Addressed TCS Chillers (09:15-09:21 AM PST today)

• TCSx Chiller:  Had level of 29.0 cm & topped off to 30.0cm by adding 75 mL.
• TCSy Chiller:  Had level of 9.5cm & topped off to 9.9cm by adding 225mL.

Koji, Craig

Tonight we investigated intensity noise further.  We locked the OMC on just the 45 MHz at 30 watts with 18 dB whitening gain, and then locked it on carrier at 20 W with 0 dB whitening gain.  This is a repeat of here.

We found that carrier and 45 MHz noise levels are similar.  (See attachment one, carrier RIN is Blue, 45 MHz RIN is Brown)

We were confused about why 100 and 1000 Hz the carrier noise is lower.  We believe that this is mostly a shot noise effect.  
For 45 MHz we have 0.5 mA on the OMC DCPDs, while for carrier we had 34.3 mA.  This corresponds to RIN shot noise levels of 2.5 × 10-8 1/rtHz and 3.1 × 10-9 1/rtHz.
- The carrier is mostly limited by the ISS second loop sensing noise, as seen from the green curve in attachment one.  
- The 45 MHz flattens out right around our shot noise limit around 100 Hz.
- Both carrier and 45 MHz have similar RIN spectra below 100 Hz.

Recap:
We know that the 9 MHz modulation depth has a real effect on the DARM noise, seen here as well.  During the holiday party we were able to improve DARM noise by increasing the DARM offset, which could be due to increased carrier drowning out the 9 MHz noise.
We also know that our OMC measured RIN in the sidebands cannot be explained from the noise in our RF AM stabilization scheme, or the ISS.
We are pretty sure that the IMC FSR is reasonably close to our 1f modulation frequency.

Conclusions:
We here note that our shot noise from the OMC sideband measurements is not good enough from 80 Hz onwards to be able to tell what our actual 45 MHz sideband RIN noise level is.  
Below 100 Hz, carrier and 45 MHz RIN seem to be limited by the same mechanism, we don't know what it is.  
9 MHz is not ever limited by shot noise in the OMC RIN measurement, and is overall worse than both carrier and 45 MHz RIN after 20 Hz, where it starts falling much more slowly. (Attachment three)

-------------------------------------------------------------------------------------------------------------

Past OMC carrier RIN measurement

We took some CHARD, DHARD P measurements tonight, neither seem responsible for the 1 Hz oscillation reported earlier.
We lost lock due to a 6.4 earthquake five hours ago.

TVo, Sheila, Koji and Jonathan

Today ASC required our attention, so we spent some time on it.  It looks like the combination of moving the spot position on ETMY and phasing AS72 have helped our build ups when the ASC is engaged at 2W, and perhaps the drop off in POP18 is a bit better. 

Dither Phases:

• We found that the sign of the Pit4 dither servo (ETMX->DARM-Xarmsoft feedback) flipped after several minutes at 20W, flipped again as we powered up to 30W, and again after several minutes at 30W. 
• In the next lock we decided to check the demod phase for the dithers at 2W DC readout.  For PIT4 we changed the phase from -25 to +40 degrees by changing the P2L gain and rotating the response into I, we then had to reduce the loop gain from 1 to 0.5.  We also changed the phase on YAW4 from -25 to +20, and reduced it's gain by half.  
• PIT3 (ITMY-> DARM -> PRM) demod phase seemed fine, YAW3 we rotated from -25 to +20. 
• At 2 W, both ETMY dithers had offsets in the Q phase, so that if rotated the phases to maximize the response in I, we would introduce and offset into the loop, so we did not make any changes to those.  However, once we powered up to 30W we saw that the PIT5 loop ran away, and that we actually had better PRC and POP18 build ups for a different spot position (ETMY P2L changed from 2.6 to 3.45).  We checked the phase on PIT5 demod again at this new spot position, and saw that both the Q offset and the response in Q were much smaller than what we had seen at 2 W. We didn't get to explore the spot positions more in that lock because we lost lock either due to an EQ or running away with PIT5.  

Phasing AS72:

• We also noticed that when we engage ASC at 2W, we have been loosing POP18 build up, so Thomas and I moved offsets in SRC1 until we recovered most of POP18 and the POP90 was lower.  We opened the loops and rotated the phased on AS72A so that the DC signals were all in the I phase, then reclosed the loop.  
• With these changes and the change in position on ETMY, we are now increasing our build ups instead of decreasing them when we engage ASC, and it seems that we are not loosing as much POP18 when we power up.  

Engaging ASC/ no more POP QPD feedback:

• We had one lockloss in ENGAGE_ASC when the PRC1,2 and SRC1 gains are increased, which looked to be due to the large transient when we increase the gain in PRC1.  This seems similar to the reason Georgia added a convergence checker 46554
• We are no longer using feedback from the POP QPD to PRM, and using only the dither loop.  We realized that what we are doing with the feedback to PRM didn't make much sense, we were engaging AC coupled feedback from the QPD at the same time as the wavefront sensor loops are engaged, then increasing the gain, then latter turning on the dither feedback when we engaged the soft loops, then turning off PRC1Y after the soft loops are on.  We decided to try leaving out the QPD feedback to PRM completely, this was fine for the engagement and power up to 20W. 
• We did some cleanup of the code in ENGAGE_ASC, moving some set up for the soft loops to PREP_ASC and getting rid of some repetitive or unnecessary steps. 

LOWNOISE ASC:

• We lost lock 3 times due to a 1 Hz instability that looked like DHARD P after going through LOW NOISE ASC.  We manually stepped through the steps in LOWNOISE_ASC the second and third time, and it seems to be engaging the DHARDP 10Hz low pass caused the instability, so we have edited the guardian to use the less aggressive 17 Hz low pass.  The problem seems to be one of the steps that happens before the DHARD_P cut off filte.  We'd recomend stepping through LOWNOISE_ASC manually next time we try to lock, the instability doesn't ring up immediately after any change, so it may be necessary to wait a few minutes at each step to see what the problem is.  
• The code in lownoise asc was a bit hard to read because of lots of unnecessary logic, we have removed some of this including logic for what to do if the input power is above 45W and counter steps that did nothing, the remaining code could still be consolidated. 
•  Hidden in some of this confusing code was a step where we lowered the gain of DSOFT Y, which was accidentally turning the QPD feedback on for DSOFT Y, which we aren't using.  We have taken it out.

PRMI to DRMI transition:

We had an accidental DRMI lock where SRM was badly mis-alinged, and we were able to walk SRM back in without immediately losing lock.  This happened at 0:59 UTC on Jan 22nd, PRCL gain was 8 and MICH gain was 2.8.

I've looked at most of the power ports of the interferometer to try to get a feel for how IFO thermalization is affecting power levels.

I looked at the time series of the PSL power, PR gain, Arms Transmission, REFL A/B LF, POP LF, and POP 18/90 I/Q from the beginning of a Jan 17 INCREASE POWER to NOMINAL LOW NOISE state transition.
- We reduced 45 modulation depth by 3 dB (from 27 to 24 dB on the 45 RFSET slider) at around 200 seconds at 20 W input power. 
- We did NOT reduce RF 9 modulation depth.
- Within the first 180 seconds, the TCS CO2 X moved from 0.81 to 0.62 W, and TCS CO2 Y moved from 0.86 to 0.07 W.

In attachment two, I fit a line to each of the early-stage power up numbers (from the first 150 seconds) to indicate what a linear response to power increases ought to look like.
When powering up from 2 to 30 W:
- The arm transmissions fall by 8%.  
- POP LF falls by 9%.  This means most of our PRG thermalization losses in build-up come from increased arm losses.
- REFL LF lost 41%.  The REFL port is more complex because the CARM loop is trying to keep it as dark as possible in carrier, and PRCL is trying to keep 9 MHz on resonance as well.  REFL LF is probably mostly seeing the collapse of the 9 MHz sidebands, along with some 8% losses from the arms, and 15% losses from the 3 dB 45 MHz mod depth reduction.
- POP 18 lost 26%.
- POP 45 lost 14%.  7% comes from the modulation depth reduction.

Skipped initial alignment. Worked ok, not sure if it saved time, I had to do quite a bit of by-hand alignment during lock acquisition.

Sheila previously suggested the ALS Y arm glitches show up when it is raining. I noticed that this afternoon ALS Y arm glitches were frequent, and it was raining. Now the y-arm glitches are gone and it is not raining. Coincidence?

Had 45 mins of NLN with no glitches, 1232065259 - 1232057789. Later in the same lock a ~50Hz scatter-looking lump showed up. First attached figure is DARM BLRMS as the lump first showed up, with excess noise in the red BLRMS progressively gets worse. The second attached figure is a spectrogram for a period later in the lock, also showing line at 46.1 Hz line that drifts in and out of the spectrum.

Moved ads dither lines (PIT3, YAW3, YAW5), implemented in guardian. The gains and phases of these loops will likely need further fine tuning in the future, so far we've successfully acquired and powered up with the new lines. In a later lock I had problems engaging the other dither loops due to a low frequency ring up. I reduced the gain of the dither loops during acquisition, and re-increased the gain in the INCREASE_POWER state so we don't spend too much time waiting for things to converge.

Had several locklosses from ENGAGE_ASC_FOR_FULL_IFO, where SRC1, SRC2, PRC1, PRC2 yaw ran away after increasing the PRC1, PRC2 and SRC1 gains. Made the convergence checker for PRC2 and SRC2 stricter and increased wait times, PRC2 is especially slow to converge.

Craig plugged in the SR785 by the PSL racks to take CARM transfer functions, this made the 120 Hz line worse, and in the next two locks we had broadband noise DARM that was coherent with excess noise also visible in SRCL and PRCL. In one lock the noise was at ~750 Hz, and the next from ~80-400 Hz. We went out to unplug the SR785 through the main door to the LVEA and lost lock when tip toeing past HAM2. In the next lock, with the SR785 unplugged, the noise was gone. We went back into the LVEA to plug the SR785 back in, this time through the back door, and again lost lock while tiptoeing past IOT2L.

Some were taken during NLN, others are full lock, but not yet NLN.  January 3rd, 13th, 15th, and 17th.

Performed the tests Fabrice suggested below, basically cleaning, inspecting measuring fibers and sleeves at the HAM6 green fiber feed-through, with no improvement.

1 - Couple a new low insertion loss fiber (let's call it 'Air-3') after the the current one ('Air-2'), and check the transmission.
2 - Remove the mating sleeve at the feedthru,  gently clean the tip of the of the feedthru ferule. Spray air in the mating sleeve and put it back on.
3 - Connect 'Air-3' to the feethru, and check the transmission.
4-  Remove 'Air-3' from the chain

Initial power 20.4mW at input coupler -> 2.6mW out on green Refl.
Take out 'Air-2' inspect, -> transmission=12.5 mW. 'Air-2' looked scratched replaced with 'Air-2_5'  transmission 12.6 mW (end A/B insertion loss .41/ .45 dB)

'Air-3' (end A/B insertion loss = .10/.15 dB). Power trans 'Air-3' = 10.5 mW. Remove mating sleeve at feed-through, clean feed-through ferrule, spray mating sleeve replace. Insert 'Air-3' transmission on green Refl = 0.5 mW

Remove 'Air-3', clean 'Air-2_5' again insert -> 1.5 mW onto green Refl.

Put original 'Air-2' back -> 2.6 mW on green refl.

This is now the max. I could not recover any more power regardless of how "loose" or "tight" the connection was. It's possible the ferrule may have been scratched or got contaminated when putting the sleeve back on (it's rather awkward). The ferrule can be cleaned but not inspected because of the location. After the 1.5mW transmission measurement with Air-2_5'  I was just glad to recover 2.6 mW transmission with 'Air-2' to green Refl, which after Nutsinee's latest post, still allows us to move on for a bit.

Note 'Air-2' had very low insertion loss at the HAM6 end (.15 dB) and this is probably why the transmission to the vacuum fiber was better than 'Air-2_5' even though 'Air-2' was scratched.

On Jan. 15th I replaced IO_MB_M3, which is in the main beam, in the IO path on the PSL.

After Robert added the tuned damper, I replaced the optic, and realigned the IMC_IN beam, using my GigE cameras, and the single bounce off of MC1 to the IMC WFS, and the results are that the IMC optics changed 5urad or less in each DOF from before the optic swap, results described in detail in alog 46436.

Robert found the jitter peak from this optic was greatly reduced, described in alog46494.

I've cleaned and inspected the optic that was removed, and found 6 small dings in an arc on the front surface, ranging from about 3mm to 16mm from the center of the 2 inch optic.

These damage sites are consistent with what was likely considered minor impact, and with the scattering from the optic.

I can't yet confirm that the scattering has been reduced with the new optic, as there was no time to evaluate.  I will evaluate at the next possible opportunity.

The attached files show two views of the optic's damage, and those features that are on the front surface are circled in yellow.  Both files are ~8MB.

Yesterday I double checked that our threshold power still makes sense compared to alog45528. This could be an indication that our crystal is doing fine. The measurement was taken with 20dB attenuator on 80MHz RF modulation.

I measured SQZ/ASQZ using Mephisto LO to get a direct comparison to alog45586. Along the way I found that I have been saturating the homodyne (I normally set the RF analyzer to 1MHz - 5MHz so I missed all the peaks at 6, 9, 12, ... MHz). I used LO power of 0.5mW and CLF of 5uW. The total power was less than what we used back in August (alog43738) which we used 1 mW LO power and 10uW seed without saturating the homodyne.

Attached a sqz/asqz result. The magenta/round datapoint were taken with 20dB attenuator on the 80MHz modulation signal, HD not saturated. The black Xes were taken without 20 dB attenuator on the 80MHz modulation signal to ensure that extra 80MHz sidebands do no harm (except that we get less non linear gain). The red round data points were taken when the HD was saturated. The CLF needed to be turned down by a factor of 10 (0.5uW total to the HD) for the HD to not saturate (where we saw no more 3MHz harmonics).

Looking at ASQZ alone it almost seems like we have 95% efficiency but you would need A LOT of phase noise to explain the squeezing. ASQZ also doesn't seem to be affected by extra phase noise introduced by the saturated homodyne, only sqz does (which makes sense, so I'm not sure if it totally explains what we measured with PSLLO here). But the fact that we can see 4dB here is a good enough indication that we can move on (and that we probably don't need to move the crystal). The solid green plot I tried giving it an arbitrary phase noise (that LLO found) just to see how that fits. If it's real I wouldn't know where that could have come from (it doesn't show on LO IMON). 

Jonathan, TJ, Jamie, Dave:

Following h1guardian1's upgrade to a 2-cpu, 40 hyper-threaded core machine on Wednesday, I started trending its ethernet port statisics Friday afternoon. Every 10 minutes ifconfig reports the port input/output errors. Jamie and TJ will see if any EPICS-CA errors correlate to ethernet port issues.

When the script was started the only error was an accumulated 1048 receive overruns (i.e. RX FIFO errors). We suspected that these may have been acquired when h1pemey/DAQ were restarted Thursday.

No further errors have been seen since program start (see attachment)

Koji, Craig

Koji has the idea that maybe our IMC FSR is not exactly equal to our RF 9 modulation frequency, and this is responsible for additional 9 MHz RF AM seen by Sheila and Jenne.
So today Koji and I moved the modulation frequency around.  We used the double modulation technique for finding the FSR by looking at the 24 + 9 = 33.178 MHz peak and trying to minimize it.  
We found that the frequency was already pretty well tuned.  We moved it around from 9.100230 MHz and found the best rejection of 33.178 MHz peak at 9.100225 MHz.  Larger movement of the modulation frequency (on the order of 10s of Hz) resulted in much higher 33.178 MHz peaks.

Then we tried repeating Sheila and Jenne's OMC 9 MHz locking test.  The results are plotted below.
The first attachment shows the OMC DCPD spectra and RF AM monitor spectra, each one appropriately calibrated into RIN, for different RF9 modulation slider values.  We see a moderate (factor of 2) win from going from +27 dB to 23.4 dB, otherwise 9 MHz RIN remains about constant.
The second attachment shows the effect of changing the 9 MHz modulation frequency by 5 Hz.  The 9 MHz RIN seems to be independent of our frequency move. 

We note that the move from 23.4 dB to 27 dB on the RF9 slider actually does increase the modulation depth, since the OMC DCPD SUM increases when this is done.  Koji suspects that his measurements from before measuring the response of the EOM suffered from broadband saturations spoiling the linear response.  Going above 23.4 dB on the 9 slider is still not recommended.
We moved the modulation frequency back to 9.100230 MHz and left it there.

-------------------------------------------------------------------------

We also injected a line into the EOM driver for the 9 to calibrate the REFL A 9I pd into units of RIN.  It seems that REFL A 9I is a pretty poor RIN sensor relative to the OMC and RF AM monitor. (attachment 3)

Koji, Craig

We became interested in Sheila and Jenne's 9 MHz RIN measurement, and so set out to measure a more complete spectrum of the 9 MHz demodulated REFL PD noise to gain some perspective on the 9 MHz amplitude modulation transmitted through the IMC.

First we shuttered the PSL and measured a full suite of dark noise spectra.  
Then we requested ITMs and SRM misaligned, PRM aligned, and 8 watts of input laser power.  This gave us 30.97 mW on REFL A LF and 28.52 mW on REFL B LF.

This is not a direct comparison of 9 MHz RIN as measured by Jenne and Sheila; their measurement is capable of picking out a single sideband, while ours is the usual carrier and sidebands beatnote.  We'll think about the comparison of each measurements' RIN.

I put a camera on the HAM2 West door, which looks at CW1, the back side of the wedge, which is carefully aligned to shoot through the IM1 tower.

This week I developed a procedure to use PRM with a single bounce, to move the REFL beam on CW1, and and look for clipping.  The procedure revealed that when PRM is in it's aligned position plus an additional 200urad in yaw, there's an increase in light on the baffle, consistent with the REFL beam starting to clip.  With PRM at -1000urad, the REFL beam is clearly seen on CW1 to thr right of the main input beam.

Using the image and a beam path simulation I wrote in matlab, I've measured the main input beam to be between 2mm and 3mm from center.

I need to adjust my HAM2 top camera to improve the view of the IO Faraday output side, which is planned for next Tuesday, in order to realign through the Faraday, to correct the mis-centering at both the input and output Calcite Wedges.

compared to the main input beam, which I've used to identify the location of the main input beam on the wedge.  A diagram and images attached.

On Monday evening, Koji and I went out and measured the shot and dark noise of the IMC/CARM PDs.  Then, Koji taught me how to think about photodiode noise.  Slides 35 and 36 of G1401145

Model:

where nV is the voltage noise floor, T is the transimpedance, e is the electron charge, R is the responsivity of e λ/(h c) = 0.858 A/W, QE is the quantum efficiency of 0.9, PDC is the DC power from the incident light, and Pdet is the power apparent from the intrinsic detector noise (i.e. the incident power where shot and dark noise are equivalent).

We fit the measured shot and dark noise spectra for IMC REFL I/Q, REFL_A 9I/Q, REFL_B 9I/Q, and at the CARM and IMC servo board OUT2 locations with the nominal gain slider settings (OUT2 connected to IN1, CMB IN1GAIN = 6 dB, IMC Board IN1Gain = 28 dB).  REFL_A spectra are plotted in attachment 3 as an example.
For dark noise I shuttered the PSL.  For shot noise I measured with 1W, 2W, and 4W input requested power, and recorded how much light was reported by the LF channels:

Light Level      REFL_A_LF [mW]     REFL_B_LF [mW]    IMC_REFL_LF [mW]
----------------------------------------------------------------------
Dark                          0                  0                   0
1W PSL                     3.51               3.24                1.87
2W PSL                     7.38               6.80                3.92
4W PSL                    15.08              13.90                8.01

Full Lock                  4.36               4.02                1.15


Then, I found the average noise floors for each spectrum between 3 and 10 kHz, and fit the model to the data.  (attachment two)
I have let the transimpedance number from the fits capture the servo board gain sliders.

Photodiode      Transimpedance [V/A]     Pdet [mW]    IN1 Gain Slider [dB]
-------------------------------------------------------------------------------------
CMBoard                        730.4          2.2                       6
IMCBoard                      2357.7          3.5                      28
IMC_REFL_I                     128.6          1.6
IMC_REFL_Q                     127.5          1.7
REFL_A_9I                      154.9          2.0
REFL_A_9Q                      151.8          2.3
REFL_B_9I                      152.4          2.1
REFL_B_9Q                      152.2          2.1

A new h1calinj model has been installed on the h1oaf1 machine.  This model will hold all front-end injection handling logic, including the injection EXC test points, under the channel prefix "H1:CAL-INJ_".  The existing injection handling code in the h1calex model ("H1:CAL-PINJX_") has been left as is for now to facilitate transitioning and testing.

The first two attached images are of the top-level contents of the new h1calinj model, and of the contents of the "INJ" CAL_INJ_MASTER2 library part that contains all the core logic.  The top of the latter shows the main injection signal flow.  The two "CW" and "TRANSIENT" filter banks at the upper left hold the EXC inputs and calibration filters for the continuous-wave (CW) and transient (TRANSIENT) injection inputs respectively.  The outputs of the two modules are summed and the overall "MASTER" output goes through an output switch ("MASTER_SW"), and then finally a switch ("END_SW") that determines which of calex or caley receives the injection signal ("H1:CAL-INJ_{X,Y}" via cdsIPCxRFM).  Below the main signal path is logic to determine the presence of signals at various points in the injection path, and bundle that info into a single status word ("STATUS_OUT").

IPC receivers for the INJ_MASTER output sent from h1calinj (H1:CAL-INJ_{X,Y}) were added to the h1calex and h1caley models.  EPICS and acquired test point monitors of the received signals were also added.

NOTE: a single 2**14 Hz, 61 us cycle delay will be added to the injection path because of the IPC needed to carry the signal from the vertex to the ends.

The new acquired fast channels are:

• H1:CAL-INJ_CW_OUT_DQ
• H1:CAL-INJ_TRANSIENT_OUT_DQ
• H1:CAL-INJ_MASTER_OUT_DQ
• H1:CAL-INJ_STATUS_OUT_DQ
• H1:CAL-INJ_X_OUT_DQ
• H1:CAL-INJ_Y_OUT_DQ

The STATUS channels (the H1:CAL-INJ_STATUS_OUT_DQ uint32 fast channel and the H1:CAL_INJ_STATUS slow channel) have the following bits:

0. MASTER
1. CW
2. TRANSIENT
3. CBC
4. BURST
5. DETCHAR
6. STOCH

A value of zero (0) indicates no signal of the specified type is present, and a value of one (1) indicates the presence of the specified signal.

The h1calinj model also holds the "TINJ" EPICS status bits, nominally set by the INJ_TRANS guardian node.  NOTE: the "H1:CAL-INJ_TINJ_" EPICS records were previously hosted by the ext_alert_ioc.py soft IOC running on the h1fescript0 machine.  The records were removed from the soft IOC and the process was restarted.

The final image attached is a new MEDM screen CAL_INJ_CONTROL2.   All functionality and status bits in the h1calinj model, and the monitors in h1calex and h1caley, are exposed.

As mentioned above, all the existing INJ infrastructure remains in place.  We leave it up to the INJ group to update the INJ_TRANS guardian, the psinject process, downstream monitors, etc.  We would like to schedule the removal of the old INJ infrastructure as soon as possible,