 60
ezcaread C1:LSC-REFL11_PHASE_R

"11" should be changed according to the PD you want to test.
"Q" should be changed to "I" depending on form which quadrature you want to eliminate the signal

The option "-g" specifies the servo gain. This specifies which slope (up or down) of the sinusoidal curve the signal is locked.
Therefore, it is important to flip the signal angle 180degree if a negative gain is used.

Note: Original phase settings before touching them

REFL11  - 19.2
REFL33   135.4
REFL55    48.0
RELF165 -118.5

Here in the measurement PRMI was locked with AS55Q (MICH) and REFL55I (PRCL)


Without no serious reason I injected a peak at 503.1Hz. This peak is not notched out by the servo. There may have been
some residual effect of the feedback loops.

PRCL: By elliminating the peak from the Q quadrature, we optimize the I phase for PRCL.

REFL11,   minimize PRCL in "Q", gain, -1, -19.3659 deg
REFL33,   minimize PRCL in "Q", gain, -1, 132.813 deg
REFL55,   minimize PRCL in "Q", gain, -1, 20.9747 deg
REFL165, minimize PRCL in "Q", gain, -1, -119.004 deg

MICH: By elliminating the peak from the I quadrature, we optimize the Q phase for MICH.
If PRCL and MICH appears at the same phase, the resulting angles shows an identical number.

REFL11,   minimize PRCL in "I", gain, -1, -28.4526 deg
REFL33,   minimize PRCL in "I", gain, -1, 65.9148 deg
REFL55,   minimize PRCL in "I", gain, -1, 12.4051 deg
REFL165, minimize PRCL in "I", gain, -0.1, -143.75 deg

Then, the signal frequency was changed to 675Hz where the notch filters in the servo is active.

PRCL: By elliminating the peak from the Q quadrature, we optimize the I phase for PRCL.

REFL11,   minimize PRCL in "Q", gain, 1, -19.5224 deg
REFL33,   minimize PRCL in "Q", gain, -1, 135.868 deg
REFL55,   minimize PRCL in "Q", gain, 1, 48.5716 deg
REFL165, minimize PRCL in "Q", gain, 1, -122.398 deg

MICH: By elliminating the peak from the I quadrature, we optimize the Q phase for MICH.
If PRCL and MICH appears at the same phase, the resulting angles shows an identical number.

REFL11,   minimize PRCL in "I", gain, -10, -73.7153 deg
REFL33,   minimize PRCL in "I", gain, -10, 135.5 deg
REFL55,   minimize PRCL in "I", gain, 10, -2.55868 deg
REFL165, minimize PRCL in "I", gain, -5, -156.135 deg

This is just a test of the REFL channels for the arms signals. ETMX or ETMY were actuated.

YARM

REFL11, minimize ETMY in "Q", gain 100 => C1:LSC-REFL11_PHASE_R = 145.694
REFL55, minimize ETMY in "Q", gain 100 => C1:LSC-REFL11_PHASE_R = -60.1512

XARM

REFL11, minimize ETMX in "Q", gain 100 => C1:LSC-REFL11_PHASE_R = 142.365
REFL55, minimize ETMX in "Q", gain 100 => C1:LSC-REFL55_PHASE_R = -68.6521

We need to correctly setup crontab or rc.local for the frontend machines.

...yet again.

lsc and sus needed mxstream restarts after I restarted the ntp on fb.

  in order to Win in Loop Tuning, you must draw a cartoon of the cost function on the whiteboard before starting. Some qualitative considerations from our Workshop:

1. We want to use the oplev servo to reduce the motion of the mirror in the frequency band where the Oplev is quieter than the mirror, w.r.t. inertial space.
2. We can estimate the true mirror motion by some simple stack / pendulum model and compare it to the Oplev noise (not the dark noise). There are several contributions to the mirror angular motion due to the cross-coupling in the stacks and pendula.
3. Below ~0.2 Hz, we think that the oplev is not the right reference, but this is not quantitative yet.
4. The high frequency noise in the OPLEV ERROR is definitely electronics + shot noise.
5. We cannot increase the gain of the loop without posting some loop measurements (Bode + steps). Also have to post estimates of how much PRCL noise is being introduced by the Oplev feedback. Oplev feedback should make less length noise than what we have from seismic.

Give us a cost function in the elog and then keep tuning.