J. Kissel

These, damped, TOP2TOP, transfer functions complete the acceptance measurement suite for the H1 SUS ITMs (see LHO aLOG 14729 for the rest of the suite). For this Phase 3B measurement, the chamber is at vacuum, the HPI and ISIs are damped and isolated, and the SUS itself is damped. 

The main chain transfer functions look exactly as expected with the LHO filter configuration that its been left in (see recent modeling efforts, and discussion of LLO filter design in LHO aLOG 14959). The reaction chain, in it's sad, sorry, copy-of-a-bad-copy, 2011, state looks like very very little thought has gone into their loop design. This is true, because it doesn't seem to matter for IFO noise or control. Some day we'll get to re-designing these loops just so they don't look so aesthetically displeasing.

Raw .xml files live here:
/ligo/svncommon/SusSVN/sus/trunk/QUAD/
H1/ITMX/SAGM0/Data/2014-12-11_0429_H1SUSITMX_M0_L_WhiteNoise.xml
H1/ITMX/SAGM0/Data/2014-12-11_0429_H1SUSITMX_M0_P_WhiteNoise.xml
H1/ITMX/SAGM0/Data/2014-12-11_0429_H1SUSITMX_M0_R_WhiteNoise.xml
H1/ITMX/SAGM0/Data/2014-12-11_0429_H1SUSITMX_M0_T_WhiteNoise.xml
H1/ITMX/SAGM0/Data/2014-12-11_0429_H1SUSITMX_M0_V_WhiteNoise.xml
H1/ITMX/SAGM0/Data/2014-12-11_0429_H1SUSITMX_M0_Y_WhiteNoise.xml
H1/ITMX/SAGR0/Data/2014-12-11_0545_H1SUSITMX_R0_WhiteNoise_L_0p1to50Hz.xml
H1/ITMX/SAGR0/Data/2014-12-11_0545_H1SUSITMX_R0_WhiteNoise_P_0p1to50Hz.xml
H1/ITMX/SAGR0/Data/2014-12-11_0545_H1SUSITMX_R0_WhiteNoise_R_0p1to50Hz.xml
H1/ITMX/SAGR0/Data/2014-12-11_0545_H1SUSITMX_R0_WhiteNoise_T_0p1to50Hz.xml
H1/ITMX/SAGR0/Data/2014-12-11_0545_H1SUSITMX_R0_WhiteNoise_V_0p1to50Hz.xml
H1/ITMX/SAGR0/Data/2014-12-11_0545_H1SUSITMX_R0_WhiteNoise_Y_0p1to50Hz.xml
H1/ITMY/SAGM0/Data/2014-12-11_0510_H1SUSITMY_M0_Mono_WhiteNoise_L_0p01to50Hz.xml
H1/ITMY/SAGM0/Data/2014-12-11_0510_H1SUSITMY_M0_Mono_WhiteNoise_P_0p01to50Hz.xml
H1/ITMY/SAGM0/Data/2014-12-11_0510_H1SUSITMY_M0_Mono_WhiteNoise_R_0p01to50Hz.xml
H1/ITMY/SAGM0/Data/2014-12-11_0510_H1SUSITMY_M0_Mono_WhiteNoise_T_0p01to50Hz.xml
H1/ITMY/SAGM0/Data/2014-12-11_0510_H1SUSITMY_M0_Mono_WhiteNoise_V_0p01to50Hz.xml
H1/ITMY/SAGM0/Data/2014-12-11_0510_H1SUSITMY_M0_Mono_WhiteNoise_Y_0p01to50Hz.xml
H1/ITMY/SAGR0/Data/2014-12-11_0629_H1SUSITMY_R0_L_WhiteNoise.xml
H1/ITMY/SAGR0/Data/2014-12-11_0629_H1SUSITMY_R0_P_WhiteNoise.xml
H1/ITMY/SAGR0/Data/2014-12-11_0629_H1SUSITMY_R0_R_WhiteNoise.xml
H1/ITMY/SAGR0/Data/2014-12-11_0629_H1SUSITMY_R0_T_WhiteNoise.xml
H1/ITMY/SAGR0/Data/2014-12-11_0629_H1SUSITMY_R0_V_WhiteNoise.xml
H1/ITMY/SAGR0/Data/2014-12-11_0629_H1SUSITMY_R0_Y_WhiteNoise.xml
exported to text files of similar name. 

Procesed measurements and saved .mat files live here:
/ligo/svncommon/SusSVN/sus/trunk/QUAD/
H1/ITMX/SAGM0/Results/2014-12-11_0429_H1SUSITMX_M0_DTTTF.mat
H1/ITMX/SAGR0/Results/2014-12-11_0545_H1SUSITMX_R0_DTTTF.mat
H1/ITMY/SAGM0/Results/2014-12-11_0510_H1SUSITMY_M0_DTTTF.mat
H1/ITMY/SAGR0/Results/2014-12-11_0629_H1SUSITMY_R0_DTTTF.mat

All have been committed to the SusSVN repo.

Frequency of DAC lock out events (requiring IOP model restart to fix). One more event (h1susb123) added since my last  entry

12/11 2014 h1susb123

10/26 2014 h1seih45

9/11 2014 h1seih45

9/10 2014 h1seih23

8/11 2014 h1susb123

8/9 2014 h1seih23

4/21 2014 h1sush2a

3/18 2014 h1sush2a

2/27 2014 h1seih23

12/16 2013 h1seih23

11/7 2013 h1sush2a

11/7 2013 h2sush34

8/8 2013 h1seih23

This is regarding the BS suspension damping using its oplev.   

I looked at the behaviour of HAM3 and BS oplev lasers over the last two hours.

The HAM3 laser (HAM3 oplev is the preferred test bed for the lasers we fix before we plug them into core optics suspensions)  has an order of magnitude lower noise in the 10 to 100Hz band.   The forest of peaks we see above 100 Hz are probably from the pier motion and not from the laser itself.

I have also looked at the 24 hr trend of the HAM3 oplev.  There are a few glitches during the time people were working in the LVEA.  I expect these will not be present when things are quiet.

Shifting the HAM3 oplev laser to BS oplev should improve the BS suspension damping.

Jeff K, Suresh, Sheila

The wind is picking up.  We are having 60 mph gusts and can hear it on the roof, so that's  for interferometery tonight.  High winds started at around 2:50 UTC, December 12th.  

S. Dwyer, K. Venkateswara

Sheila turned WFS on and locked the X-arm using the green laser. I then took measurements with and without the LLO sensor correction on ITMX and ETMX. I also tried out the low-frequency sensor correction (using Rich's SC filter and the tilt-subtracted sensor ), only at ETMX. ASD plots are attached showing various sensors. And a comparison of the control signal for the different configuraitons.

It is likely that ALS is not sensitive enough below 1 Hz to see the effect of the sensor correction.

I began realigning the PRC probe beam (sent into HAM2 via IOT2R) to the PRM. First of all I aligned the beam reflected from PRM and transmitted through IM4 back into the aux laser path, using an iris, centering on lenses, and the FI aperture as a guide. This required some significant repositioning of the lower IOT2R periscope mirror, because this was adjusted in order to dump the IM4 trans beam recently for the benefit of the ISS photodiode. I then turned on the aux laser and overlapped the main beam (as best as possible just by looking at the spots on a card at two locations). With the main beam now coarsely overlapped with the aux beam, I aligned it to the RFPD that is used to generate the main laser / aux laser beat signal, maximizing the DC output of the PD. I observed the RF output on the HP spectrum analyzer, but wasn't able to see a beat signal (this might require more tuning of the aux laser temperature offset). One thing I did notice, however, were big peaks in the spectrum at the 9MHz and 45MHz sideband frequencies. These peaks remained when the aux laser was turned off. Since this beam is just the forward beam from the IMC, one would hope that the RF sidebands were pure phase modulation... is this an indication of unwanted RFAM in the IMC trans beam already? I don't know for sure what the IFO state was when I observed this, though I do know that the PRM was deliberately misaligned by this time.

Filiberto and Aaron pulling cables in LVEA

07:41 Karen and Cris cleaning in the LVEA
08:30 Jeff B. to LVEA to swap dust monitors
08:41 Jeff B. back
08:48 Jeff B. to end Y to look at dust monitors
09:14 Corey and Suresh to the squeezer bay
09:17 Betsy to the LVEA
09:23 Jeff B. back from end Y
09:32 Betsy to mid Y
09:34 Christina, Cris and Karen to end Y for first cleaning of BSC10
09:52 Gerardo back from end Y (turned on clean room over BSC10)
10:08 Pepsi truck through gate
10:09 Betsy back from mid Y
10:41 Andres to high bay
10:41 First cleaning of BSC10 done
10:58 Andres back from high bay
11:08 Daniel and Elli to end Y to work on WFS
11:52 Cyrus upgrading firmware on surveillance cameras
12:08 Daniel and Elli back from end Y
12:42 Betsy to mid Y
14:47 Gerardo to H2 PSL enclosure

David O., Elli and Greg working on HWS table near HAM6

The last couple of days Krishna and I have been working on sensor correction everywhere, but have been having troubles on HAM3. This is not exactly related to that effort, but we noticed that there is a very Q-y feature in the HAM3 ISI and SUS spectra. Jeff helped me dig into this and we are pretty sure this is not due to the suspensions, but it's been there the last couple of days, with and without sensor correction. It doesn't show in HEPI or the ground spectra, so I'm not sure where this comes from. When the IMC isn't being used I will try turning isolation loops off to see if the ISI is causing it. Attached spectra is of the MC2 suspension point(ISI GS-13's), but PR2, MC2 OSEMS and ISI CPS's see this, as well.

A new type of filter block is now available in the real-time system. The Integrator Filter Module (IFM) implements a single integrator only, but has a couple of additional features compared to a standard filter module:

• The digital filter uses direct form type I representation with the old output as the filter history value.
• The output limiter is enforced both on the output itself as well as on the history value. This prevents the run-away integrator.
• The integrator has a bleed-off feature. When the bleed is engaged, the filter output ramps towards zero at a specified rate.
• An optional bias value can be added to the filter output. This makes it easy to support a PZT offset, for example.
• Most other features of the standard filter module, such as hold, test outputs, etc. are available as well.
• Turning the filter off will reset its history to the input value. No ramp feature at the moment.

The attached screen shot represents the new IFM. The available parameters are

• Input offset with separate enable switch.
• Input gain.
• On/off switch.
• Filter on/off switch.
• Unity gain frequency (must be non-negative).
• Bleed rate (must be non-negative) with separate enable switch.
• Limiter value (must be non-negative) with separate enable switch.
• Output on/off switch.
• Output bias with separate enable switch.
• Output hold switch.
• Decimation switch for OUT16 channel.

The testpoint and EPICS channel names are the same as in the standard filter module:

• IN1/INMON: before the input on/off switch.
• EXC: added after the input on/off switch.
• IN2: after the excitation is added.
• OUT/OUTMON: after the limiter, but before the output on/off switch and bias.
• OUTPUT: at the output of the module.
• OUT16: same as OUTPUT but with an optional 8Hz decimation filter.

The SFM uses a bit encoded momentary switch for its enables. However, this requires state notation code support; and the IFM uses individual EPICS bi channels.

Hanford transportation will haul on day shift and swing shift on Friday 12/12, and ERDF will operate a Friday day shift.  No weekend work is scheduled.

Kiwamu, Dave O, Elli

Here are better quality images of ETMx.  These images were taken with the same method as for the ETMy images in alog 15547.  The exposure level was 500000 microseconds (compared to 10000 for ETMy.)

The camera isn't well focused, because we choose to increase the camera apperature to maximum, which limited our ability to focus.

I was trying to understand a strange temperature bump at MIDX VEA and Richard noticed that there had been a very abrupt  excursion outside.

This event turned out to be responsible for the VEA bump I observed at MIDX.  The outside temperature rose from 42F to 54F in about 45 minutes

Take a look at these two outside temperatures from X and Y recorded yesterday around 8 pm.

The end stations were not affected since we still have chillers operating there.

JimW did this Tilt Decoupling 14 April; those values he came up with have managed to survive and were still in the CPS Align Matrix.  Since we've done lots of changes in many places, we thought a check of tilt decoupling was in order.  This morning I repeated the measurement and adjusted the numbers around these historic values.  I changed both RX & RY decouple numbers at the same time during the tests.

The first attached is the results for driving in the Y direction.  The left column is for the in line coupling; that is, pitching response adjusted by the RX-Y component of the ALIGN matriix.  The right column shows the cross response--rolling about the Y axis adjusted by the RY-Y Matrix element.

The blue traces are with the original values, -.001/.0005 for RX/RY to Y elements.  Green is the change to -.0015/.00075; brown are with the elements at -.0005/.000 for RX/RY.
 The red traces are with the final desired values.  Looking at the left inline values, the original blue trace with RX-Y at -.001 looks best.  For the cross term RY-Y, it looks best at .00075 shown on the green trace--notice the greatly reduced coherence and lowered amplitude in the lower right and upper right plots respectively.  When these 'best' values are used for the red results, things don't look quite as good as the green and the blue but may be the best compromise.  This really shows the couplings here and the pain given the length of these measurements.

The second attachment shows driving in X.  The blue curves show the original values for RX-X & RY-X.  These are the best numbers as my bracketing around these makes the coupling higher seen in the green and red curves: The left column has the blue most linear and in the right column you see the lowest coherence and magnitude for the blue as well.

Templates for these measurements are in /ligo/svncommon/SeiSVN/seismic/BSC-ISI/H1/ETMY/Data/Transfer_Functions/Measurements/Tilt_Decouple/

As a preparation for doing some DRMI ASC, I've added a loop that servos SR2 to the AS_C QPD.  (see LLO alog 15840) 

As Keita centered AS_C after he finished unclipping the output arm (alog 15145), we can servo SR2 to this QPD to avoid clipping on the Faraday.  The input and output matrix elements are 1, the output goes to M2, the gain for PIT is 500 and Yaw is 3000.  These are both low bandwidth loops for now.  The hope is that this will make using WFS to control the SRM more robust, as was seen at LLO.  We'll see when we get a chance to lock DRMI.  

As proposed, WFSA (hard sensor) is fed back to the ITMX and WFSB (soft sensor) to the TMSX.

In the attached, you can see that even when I made a terrible kick such that the transmission dropped below 50%, it eventually came back on its own.

Short lock losses don't kick the WFS out of range, and as of now it is maintaining green transmission between 1.15 and 1.18 for 30 min.

ITMX beam position control (fed back to ETMX) is not implemented yet, but should be easier than WFSs themselves.

Eventually we might be able to change the scheme such that hard is fed back to hard, soft to soft, and then ITM beam position to TMS, but I'm doing it this way as it is easier.

Before doing camera, however, I'll move to Y arm.

Initial attempts to take undamped TFs on ITMX & ITMY exhibited rung up P & R modes (see LHO aLOG entry 14653). For the next attempt, fine tuning of excitation amplitudes was necessary to avoid ringing up these modes.

Phase 3b (in-vacuum) undamped TF measurements have been taken for ITMX & ITMY (QUAD) suspensions as follows:-  

- ITMX M0-M0 undamped results (2014-10-30_0700_H1SUSITMX_M0_ALL_TFs.pdf)
- ITMX R0-R0 undamped results (2014-10-30_0700_H1SUSITMX_R0_ALL_TFs.pdf)
- ITMY M0-M0 undamped results (2014-10-28_1200_H1SUSITMY_M0_ALL_TFs.pdf)
- ITMY R0-R0 undamped results (2014-10-28_1200_H1SUSITMY_R0_ALL_TFs.pdf)

ISI Status: ISI's damped and FULLY_ISOLATED via Guardian.

ITMX & ITMY undamped TFs above have been compared with other similar QUADs at the same phase of testing (allquads_2014-10-30_AllQUADS_Doff_Phase3b_ALL_ZOOMED_TFs.pdf). The plot key is as follows:-

Blue Trace = Model Prediction (fiber/thincp).
Orange Trace = L1 ITMX (fiber 2013−09−04), Phase 3b.
Black Trace = L1 ITMY (fiber 2013−09−05), Phase 3b.
Magenta Trace = H1 ITMY (fiber 2014−10−28), Phase 3b.
Cyan Trace = H1 ITMX (fiber 2014−10−30), Phase 3b.

Summary:

M0-M0, main chain TFs are a very good fit to the model, for all DOFs, with only some minor cross-couplings from P2V.

R0-R0, reaction chain TFs agree with the model predictions and are consistent with similar QUADs. The largest deviation from the model can be seen with the ~1.45 Hz P mode, a consequence of the harness routing stiffening the suspension, seen before. Some minor cross-couplings are also present: from P2L, P2R, and P2V only for ITMY. Damped TFs should be taken to verify that damping loops suppress these cross-couplings.

All data, scripts and plots have been committed to the sus svn as of this entry.