DQ shift summary: LHO Sep 3 - 6 (1125273617 - 1125619217)

There were six separate locks during this shift, including a new record lock stretch of 25 hours. Typical inspiral range was ~ 75 Mpc. At least two locklosses were caused by earthquakes. Total observation time 57 hours (duty cycle 59%).

Very loud (SNR ~ 1e3) glitches associated with range drops and ETMY saturations continue (roughly 40 during this shift). The correlation between glitches and saturations was investigated; it was found that every SNR > 1000 Omicron trigger (and most with SNR 100 – 1000) was simultaneous with a range drop and an ETMY saturation. Considering the possibility that the 'dust glitches' and the ETMY saturations are actually the same thing. (More detailed alog on this coming soon; details and follow-up at PreO1WorstGlitches wiki).

Spectrograms showed some excess noise in several broadband lines between 10 and 40 Hz on Thursday, and a set of narrower lines between 10 and 50 Hz on Friday. Cause unknown.

The third observation period on Saturday had a significantly higher strain noise floor and lower range (~ 60 Mpc). Robert suggested that this was high frequency noise in an oscillator ( alog). Not followed up.

The periodic 60 Hz glitch which occurs every 72 minutes continues, with slightly lower SNR than during ER7. They are vetoed very efficiently by H1:SUS-ETMY_L2_WIT_L_DQ.

More details can be found at the DQ shift wiki page.

H1 SUS ETMY PUM coil driver electronics analysis

Jeffrey K, Kiwamu I, Darkhan T

Overview

In this alog we present a summary of H1 SUS ETMY PUM coil driver electronics measurements (LHO alog 20846) analysis. PUM coil driver circuit consist of switchable low-pass filter and acquisition circuits (DCC D070483). Fitted zeros and poles of these circuits are given in Table 1:

                                            Table 1. H1:SUSETMY-PUM driver fit summary
==================================================================================================================================
                       LP1                        Acq=0                Acq=1                 PUM total in state 3 (nominal)
                 ( z : p ) [Hz]              ( z : p ) [Hz]       ( z : p ) [Hz]                     ( z : p ) [Hz]
----------------------------------------------------------------------------------------------------------------------------------
 UL    ( 5.62, 21.41 : 0.50, 240.86 )    ( 11.87 : 119.62 )    ( 1.31 : 106.12 )    ( 5.62, 21.41, 11.87 : 0.50, 240.86, 119.62 )
 LL    ( 5.70, 21.66 : 0.50, 242.82 )    ( 11.81 : 120.07 )    ( 1.30 : 106.42 )    ( 5.70, 21.66, 11.81 : 0.50, 242.82, 120.07 )
 UR    ( 5.74, 21.38 : 0.50, 241.15 )    ( 11.84 : 119.68 )    ( 1.30 : 105.43 )    ( 5.74, 21.38, 11.84 : 0.50, 241.15, 119.68 )
 LR    ( 5.78, 21.09 : 0.50, 239.51 )    ( 12.75 : 118.38 )    ( 1.45 : 106.21 )    ( 5.78, 21.09, 12.75 : 0.50, 239.51, 118.38 )

 mean  ( 5.71, 21.37 : 0.50, 241.00 )    ( 11.97 : 119.56 )    ( 1.34 : 106.04 )    ( 5.71, 21.37, 11.97 : 0.50, 241.00, 119.56 )
==================================================================================================================================

Overall, uncertainties in the weighted means of fitted zero and pole frequencies for low-pass filter are under 0.15%, and for acquisition circuit are under 0.30%, complete table of uncertainties are given in tables 2 and 3.

Fitted values for low-pass filter were within ~1.5 Hz of the values we expected to have according to T1100507 (except for high frequency pole which came out to be at ~240 Hz as opposed to 250 Hz in the document).

Acquisition curcuit TFs showed roll-off (perhaps some combination of poles) after 1 kHz, and roll-off of the phase in acquisition circuit TF is noticable after 100 Hz. Since we don't understand the cause of this roll-off, we tried to elimite this effect from fitting of known zeros and poles (from T1100507) by reducing fitting range to [0.1 80.0] Hz. The results were mostly at the expected frequencies except for high frequency poles: in state 1 we got a pole at ~119 Hz as opposed to 110 Hz suggesed by the document, and in state 2 we got a pole at ~106 Hz as opposed to 80.5 Hz from the document.

Fitted zero-pole-gain models of UIM, PUM, ESD drivers' electronics TFs will be implemented in DARM OLG TF model for ER8/O1. In fact, models of PUM coil-driver electronics TFs from this analysis have been already implemented in DARM OLG TF model for ER8/O1 (see LHO alog 21250).

Details

In this analysis we used measurements of the H1 SUS ETMY drivers from Aug. 24, 2015 (LHO alog 20846). The measured transfer functions between excitation and readout signals, apart from driver itself, include also frequency dependent effects from IOP upsampling, digital anti-imaging, analog anti-imaging, and analog anti-aliasing filters given in the diagram below (also explained in LHO alog comment 21127):

Assuming that the FAST IMON does not have a frequency dependent effects and dividing out the IOP digital anti-imaging, the analog anti-imaging and ani-aliasing filter TFs from the measurements, should give us the PUM coil-driver electroinics TFs for each of the quadrants in 4 different opereating states. Idially the PUM coil driver electronics (DCC D070483) transfer functions in 4 operating states should correspond to a state machine diagram given in T1100507, however LISO zero-pole fitting results of the measured PUM coil driver TFs (see Table 1) did not entirely agree with the diagram.

From the diagram given in T1100507 we can see that measred TFs of the low-pass filter can be isolated from all of the other frequency dependences by taking the TF ratios of State 3 / State 1 and State 4 / State2, and TF ratios of the acquisition circuit at two acq. bit states can be isolated by taking TF ratios of State 2 / State 1 and State 4 / State 3. The consistency between two different configurations (and the residuals between the two) can be seen in figures below:

Low-Pass filter

The fitted zeros and poles of the low-pass filter for each of the quadrants from two different TF ratio combitations (model) are given in table 2; plot of the model zero-pole TF vs. measurement and the residuals are shown in the plots under the table. Fitting was done on a logarithmically spaced frequency vector in range [0.1 1200] Hz.

                             Table 2. H1:SUSETMY-PUM driver LP1 fit details
========================================================================================================
      Measurement                                    Fitted LP1 and fit uncertainty
                                                             ( z : p ) [Hz]
--------------------------------------------------------------------------------------------------------
 UL: St. 3 TF / St. 1 TF     ( 5.71 +/- 0.19 %, 21.20 +/- 0.19 % : 0.50 +/- 0.13 %, 240.90 +/- 0.13 % )
 UL: St. 4 TF / St. 2 TF     ( 5.51 +/- 0.20 %, 21.67 +/- 0.20 % : 0.49 +/- 0.14 %, 240.80 +/- 0.14 % )
 UL:   weighted mean         ( 5.62 +/- 0.14 %, 21.41 +/- 0.14 % : 0.50 +/- 0.09 %, 240.86 +/- 0.09 % )

 LL: St. 3 TF / St. 1 TF     ( 5.77 +/- 0.20 %, 21.54 +/- 0.20 % : 0.51 +/- 0.13 %, 242.55 +/- 0.13 % )
 LL: St. 4 TF / St. 2 TF     ( 5.63 +/- 0.21 %, 21.80 +/- 0.21 % : 0.50 +/- 0.15 %, 243.15 +/- 0.15 % )
 LL:   weighted mean         ( 5.70 +/- 0.14 %, 21.66 +/- 0.14 % : 0.50 +/- 0.10 %, 242.82 +/- 0.10 % )

 UR: St. 3 TF / St. 1 TF     ( 5.81 +/- 0.19 %, 21.22 +/- 0.19 % : 0.51 +/- 0.12 %, 241.16 +/- 0.13 % )
 UR: St. 4 TF / St. 2 TF     ( 5.65 +/- 0.21 %, 21.59 +/- 0.21 % : 0.50 +/- 0.15 %, 241.14 +/- 0.15 % )
 UR:   weighted mean         ( 5.74 +/- 0.14 %, 21.38 +/- 0.14 % : 0.50 +/- 0.10 %, 241.15 +/- 0.10 % )

 LR: St. 3 TF / St. 1 TF     ( 5.94 +/- 0.19 %, 20.58 +/- 0.19 % : 0.50 +/- 0.12 %, 239.00 +/- 0.12 % )
 LR: St. 4 TF / St. 2 TF     ( 5.63 +/- 0.19 %, 21.66 +/- 0.19 % : 0.50 +/- 0.13 %, 240.12 +/- 0.13 % )
 LR:   weighted mean         ( 5.78 +/- 0.14 %, 21.09 +/- 0.14 % : 0.50 +/- 0.09 %, 239.51 +/- 0.09 % )

========================================================================================================

Weighed means of the low-pass filter parameters of each of the quadrants (UL, LL, UR, and LR) from State 3 / State 1 and State 4 / State 2 measurements and weighted mean of all 4 quandrants are given in the figure below:

Acquisition circuit

The same technique of taking the transfer functions of two different states to isolate TFs of acquisition circuit (TF when Acq. bit=1 / TF when Acq. bit=0) showed some extra frequency dependend effect, a pole at about ~1kHz and missing polt at 80.5 Hz (instead the pole was around 110 Hz); we're not sure what's the source of this effect (see attached figures st2/st1, st4/st3). Because we're not sure which of two states (acq.bit=0 / acq.bit=1) causes this frequency dependence we decided not to use measurement in state 1 (not a ratio of measurements in two different states) to get acquisition circuit TF for acq.bit=0 and state 2 for acq.bit=1.

FItted acq. circuit zeros and poles (model) are given in table 3; model vs. measurement and residual plots are given in the plot under the table. Fitting was done on a logarithmically spaced frequency vector in range [0.1 80] Hz.

                     Table 3. H1:SUSETMY-PUM driver acq. circuit fit details
==================================================================================================
               Fitted Acq0 and fit uncertainty                Fitted Acq1 and fit uncertainty
                        ( z : p ) [Hz]                                 ( z : p ) [Hz]
--------------------------------------------------------------------------------------------------
 UL     ( 11.87 +/-  0.05 % : 119.62 +/-  0.10 % )      ( 1.31 +/-  0.21 % : 106.12 +/-  0.27 % )
 LL     ( 11.81 +/-  0.05 % : 120.07 +/-  0.09 % )      ( 1.30 +/-  0.21 % : 106.42 +/-  0.27 % )
 UR     ( 11.84 +/-  0.04 % : 119.68 +/-  0.08 % )      ( 1.30 +/-  0.20 % : 105.43 +/-  0.26 % )
 LR     ( 12.75 +/-  0.06 % : 118.38 +/-  0.12 % )      ( 1.45 +/-  0.20 % : 106.21 +/-  0.26 % )
==================================================================================================

We see that in both cases, acq.bit=0 and acq.bit=1 we see roll-off in the measurements over 1kHz which is not explained in T1100507. We also see that pole at 80.5 Hz did not show up in the measurements, instead we saw a pole at about 110 Hz.

in this analysis we did not try to estimate DC gain of the coil driver TFs, but only frequency dependent effects.

Scripts and Plots in the SVN

Scripts were committed to the calibration SVN:

comparison script for the same TF using two different configurations:

CalSVN/aligocalibration/trunk/Runs/ER8/H1/Scripts/Electronics/compare_PUM_measState3_measState4divState2.m

measurement parameters are in

CalSVN/aligocalibration/trunk/Runs/ER8/H1/Scripts/Electronics/H1SUSETMY_PUMDriver_$(state)_param_$(gps_time).m

the script that loads all of the measurement parameters, calls the fitting function and produces the plots:

CalSVN/aligocalibration/trunk/Runs/ER8/H1/Scripts/Electronics/runFit_H1SUSETMY_PUMDriver.m

all of the Matlab functions used in the analysis are commtted to the same directory

Plots were committed to the calibration SVN:

CalSVN/aligocalibration/trunk/Runs/ER8/H1/Results/Electronics/2015-09-04_H1SUSETMY_PUMDriver_$(short_description).pdf

new script for loading coil mismatch into CALCS

As we recently confirmed, the coil driver analog circuits are not accurately compensated in the suspension front end model (see alog 21189 for EY L3 stage, alog 21189 for EY L1 stage, and Darkhan is making alog about the EY L2 stage). This results in an extra frequency dependent response in the actuators which would be flat if the compensation was perfect. Since we will run the interferometer in this way without accurately compensating the coil driver responses, we need a correction filter to take this into account in CAL-CS. So I made a matlab script to pick up the latest DARMOLGTF model, which already includes the precise measurement of the analog coil drivers, and then install them in CAL-CS_DARM_ANALOG_ETMY filters.

The script can be found at:

aligocalibration/trunk/Runs/ER8/H1/Scripts/CALCS/quack_coilmismatch_into_calcs.m

The script simply loads the latest DARM open loop model by calling a matlab function, H1DARMOLGTFmodel_ER8, and then processes the mismatch into one digital filter for each stage by combining the digital compensation filter and corresponding analog responses. Currently it handles only L2 and L3 stages since we are still working on the L1 stage. Using autoquack, it is going to install the mismatch filters into CALCS. I have ran this code multiple times and seemed to run without an issue so far. The filters are already installed at the end of the actuator chain as seen in the attached plot -- FM6 of each stage is the one. Note that they are not enabled yet since we need to install the latest suspension model in CAL-CS in order to be accurate.

Unused PSL chilled water flow still dominant exciter of PSL periscope in 300 Hz region.

We did not get the factor of 5 reduction in jitter coupling from second loop ISS and PZT servos that I had hoped for, and resonances of the top mirror mount on the periscope are visible in DARM near 300 Hz. Jason and I altered PSL crystal chiller flow rates to see if they affected PSL periscope motion. Figure 1 shows that, indeed, a 2% decrease in flow rate (as low as we could go using the bypass valve) dropped the periscope motion by 8%. Also, a 10% increase in flow gave a 29% increase in periscope motion.

The great majority of the flow is through the unused HPO. A small fraction of the flow is needed for the front-end laser that we are using. We could reduce the flow and probably remove the peaks from DARM if we turn down the total flow while increasing the fraction to the front-end laser.

Robert, Jason

H1 in Observing/Undisturbed Mode

Ed, Sheila, Jeff K.

23:20UTC Observation Intent Bit set. This is following some debate as to the  "wait" time for the IFO to "cool" down before setting the bit.

Shift Summary -Day

ALL TIMES IN UTC

 

ACTIVITY LOG:

• 17:30 Kissel on site

• 18:44 ETMY Saturation

• 19:00 Kiwamu on site

• 19:45  Sheila on site

• 20:22  Sudarshan on site

• 21:02 ETMY Saturation

• 21:13  Wind picked up from 5 to about 15 in the last couple of minutes. No biggie.

• 21:18  Looks like another quake from the same New Zealand area is shaking us causing SRM saturations.

• 21:39  Sheila taking advantage of this Lockloss to try a different PRM offload to be able to ride out earthquakes better.

• 22:36 Jeff B on site for evening shift

• 22:46 Darkham on site

• Handing off to Jeff - Undisturbed

LOCK LOG:

• 16:58 DRMI finally able to lock

• 17:06 Lockloss at SWITCH_TO_QPDS

• 17:21 DRMI locked

• 17:24 Lockloss  @ DRMI_ON_POP

• 17:43  DRMI locked

• 17:56 Locked @ NOMINAL_LOW_NOISE 66Mpc

• 21:33 Lockloss - New Zealand quake

 

Summary:

• Not too much to summarize. The time logs tell the story.

Lockloss

21:33UTC

H1 IFO in Undisturbed/Observing Mode

18:02UTC Intent Bit set

CDS model and DAQ restart report, Saturday and Sunday 5th,6th September 2015

ER8 days 20,21. No restarts reported.

OPS Owl shift summary

After the Earth settled down, I ran through an initial alignment (including dither alignment of TMSy) since the green alignment wasn't very good.  Lots of locklosses on the way up at various places.

11:48 Dan leaves for Italy

12:00 Begin initial alignment

12:08 I notice that ALS_-Y_WFS_DOF_3_P_GAIN and ALS_-Y_WFS_DOF_3_Y_GAIN are flagged in SDF.  I assume this is a result of the dither alignment of TMSy.  I revert them.

13:12 lockloss @ SWITCH_TO_QPDS

13:25 lockloss @ RF_DARM

13:36 lockloss @ DRMI_ON_POP, possibly due to PRM saturation

13:43 lockloss @ SWITCH_TO_QPDS

13:46 lockloss @ FIND_IR

13:54 lockloss @ SWITCH_TO_QPDS

14:13 ISC_LOCK is stalled at DC_READOUT_TRANSITION.  Looking at the code and OMC_LOCK log, I try re-requesting READY_FOR_HANDOFF in OMC_LOCK.  It works, and ISC_LOCK carries on.

14:16 lockloss at ENGAGE_ISS_2ND_LOOP, conicident ETMy saturation

14:18 Seismos start showing more activity.  USGS site reports another 6.2 EQ in New Zealand.  The thunder from down under does us in again.

14:47 Observatory mode set to Earthquake

15:00 Good luck Ed!

 

OMC visibility measurement

Dan, Travis

During the earthquake tonight we finally closed the loop on the HAM6 power budget measurements.  This is something Koji had asked for many, many months ago.  orz

With IMC-PWR_IN at 2.2388 +/- 0.0004 watts (not sure about the calibration for this channel) and a single bounce off ITMY, we collected the following data (30 seconds of slow channels, uncertainties are stdevs) with the OMC locked & unlocked:

	OMC Unlocked	OMC Locked
AS_A_DC_SUM_OUT [cnts]	436.3 +/- 0.6	436.3 +/- 0.7
AS_B_DC_SUM_OUT [cnts]	427.3 +/- 0.07	427.5 +/- 0.7
AS_C_SUM_OUT [cnts]	289.8 +/- 0.2	289.8 +/- 0.3
OMCR_A_SUM_OUT [cnts]	142.1 +/- 0.2	23.3 +/- 0.1
OMCR_B_SUM_OUT [cnts]	150.7 +/- 0.1	22.9 +/- 0.1
DCPD_A_OUT [mA]	--	1.775 +/- 0.0023
DCPD_B_OUT [mA]	--	1.761 +/- 0.0023
DCPD_SUM_OUT [mA]		3.537 +/- 0.0046

 

AS_C has 36dB of whitening gain.  The OMCRs and DCPDs have no DC whitening.  I'm not sure about AS_A and AS_B (I think they have zero whitening, but I haven't looked at the electronics chain for the WFS DC signals so I'm not sure of the transimpedance, etc).  Since AS_C gets 400ppm of the light entering HAM6, the total light into the chamber is:

289 counts / (1638.4 counts/V * 0.8 A/W * 1000 V/A * 2 * 36dB * 400ppm) = 4.38 mW

(PD responsivity = 0.8, QPD transimpedance = 1000 ohms, differential output = 2x, whitening gain = 36dB)

The alignment into the OMC is probably not perfect in this state, but the mode-matching for the carrier should be better than 90% -- same as here, since we apply central heating to ITMX.  The modulation depths are also the same; the power in the carrier should be 94.3% of the total.

OMC dither lines moved above 2kHz

Dan, Travis

Following up on the measurement of the tip-tilt response at high frequencies, I've moved the OMC alignment dither lines up above 2kHz.  The new frequencies are [2125.1, 2150.1, 2175.1, 2200.1] Hz.  The amplitudes were slightly increased to maintain the same line height after the reduction in the tip-tilt response at high frequency.  In the filter banks following the demodulation, I changed the frequencies of the elliptic filters that are used to notch the line frequencies post-demodulation.

To update the control scheme, I adjusted the phase rotation of the demodulated signals such that offsets in the alignment loops only showed up in the I phase signals.  Striptools are the best way to do this.  Then, I reran the ditherDCsense script to calculate a new sensing matrix, and closed the loops.  Things looked pretty stable, so let's stick with this new arrangement.  I attach a screenshot of the new sensing matrix.

 

If for some reason people get spooked and want to revert, do the following:

 - set the frequencies back to what they were before: [P1, P2, Y1, Y2] --> [1675.1, 1700.1, 1725.1, 1750.1]

 - set the oscillator CLK_GAINs to 100.  The current CLK_GAINs are [150, 120, 100, 130].

 - in the dither demodulation, for filter banks {P,Y}{1,2}_X_{SIN,COS}, switch from FM7,9 to FM8,10.  These are pairs of ELPs that notch the dither line in the demodulated signal.  They're used to squash high-frequency stuff from the control signal.  It's probably excessive.

 - set the R demod phases to 90.  They are currently 135, 150, 150, and zero.

 - use conlog to revert the sensing matrix.  Look for changes to channels of the form "H1:OMC-ASC_DACTMAT*".

 

The SDF table has been updated with these new settings.  We were all set to go back to science mode when the IFO lost lock, after 28 hours of good data.  The lockloss was due to PRM M3 hitting the rails (see second plot), maybe because of a 6.4 earthquake, although the lockloss was about twenty minutes before the arrival time of the P-waves according to terramon.

The CW group is appreciative!

J. Kissel

To answer Peter's question, for an original amplitude of 
       Freq    Amp
      [Hz]    [ct]
OM1   2125.1   150
      2150.1   120
OM3   2175.1   100
      2200.1   130
defined by the oscillator's clock gain in the OMC model, we propogatw it to the SUS model to the DRIVEALIGN matrix (which has ~10% off-diagonal elements), then through EULER2OSEM matrix, the DAC gain and the HAM-A coil driver + BOSEM coil transconductance, the current across the coil (assuming a 20/2^16 [V/ct], 16 bit DAC, and a HAM-A coil driver + BOSEM transconductance of 0.988 [mA/V] from T1200264):
Coil       UL          LL         UR          LR
          [mA]        [mA]       [mA]        [mA]
OM1      0.306       0.604      0.604       0.306
OM3      0.096       0.778      0.778       0.096
i.e. a coil current of at most ~0.8 [mA] at these frequencies,

And the displacement this causes on the HTTS (using 0.021 [N/A] from G1100968, a 0.048225 lever arm from D1001428, and the HTTS dynamical model at each frequency from using the httsopt_damp parameter set for the ssmake1MBf production single-stage suspension model):
                   P            Y
                 [rad]        [rad]
OM1   2125.1    3.37e-10    2.74e-10
      2150.1    3.21e-10    2.62e-10

OM3   2175.1    2.52e-10    2.01e-10
      2200.1    2.41e-10    1.92e-10
i.e. a displacement of roughly 3e-10 [rad] in both pitch and yaw at these frequencies.

Mid OWL Shift Summary

Nutsinee handed over an IFO that had been locked for over 28 hours.  Unfortunately, an EQ in New Zealand took us down.  It appears that PRM had saturated, taking down DRMI a few minutes before the EQ arrived, but the EQ has prevented us from relocking for over an hour now as the Earth calms back down. 

7:30 LLO dropped lock, went to commissioning mode for a few minutes so Nutsinee and Dan could look at a few things

9:08 Lockloss (Guardian reports that the OMC SW watchdog has tripped and that DRMI has unlocked)

9:40 I begin initial alignment after seeing that ALS isn't well aligned post lockloss

9:44 After noticing that the alignment is wavering, Dan notices that TerraMon is reporting an EQ in New Zealand.  Site seismos start reporting the same things soon after.  Initial alignment is paused to wait for calmer conditions.

9:44 I notice a timing error on the CDS overview for ETMx.  ETMy already has one which a note from Ed says should be reset after lockloss.

10:53 ETMy timing error reset (as per Ed's note).  ETMx timing error is NOT reset (not sure if I should, emailed DaveB for advice).

EVE Shift Summary

(All time in UTC)

23:00 Take over from Ed. IFO has been locking and observing. Dick's inside optics lab.

01:21 Dick out of the lab.

05:00 Dan's on site. Will be here for the rest of the night.

07:00 The ifo has been locking over 26 hours now! Handing off to Travis.

 

- Quiet shift. Wind still below 20 mph. No seismic activity. Saturation alarm hasn't been complaining about anything else other than ETMY, which has been the only source of big glitches on the DMT Omega.

- Camera 3 control was very slow. At one point I thought it was busted. Seems to be working fine again.

Mid EVE shift Summary

The interferometer is still locking and going strong. Wind below 20 mph. Quiet seismic activity. No one on site but me. I'm hoping for another record breaking lock stretch!

LLO has been locking and observing over the past two hours.

some updated on analysis codes for suspension calibration

I worked a bit more on the analysis codes that Jeff has written (alog 21015 and alog 21049). I was not able to finish deriving the suspension scale factors. Tomorrow, Jeff and Darkhan will pick up them at the point where I left and will continue working on it.

Major updates:

• DARMOLGTFmodel now incorporates not only each ESD driver circuit measurement but also each PUM coil driver measurement.
  • Darkhan is going to alog some detail of the PUM driver measurement and fitting.
• The ALS Diff measurements are now analyzed using the latest DARMOLGTF model.
• Also, the Pcal measurements are now analyzed using the latest DARMOLGTF model.
  • By eyeballing the resultant plots, I see that the inclusion of the ESD driver and PUM coil driver circuits made the model more accurate. This is encouraging.

Next steps:

• We need to include the latest DARMOLGTF model in the free-swinging Michelson analysis.
• UIM driver circuit needs to be updated. DARMOLGTF still assumes a perfect cancelation between the digital and analog filters.

 

Some notes:

In the course of the code update, I changed the file organization structure of the ALS diff and Pcal scripts so that they have only one analysis code which can be invoked by specifying a set of data parameters. This organization approach is the same as those for the DARM open loop analysis.

As for ALS diff, the core analysis code is

aligocalibration/trunk/Runs/ER8/H1/Scripts/ALSDiff/Matlab/analyze_alsdiff.m

and the parameter files are in the same directory and named as

H1ALSDIFFparams_20150826.m

H1ALSDIFFparams_20150828.m

H1ALSDIFFparams_20150829.m

As for Pcal, the core analysis code is

aligocalibration/trunk/Runs/ER8/H1/Scripts/PCAL/analyze_pcal.m

and the parameter files are in the same directory and named as

H1PCALparams_20150826.m

H1PCALparams_20150828.m

H1PCALparams_20150829.m

Each parameter file loads H1DARMOLGTFmodel_ER8 with the latest electronics information and subsequently returns the parameter structure, or the familiar variable "par". As usual the parameter structure can be then fed to the analysis code (e.g. analyze_alsdiff.m or analyze_pcal.m) as an input argument. I have not carefully thought about the final return variables from the core analysis codes, but in principle we can write them such that they return the calibrated suspension transfer functions in meters/counts together with some error bars and also perhaps some statistical values. In this way, the final step of performing apple-to-apple comparison between various measurements from various dates can be less painful.

I tried propagating the same file organization structure to the free-swinging Michelson codes, but apparently I am running out my energy for the night and did not finish it yet.