Alexa, Evan, Kiwamu

We observed some new features which are related to the SRC mode hopping.

• We noticed that when SRM was slightly misaligned from the optimum angle in pithc by approximately 15 urad, it seems to reduce the number of mode hops in SRC.
  • Of course, this reduced the power build up in the signal recycling cavity by some amount, but it was more stable due less mode hoppings.
• We starting putting an offset in the SRC operating point in order to study if the signal really has multiple zero crossings.
  • When LSC-SRCL had an offset of -800 cnts it became much more stable than it was in the 10 W configuration. We could even stably go to -5000 cnts by expanding the linear range by the normalization with AS_RF90.
  • On the other hand, when we went to the positive side in the offset (e.g. ~ 400 conts), we could easily lose the lock due to the instability or the hopping.
  • This indicates that the error signal for SRCL is asymmetrical about 00-mode's zero crossing.
• We checked the same asymmetrical error signal both in 10 W and 1 W configurations and found that the instability seemed to be further away from the 00-mode in the case of 1 W.
• We also checked the level of RFAM in all the REFL detectors by directly detecting the prompt reflection of PRM while the interferometer was misaligned.
  • Even though both in-vac and in-air RF45_I signals showed a discrete jump in the RFAM offset, the amount of the offsets were much smaller -- the biggest offset was about -22 cnts in REFL_A_RF45_I.
  • Apart from REFL_RF45s, all the detectors seemed to have a linar relation in the RFAM offsets with respect to the PSL power.
• Note that we had to re-align the DRMI after three lock or so losses. Otherwise, we were not able to smoothly acquire the lock.

Kiwamu, Alexa

Today we looked at the demod phases for the REFLAIR RFPDs. The result was inconclusive. We excited PRM_M3_LOCK_L at 3.25Hz with an amplitude of 10000cts. We then examined the magnitude of transfer function of the I/Q error signals to this excitation at 1W, 5W, 7W, and 10W incident laser power. The data was taken with RELFAIR_A_RF45 at -135deg, RELFAIR_A_RF9 at 93deg, REFLAIR_B_RF27 at 107.8deg, and RELFAIR_B_135 at -30 deg following our nominal configuration. Attached show the magnitude vs power for the 1f and 3f signals, along with a linear fit. RELFAIR_A_RF9 behaves as expected with a linear reponse. During the measurement, REFLAIR_A_RF45 was not very coherent with the excitation, which explains why 45Q's response does not look very good. REFLAIR_B_RF27 looks fine as well; but RELFAIR_B_RF135 does not look right; it appears I might have missed a sign flip in the TF. But no major red flags...

J. Kissel, K. Venkateswara

The output filter for the BRS included two zeroes at 8.8 mHz and two poles at 1 mHz to compensate for the real pole of the beam-balance and a zero due to the gravitational spring (proportional to d, distance between CoM and pivot). Based on the data during windy periods (see 13563 and 14422), it looks like d is roughly -35 +/- 5 microns, which corresponds to an imaginary zero at ~7.3 mHz. Since Foton doesn't allow imaginary poles, I put in a complex pole with Q of 3 as an approximation but this means that the output is going to be incorrect between 5-10 mHz.

With this new output filter in place, the tilt-subtraction is working well above ~ 20 mHz. The attached pdf shows the ASD for the ground seismometer (blue), the tilt-subtracted super-sensor (red) and the tilt-correction output (green). Wind-speeds were in the range of 20-30 mph during this measurement. Note that while the super-sensor is lowered by a factor of ~5 at 50 mHz, the subtraction near 10 mHz is limited by the approximation I made above. If we could add the imaginary pole in Foton, the subtraction would be better.

I investigated why the H1 frame is about 20% larger than the L1 frame. Using the set of INI files in the running configuration for both DAQs, I ran them through Jim's inicheck program. I removed the systems which are H1 only (PEM at midstations and SUS-QUADTST in LVEA). To remove the frame size compression ration, I am comparing raw data rates not size of frames on disk.

So H1 has 0.4% more slow and 0.25% LESS fast channels but 4% MORE data in the science frame.

I took the H1 and L1 science frame channels lists, removed all 16Hz channels, converted L1 to H1, sorted them alphabetically and compared for fast channels missing from L1 frames or having higher datarates in the H1 frame. The PEM system stood out with the biggest differences.

This summer Robert asked for many 2kHz channels to be increased to 8kHz and some added to the science frame. Looking at the PEM channels for EX, EY and CS, H1 has 1.8MB/s more data than L1. In the above table, 28.9 + 1.8 = 30.7, close to the H1 number of 30.2 (difference is most probably due to L1-only channels like CS-CAL).

As an aside, I was surprised that the science uncompressed rate was so high compared with the frame size. For H1 this ratio is 30.2 vs 12.5 which is a compression factor of 2.4. Greg confirmed that he is seeing compression rates around 2.7 using frcheck.

I also found that the compression rate and therefore the frame size depends on the state of the interferometer. As one would suspect, if ADC data has higher AC components, the compression factor will decrease. For example, I trended the frame size against the Guardian state of the DRMI locking system, there is a clear correlation (see attached plot). The size of the frame has varied by 6%.

Here is the list of commissioning task for the next 7-14 days:

Locking team:

1. Measure recycling gain and try explaining low value.
2. Measure power dependence of demod phase in DRMI.
3. Check demod phase after the IMC as function of power.
4. Change 1f/5f relative phase and look at DRMI stability.
5. Common hand-off to transmitted power.
6. Lock interferometer.

Alignment team:

1. Revisit EY green WFS beam path and look for ghost beams.
2. Make EY green WFS work again.
3. Install EX green WFS auto-centering hardware.
4. Commission an arm alignment controls topology which can be used for initial alignment.
5. Integrate green WFS into initial alignment.
6. Commission more WFS dofs in DRMI. 
7. Improve stability of DRMI alignment.

SEI/SUS team:

1. Install new models for violin mode damping.
2. Commission violin mode damping.
3. Install new ESD linearization code.
4. Commission ESD linearization.
5. Pump HAM1 to reduced periscope vibrations(?)

RF:

1. Make an assessment of RF cross talk into the IMC AOM.
2. Investigate EMI radiation by the VCOs and the fixed frequency OCXO.

TCS:

1. Finish commissioning  ring heaters and CO2 lasers.
2. Optimize mode matching and Michelson contrast defect.

Daniel, Thomas and shivaraj

This morning we installed Beckhoff software for the PCal at the X-end station. During the installation we restarted the Beckhoff system a couple of times. Since the PCal setup is not connected yet, we didn't test the singals. One thing that came up during the software installtion was the naming of PCal channels. Previoulsy it was decided that the PCal channels would come under CAL subsytem and would read such as H1:CAL-PCALX_SHUTTERSTATUS. However 'CAL' is a restricted word and hence we couldn't use it in the channel name as we wanted. For time being we have used 'PCL' instead of 'CAL' and the channels now read such as  H1:PCL-PCALX_SHUTTERSTATUS.

J. Kissel, K. Venkateswara

After the ground super-sensor sensor correction filter model changes to the H1ISIETMX models (see LHO aLOGs 14408 and 14452), these changes needed reflecting on the BRS overview screen.

In doing so, I identified a few bugs in the implementation that we'll fix tomorrow when we add sending the GND super sensor to HEPI to try out sensor correction there. The bugs to fix:
- The gravitational gradient damping mechanism's control output signal is monitored in the front end. That control signal's channel name has a spurious copy-and-paste "1" at the end of it, i.e.
H1:ISI-GND_BRS_ETMX_DAMPCTRLMON1
We'll correct it to be
H1:ISI-GND_BRS_ETMX_DAMPCTRLMON
- The newer filters in the GND_SENCOR block needs to have the chamber name in its filter banks, like the filters in the GND_BRS block. In addition, if we do get more of these BRSs, it'll be essential to keep clear which sensor is getting corrected and in which direction. So, the channels should change from 
H1:ISI-GND_SENSCOR_ROTVELCORR
H1:ISI-GND_SENSCOR_TORQUECORR
to
H1:ISI-GND_ETMX_SENSCOR_ROTVEL_STS_X
H1:ISI-GND_ETMX_SENSCOR_TORQUE_STS_X
(where the "CORR" after "ROTVEL" and "TORQUE" are redundant with the "COR" in "SENSCOR," so we'll remove that as well).

Rebooted the projector0 computer, it appears that dmtviewer has a memory leak. Dmtviewer had been running for a bit over a month.  After reboot, restarted the dmtviewer with the seismic FOMs.

On the BSC HEPI L4C input bank there are two filters, one calibrates the raw signal to nm (Cal) and the second contains the sensor symmetrization feature around at 1hz (Sym.)

On the BSC-ISI HEPI L4C, there is only a Cal filter but this also contains the resonance notch.  Attached is the foton comparison, they are close but not exactly the same.  I'm waiting for confirmation from MIT but I think these should be the same.  Along the same lines I'm correcting the BSC-ISI HEPI L4Cto Cart matrix which I think too should be the same as the BSC-HEPI L4C to Cart matrix.

The BSC-ISI HEPI L4C filters should be the same as the BSC-HEPI L4Cs.

Beckhoff work at EX, othe Pcal work, tomorrow taking hardware to EY
HAM4&5 oplev work tomorrow
Oplev calibration on-going
ESD linearization algorithm
A number of SEI fixes, mostly software. Hugh will be working on ITMX HEPI tomorrow during maintenance, will probably require turning HEPI off.
 

Took HAM6 off line to correct the matrix that was wrong.  Made safe.snap & uploaded it to the SVN.  HAM6 SEI back under guardian control.

model restarts logged for Sat 18/Oct/2014
2014_10_18 05:45 h1fw0
2014_10_18 13:27 h1fw1

both unexpected restarts.

model restarts logged for Fri 17/Oct/2014
2014_10_17 07:32 h1fw1

unexpected restart of fw1, looks like additional memory did not fix this. Time will tell if frequency of restarts has been improved.

Alexa, Sheila, Dan, Kiwamu

I was about to post the alog below, when we decided to go back to 1 Watt since everything else we have tried to get DRMI stable at 10 Watts has failed.  At 1 Watt we had no mode hopping issues, we aquire lock within 10 minutes or less, and it is stable once we lock.   This sounds like a familiar story.....

Right now we are trying to transition to 3F, we didn't suceed at 1 Watt so we are now trying at 2.9 Watts. 

The old, sad alog:

Today we tried to get PRMI to be stable again. Our working theory is that we are having these mode hopping problems beause of some bad alignment.  Our intial alingment scheme right now doesn't fix the alingment of PR3, so we have tried various methods to recover this alingment to a time when we had more stable locking.

• We have added states to the ISC_DOF guardian (what we are now calling the guardian we have been using for inital alingment) that align PR3 using a servo on the GigE camera on ISCT1.  It seems to work, we have the current position as the target right now, but this obviously isn't optimal. Once we find the good alingment we should both center the POP QPDs and mark the location of the ALS X beam
• The most sucsessfull method we tried was based on trending old offsets, locking the IR to the X arm with refl wfs actuating on IM4 +PR2 to keep the input beam aligned down the X arm while we moved PR3 a few microradians.  We then locked PRX and used the refl wfs to actuate on PRM.  We found that our PRX +wfs don't yeild a repeatable alingment of PRM.  We tried to walk ourselves towards the old alignment using this method, but found that our best PRMI build up was in the other direction.  The result We were able to get the build up in PRMI better (we went from about 150 counts in POP 18 at max to almost 200 (perviously our best PRMI build up was 220), however we are still mode hopping.  Until the last two days, we hadn't seem this mode hopping in PRMI, just DRMI.  A video is attached, at the end of the video we drop the lock.
• We tried some other methods, like simply restoring to our old offsets, and we tried locking the input beam alignment to the X arm while we move PR3 to get the beam spots on the POP sled back to the position they were at when we had stable DRMI late last week.  (both of these methods were complete failures, resulting in very low or no flashes in PRMI)

We also got a DRMI lock, which lasted only a few minutes, and dropped at 1:53:07 october 18 UTC, that is the second video attached.   Our big victory today was that Alexa killed the fly that has been hanging out in the control room.

Krishna had suggested that we could use an intermediate blend filter for the BSC (more aggressive than the 750mhz blends, less aggressive than the LLO blends), to take advantage of the sensor correction. At the SEI call today, it was suggested that we try the 01_28 blend being used on the HAMs, which is a ~250mhz blend. After a little hacking, I was able to get something that could be installed. All I had to do was make extra copies of the filters in a mat file and change some variable names, but it took a little while to figure that out. The blend routine was then able to compute the complementary filters for the T240 and L4C's. The X and Y blends for St2 look a little funny, they aren't quite complementary above the blend, but we don't need these filters for St2. See attached plots for complementary forms, St1 on the first page, St2 on the second page.

I won't install these new filters until next week, so I won't risk disrupting commissioners.

Replaced non functioning whitening chassis S1101631 with S1101602 on 10/15

J. Warner, K. Venkateswara

Using Jeff K.'s model modifications, we were able to incorporate BRS tilt-subtraction to the GND_STS which was then output to STS_C channel (which was empty previously). This was then used for sensor correction by changing the cartesian conversion matrix. The tilt subtraction filter uses a simple acceleration to velocity converter, a gain matching and a high-pass filter.

Rich M. wrote for us a neat IIR filter to try out for the sensor correction filter bank. After the first attempt doubled the ground signal, we used this with a gain of -1.0 which worked well :)

We wanted to try out this sensor correction yesterday, but there was a lot of activity at ETMX which was disturbing the BRS as shown in the attached .png file. The good news is that immediately after the activity ceased, the BRS output settled down so the damper continues to work as expected.

This morning we had some time to try it out before EX activity resumed. The file NoSensCorr45mHzblnd.pdf shows the current ISI stage 1 performance with the 45 mHz blend filter. As I've mentioned before, notice the nice coherence with the GND_STS_X and ST1 T240_X between 0.1 to 0.6 Hz. The Y-axis is in counts (velocity).

For the uninitiated, sensor correction adds the ground sensor signal to the position sensor, thus it is effective only below the onboard inertial sensor blend frequency. In this case, since we wanted to do sensor correction in the 0.1-0.5 Hz band, we shifted the inertial sensor blend to 250 mHz. The two files NoSensCorr250mHzblnd.pdf and SensCorrOn250mHzblnd.pdf shows the Stage 1 motion without and with sensor correction respectively. Even with sensor correction on, this configuration seems to be a factor 2-5 worse in the 0.1 to 0.5 Hz band, but it is very interesting that the coherence with GND_STS_X is much smaller. And there is significant coherence with ST1_T240_Z at the microseismic peak. This might be related to what I mentioned before in 14426.

But to our surprise we were injecting noise below 0.1 Hz. I then rechecked our tilt-subtraction filter and it turns out we had one filter gain wrong. I've corrected it and I verified that the tilt-subtracted super-sensor was indeed smaller than the ordinary ground sensor. Unfortunately, EX activity resumed just then and we couldn't verify that the sensor correction was lowering the noise below 0.1 Hz. The ISIs were restored to the 45 mHz blends for commissioning activity.

Summary: Our first attempt at trying sensor correction with the ground 'super-sensor' was not terrible! :) We will do a more refined attempt tomorrow. We will also try doing sensor correction using HEPI which may give slightly different results.